合并 TURBO-615-RecognitionAndPositioning 到 ResolveConflicts,保留双方冲突内容

This commit is contained in:
李伟
2026-05-20 15:32:43 +08:00
44 changed files with 25278 additions and 89 deletions
File diff suppressed because it is too large Load Diff
Binary file not shown.
+2 -1
View File
@@ -1,4 +1,5 @@
{
"Language": "zh-CN",
"LogLevel": "Debug"
"LogLevel": "Debug",
"CameraType": "Hikvision"
}
+147 -2
View File
@@ -10,6 +10,7 @@ public static class PixelConverter
{
/// <summary>
/// 将原始像素数据转换为 WPF 的 BitmapSource 对象。
/// 支持 Mono8、BGR8、RGB8、BGRA8 以及 Bayer 8-bit 格式(自动解码为 BGR24)。
/// 返回的 BitmapSource 已调用 Freeze(),可跨线程访问。
/// </summary>
public static BitmapSource ToBitmapSource(byte[] pixelData, int width, int height, string pixelFormat)
@@ -19,11 +20,23 @@ public static class PixelConverter
if (height <= 0) throw new ArgumentException("Height must be a positive integer.", nameof(height));
ArgumentNullException.ThrowIfNull(pixelFormat);
var (format, stride) = pixelFormat switch
string normalized = NormalizePixelFormat(pixelFormat);
// Bayer 格式需要解码
if (normalized.StartsWith("Bayer"))
{
byte[] bgrData = DemosaicBayer(pixelData, width, height, normalized);
var bmp = BitmapSource.Create(width, height, 96, 96, PixelFormats.Bgr24, null, bgrData, width * 3);
bmp.Freeze();
return bmp;
}
var (format, stride) = normalized switch
{
"Mono8" => (PixelFormats.Gray8, width),
"BGR8" => (PixelFormats.Bgr24, width * 3),
"BGRA8" => (PixelFormats.Bgra32, width * 4),
"RGB8" => (PixelFormats.Rgb24, width * 3),
_ => throw new NotSupportedException($"Pixel format '{pixelFormat}' is not supported.")
};
@@ -31,4 +44,136 @@ public static class PixelConverter
bitmap.Freeze();
return bitmap;
}
}
/// <summary>
/// 将不同 SDK 的像素格式名称统一为标准名称。
/// </summary>
private static string NormalizePixelFormat(string pixelFormat)
{
if (pixelFormat is "Mono8" or "BGR8" or "BGRA8" or "RGB8")
return pixelFormat;
var upper = pixelFormat.ToUpperInvariant();
if (upper.Contains("MONO8")) return "Mono8";
if (upper.Contains("BGR8")) return "BGR8";
if (upper.Contains("BGRA8")) return "BGRA8";
if (upper.Contains("RGB8") && !upper.Contains("BAYER")) return "RGB8";
// Bayer 格式
if (upper.Contains("BAYERRG8") || upper.Contains("BAYER_RG8")) return "BayerRG8";
if (upper.Contains("BAYERGR8") || upper.Contains("BAYER_GR8")) return "BayerGR8";
if (upper.Contains("BAYERGB8") || upper.Contains("BAYER_GB8")) return "BayerGB8";
if (upper.Contains("BAYERBG8") || upper.Contains("BAYER_BG8")) return "BayerBG8";
return pixelFormat;
}
/// <summary>
/// 简单 Bayer 解码(双线性插值),输出 BGR24。
/// </summary>
private static byte[] DemosaicBayer(byte[] bayer, int width, int height, string pattern)
{
// pattern: BayerRG8, BayerGR8, BayerGB8, BayerBG8
// RG: R G GR: G R GB: G B BG: B G
// G B B G R G G R
int rRow, rCol; // 红色像素在2x2块中的位置
switch (pattern)
{
case "BayerRG8": rRow = 0; rCol = 0; break;
case "BayerGR8": rRow = 0; rCol = 1; break;
case "BayerGB8": rRow = 1; rCol = 0; break;
case "BayerBG8": rRow = 1; rCol = 1; break;
default: rRow = 0; rCol = 0; break;
}
byte[] bgr = new byte[width * height * 3];
for (int y = 0; y < height; y++)
{
for (int x = 0; x < width; x++)
{
int srcIdx = y * width + x;
int dstIdx = (y * width + x) * 3;
// 确定当前像素在 Bayer 模式中的角色
int py = (y + rRow) % 2; // 0=红行, 1=蓝行
int px = (x + rCol) % 2; // 0=红列/蓝列, 1=绿列
byte r, g, b;
if (py == 0 && px == 0)
{
// 红色像素位置
r = bayer[srcIdx];
g = AvgNeighbors4(bayer, width, height, x, y);
b = AvgDiagonal(bayer, width, height, x, y);
}
else if (py == 1 && px == 1)
{
// 蓝色像素位置
b = bayer[srcIdx];
g = AvgNeighbors4(bayer, width, height, x, y);
r = AvgDiagonal(bayer, width, height, x, y);
}
else if (py == 0 && px == 1)
{
// 绿色像素(红行)
g = bayer[srcIdx];
r = AvgHorizontal(bayer, width, x, y);
b = AvgVertical(bayer, width, height, x, y);
}
else
{
// 绿色像素(蓝行)
g = bayer[srcIdx];
b = AvgHorizontal(bayer, width, x, y);
r = AvgVertical(bayer, width, height, x, y);
}
bgr[dstIdx] = b;
bgr[dstIdx + 1] = g;
bgr[dstIdx + 2] = r;
}
}
return bgr;
}
private static byte AvgNeighbors4(byte[] data, int w, int h, int x, int y)
{
int sum = 0, count = 0;
if (x > 0) { sum += data[y * w + x - 1]; count++; }
if (x < w - 1) { sum += data[y * w + x + 1]; count++; }
if (y > 0) { sum += data[(y - 1) * w + x]; count++; }
if (y < h - 1) { sum += data[(y + 1) * w + x]; count++; }
return count > 0 ? (byte)(sum / count) : (byte)0;
}
private static byte AvgDiagonal(byte[] data, int w, int h, int x, int y)
{
int sum = 0, count = 0;
if (x > 0 && y > 0) { sum += data[(y - 1) * w + x - 1]; count++; }
if (x < w - 1 && y > 0) { sum += data[(y - 1) * w + x + 1]; count++; }
if (x > 0 && y < h - 1) { sum += data[(y + 1) * w + x - 1]; count++; }
if (x < w - 1 && y < h - 1) { sum += data[(y + 1) * w + x + 1]; count++; }
return count > 0 ? (byte)(sum / count) : (byte)0;
}
private static byte AvgHorizontal(byte[] data, int w, int x, int y)
{
int sum = 0, count = 0;
if (x > 0) { sum += data[y * w + x - 1]; count++; }
if (x < w - 1) { sum += data[y * w + x + 1]; count++; }
return count > 0 ? (byte)(sum / count) : (byte)0;
}
private static byte AvgVertical(byte[] data, int w, int h, int x, int y)
{
int sum = 0, count = 0;
if (y > 0) { sum += data[(y - 1) * w + x]; count++; }
if (y < h - 1) { sum += data[(y + 1) * w + x]; count++; }
return count > 0 ? (byte)(sum / count) : (byte)0;
}
}
+2 -2
View File
@@ -11,8 +11,8 @@ public class CameraFactory : ICameraFactory
return cameraType switch
{
"Basler" => new BaslerCameraController(),
// "Hikvision" => new HikvisionCameraController(),
_ => throw new NotSupportedException($"不支持的相机品牌: {cameraType}")
"Hikvision" => new HikvisionCameraController(),
_ => throw new NotSupportedException($"Unsupported Camera Type: {cameraType}")
};
}
}
@@ -0,0 +1,535 @@
using MvCameraControl;
using Serilog;
namespace XP.Camera;
/// <summary>
/// 海康威视相机控制器,封装 MvCameraControl.Net SDK 实现 <see cref="ICameraController"/>。
/// </summary>
/// <remarks>
/// <para>所有公共方法通过内部 <c>_syncLock</c> 对象进行 lock 同步,保证线程安全。</para>
/// <para>事件回调(ImageGrabbed、GrabError)在 SDK 回调线程上触发,不持有 _syncLock,避免死锁。</para>
/// </remarks>
public class HikvisionCameraController : ICameraController
{
private static readonly ILogger _logger = Log.ForContext<HikvisionCameraController>();
private static bool _sdkInitialized;
private static readonly object _sdkInitLock = new();
private readonly object _syncLock = new();
private IDevice? _device;
private CameraInfo? _cachedCameraInfo;
private bool _isConnected;
private bool _isGrabbing;
public HikvisionCameraController()
{
// SDK 初始化延迟到 Open() 中执行
}
/// <inheritdoc />
public bool IsConnected
{
get { lock (_syncLock) { return _isConnected; } }
}
/// <inheritdoc />
public bool IsGrabbing
{
get { lock (_syncLock) { return _isGrabbing; } }
}
/// <inheritdoc />
public event EventHandler<ImageGrabbedEventArgs>? ImageGrabbed;
/// <inheritdoc />
public event EventHandler<GrabErrorEventArgs>? GrabError;
/// <inheritdoc />
public event EventHandler? ConnectionLost;
/// <inheritdoc />
public CameraInfo Open()
{
lock (_syncLock)
{
if (_isConnected && _cachedCameraInfo != null)
{
_logger.Information("Hikvision camera already connected, returning cached info.");
return _cachedCameraInfo;
}
try
{
_logger.Information("Opening Hikvision camera connection...");
// 确保 SDK 初始化
EnsureSdkInitialized();
// 枚举设备
DeviceTLayerType layerType = DeviceTLayerType.MvGigEDevice
| DeviceTLayerType.MvUsbDevice;
List<IDeviceInfo> deviceInfoList;
int ret = DeviceEnumerator.EnumDevices(layerType, out deviceInfoList);
_logger.Information("EnumDevices(GigE|USB) returned: 0x{RetCode:X8}, device count: {Count}",
ret, deviceInfoList?.Count ?? 0);
// 如果没找到,分别尝试
if (ret == MvError.MV_OK && (deviceInfoList == null || deviceInfoList.Count == 0))
{
// 单独尝试 GigE
List<IDeviceInfo> gigeList;
int retGige = DeviceEnumerator.EnumDevices(DeviceTLayerType.MvGigEDevice, out gigeList);
_logger.Information("EnumDevices(GigE only) returned: 0x{RetCode:X8}, count: {Count}",
retGige, gigeList?.Count ?? 0);
// 单独尝试 USB
List<IDeviceInfo> usbList;
int retUsb = DeviceEnumerator.EnumDevices(DeviceTLayerType.MvUsbDevice, out usbList);
_logger.Information("EnumDevices(USB only) returned: 0x{RetCode:X8}, count: {Count}",
retUsb, usbList?.Count ?? 0);
// 合并结果
deviceInfoList = new List<IDeviceInfo>();
if (gigeList != null) deviceInfoList.AddRange(gigeList);
if (usbList != null) deviceInfoList.AddRange(usbList);
}
if (ret != MvError.MV_OK)
{
throw new CameraException($"Enumerate Hikvision devices failed: 0x{ret:X8}");
}
if (deviceInfoList == null || deviceInfoList.Count == 0)
{
throw new DeviceNotFoundException("No Hikvision camera device found.");
}
// 选择第一个设备
IDeviceInfo deviceInfo = deviceInfoList[0];
_logger.Information("Found Hikvision device: {Model} (SN: {Serial})",
deviceInfo.ModelName, deviceInfo.SerialNumber);
// 创建设备
_device = DeviceFactory.CreateDevice(deviceInfo);
// 打开设备
ret = _device.Open();
if (ret != MvError.MV_OK)
{
_device.Dispose();
_device = null;
throw new CameraException($"Open Hikvision device failed: 0x{ret:X8}");
}
// GigE 设备优化包大小
if (_device is IGigEDevice gigEDevice)
{
int packetSize;
ret = gigEDevice.GetOptimalPacketSize(out packetSize);
if (ret == MvError.MV_OK && packetSize > 0)
{
_device.Parameters.SetIntValue("GevSCPSPacketSize", packetSize);
_logger.Debug("Set GigE packet size to {PacketSize}", packetSize);
}
}
// 配置软件触发模式
_device.Parameters.SetEnumValueByString("TriggerMode", "On");
_device.Parameters.SetEnumValueByString("TriggerSource", "Software");
// 彩色相机:尝试设置输出为 BGR8 以便直接显示
// 如果相机不支持 BGR8(如只支持 Bayer),则保持默认
int fmtRet = _device.Parameters.SetEnumValueByString("PixelFormat", "BGR8Packed");
if (fmtRet != MvError.MV_OK)
{
// 尝试 Mono8(黑白相机)
fmtRet = _device.Parameters.SetEnumValueByString("PixelFormat", "Mono8");
}
_logger.Debug("Set PixelFormat result: 0x{Ret:X8}", fmtRet);
_cachedCameraInfo = new CameraInfo(
ModelName: deviceInfo.ModelName ?? "",
SerialNumber: deviceInfo.SerialNumber ?? "",
VendorName: deviceInfo.ManufacturerName ?? "",
DeviceType: deviceInfo.TLayerType.ToString()
);
_isConnected = true;
_logger.Information("Hikvision camera connected: {ModelName} (SN: {SerialNumber})",
_cachedCameraInfo.ModelName, _cachedCameraInfo.SerialNumber);
return _cachedCameraInfo;
}
catch (Exception ex) when (ex is not CameraException)
{
_device?.Dispose();
_device = null;
_logger.Error(ex, "Failed to open Hikvision camera.");
throw new CameraException("Failed to open Hikvision camera device.", ex);
}
}
}
/// <inheritdoc />
public void Close()
{
lock (_syncLock)
{
if (!_isConnected)
{
_logger.Information("Hikvision camera not connected, Close() ignored.");
return;
}
try
{
if (_isGrabbing)
{
StopGrabbingInternal();
}
_logger.Information("Closing Hikvision camera connection...");
_device?.Close();
_device?.Dispose();
_device = null;
_isConnected = false;
_cachedCameraInfo = null;
_logger.Information("Hikvision camera connection closed.");
}
catch (Exception ex) when (ex is not CameraException)
{
_device = null;
_isConnected = false;
_isGrabbing = false;
_cachedCameraInfo = null;
_logger.Error(ex, "Error while closing Hikvision camera.");
throw new CameraException("Failed to close Hikvision camera device.", ex);
}
}
}
/// <inheritdoc />
public void StartGrabbing()
{
lock (_syncLock)
{
EnsureConnected();
if (_isGrabbing)
{
_logger.Information("Already grabbing, StartGrabbing() ignored.");
return;
}
try
{
_logger.Information("Starting Hikvision grabbing with software trigger...");
// 设置缓存节点数
_device!.StreamGrabber.SetImageNodeNum(5);
// 注册回调
_device.StreamGrabber.FrameGrabedEvent += OnFrameGrabbed;
// 开始采集
int ret = _device.StreamGrabber.StartGrabbing();
if (ret != MvError.MV_OK)
{
_device.StreamGrabber.FrameGrabedEvent -= OnFrameGrabbed;
throw new CameraException($"Start grabbing failed: 0x{ret:X8}");
}
_isGrabbing = true;
_logger.Information("Hikvision grabbing started.");
}
catch (Exception ex) when (ex is not CameraException)
{
_logger.Error(ex, "Failed to start Hikvision grabbing.");
throw new CameraException("Failed to start grabbing.", ex);
}
}
}
/// <inheritdoc />
public void ExecuteSoftwareTrigger()
{
lock (_syncLock)
{
if (!_isGrabbing)
{
throw new InvalidOperationException("Cannot execute software trigger: camera is not grabbing.");
}
try
{
int ret = _device!.Parameters.SetCommandValue("TriggerSoftware");
if (ret != MvError.MV_OK)
{
throw new CameraException($"Execute software trigger failed: 0x{ret:X8}");
}
}
catch (Exception ex) when (ex is not CameraException and not InvalidOperationException)
{
_logger.Error(ex, "Failed to execute software trigger.");
throw new CameraException("Failed to execute software trigger.", ex);
}
}
}
/// <inheritdoc />
public void StopGrabbing()
{
lock (_syncLock)
{
if (!_isGrabbing) return;
StopGrabbingInternal();
}
}
/// <inheritdoc />
public double GetExposureTime()
{
lock (_syncLock)
{
EnsureConnected();
IFloatValue floatValue;
int ret = _device!.Parameters.GetFloatValue("ExposureTime", out floatValue);
if (ret != MvError.MV_OK)
throw new CameraException($"Get ExposureTime failed: 0x{ret:X8}");
return floatValue.CurValue;
}
}
/// <inheritdoc />
public void SetExposureTime(double microseconds)
{
lock (_syncLock)
{
EnsureConnected();
// 关闭自动曝光
_device!.Parameters.SetEnumValueByString("ExposureAuto", "Off");
int ret = _device.Parameters.SetFloatValue("ExposureTime", (float)microseconds);
if (ret != MvError.MV_OK)
throw new CameraException($"Set ExposureTime failed: 0x{ret:X8}");
_logger.Information("Hikvision exposure time set to {Microseconds} µs.", microseconds);
}
}
/// <inheritdoc />
public double GetGain()
{
lock (_syncLock)
{
EnsureConnected();
IFloatValue floatValue;
int ret = _device!.Parameters.GetFloatValue("Gain", out floatValue);
if (ret != MvError.MV_OK)
throw new CameraException($"Get Gain failed: 0x{ret:X8}");
return floatValue.CurValue;
}
}
/// <inheritdoc />
public void SetGain(double value)
{
lock (_syncLock)
{
EnsureConnected();
_device!.Parameters.SetEnumValueByString("GainAuto", "Off");
int ret = _device.Parameters.SetFloatValue("Gain", (float)value);
if (ret != MvError.MV_OK)
throw new CameraException($"Set Gain failed: 0x{ret:X8}");
_logger.Information("Hikvision gain set to {Value}.", value);
}
}
/// <inheritdoc />
public int GetWidth()
{
lock (_syncLock)
{
EnsureConnected();
IIntValue intValue;
int ret = _device!.Parameters.GetIntValue("Width", out intValue);
if (ret != MvError.MV_OK)
throw new CameraException($"Get Width failed: 0x{ret:X8}");
return (int)intValue.CurValue;
}
}
/// <inheritdoc />
public void SetWidth(int value)
{
lock (_syncLock)
{
EnsureConnected();
int ret = _device!.Parameters.SetIntValue("Width", value);
if (ret != MvError.MV_OK)
throw new CameraException($"Set Width failed: 0x{ret:X8}");
_logger.Information("Hikvision width set to {Value}.", value);
}
}
/// <inheritdoc />
public int GetHeight()
{
lock (_syncLock)
{
EnsureConnected();
IIntValue intValue;
int ret = _device!.Parameters.GetIntValue("Height", out intValue);
if (ret != MvError.MV_OK)
throw new CameraException($"Get Height failed: 0x{ret:X8}");
return (int)intValue.CurValue;
}
}
/// <inheritdoc />
public void SetHeight(int value)
{
lock (_syncLock)
{
EnsureConnected();
int ret = _device!.Parameters.SetIntValue("Height", value);
if (ret != MvError.MV_OK)
throw new CameraException($"Set Height failed: 0x{ret:X8}");
_logger.Information("Hikvision height set to {Value}.", value);
}
}
/// <inheritdoc />
public string GetPixelFormat()
{
lock (_syncLock)
{
EnsureConnected();
IEnumValue enumValue;
int ret = _device!.Parameters.GetEnumValue("PixelFormat", out enumValue);
if (ret != MvError.MV_OK)
throw new CameraException($"Get PixelFormat failed: 0x{ret:X8}");
return enumValue.CurEnumEntry.Symbolic;
}
}
/// <inheritdoc />
public void SetPixelFormat(string format)
{
lock (_syncLock)
{
EnsureConnected();
int ret = _device!.Parameters.SetEnumValueByString("PixelFormat", format);
if (ret != MvError.MV_OK)
throw new CameraException($"Set PixelFormat failed: 0x{ret:X8}");
_logger.Information("Hikvision pixel format set to {Format}.", format);
}
}
/// <inheritdoc />
public void Dispose()
{
Close();
GC.SuppressFinalize(this);
}
// ══════════════════════════════════════════════════════════════
// 私有方法
// ══════════════════════════════════════════════════════════════
/// <summary>
/// SDK 回调:图像采集完成
/// </summary>
private void OnFrameGrabbed(object? sender, FrameGrabbedEventArgs e)
{
try
{
var frameOut = e.FrameOut;
if (frameOut == null || frameOut.Image == null)
{
_logger.Warning("Hikvision OnFrameGrabbed: FrameOut or Image is null");
GrabError?.Invoke(this, new GrabErrorEventArgs(-1, "FrameOut or Image is null."));
return;
}
var image = frameOut.Image;
int width = (int)image.Width;
int height = (int)image.Height;
int imageSize = (int)image.ImageSize;
string pixelFormat = image.PixelType.ToString();
// 提取像素数据
byte[] pixelData = image.PixelData ?? Array.Empty<byte>();
_logger.Debug("Hikvision frame: {Width}x{Height}, format={Format}, dataLen={Len}",
width, height, pixelFormat, pixelData.Length);
if (pixelData.Length == 0)
{
_logger.Warning("Hikvision OnFrameGrabbed: PixelData is empty");
return;
}
var args = new ImageGrabbedEventArgs(pixelData, width, height, pixelFormat);
ImageGrabbed?.Invoke(this, args);
}
catch (Exception ex)
{
_logger.Error(ex, "Exception in Hikvision OnFrameGrabbed handler.");
}
}
private void StopGrabbingInternal()
{
if (!_isGrabbing) return;
try
{
_device?.StreamGrabber.StopGrabbing();
if (_device != null)
_device.StreamGrabber.FrameGrabedEvent -= OnFrameGrabbed;
_isGrabbing = false;
_logger.Information("Hikvision grabbing stopped.");
}
catch (Exception ex) when (ex is not CameraException)
{
_isGrabbing = false;
_logger.Error(ex, "Error while stopping Hikvision grabbing.");
throw new CameraException("Failed to stop grabbing.", ex);
}
}
private void EnsureConnected()
{
if (!_isConnected)
throw new InvalidOperationException("Hikvision camera is not connected. Call Open() first.");
}
/// <summary>
/// 确保 SDK 全局初始化(只调用一次)
/// </summary>
private static void EnsureSdkInitialized()
{
if (_sdkInitialized) return;
lock (_sdkInitLock)
{
if (_sdkInitialized) return;
try
{
int ret = SDKSystem.Initialize();
if (ret != MvError.MV_OK)
{
_logger.Error("Hikvision SDK Initialize failed: 0x{ErrorCode:X8}", ret);
throw new CameraException($"Hikvision SDK Initialize failed: 0x{ret:X8}");
}
_sdkInitialized = true;
_logger.Information("Hikvision SDK initialized successfully.");
}
catch (Exception ex) when (ex is not CameraException)
{
_logger.Error(ex, "Failed to initialize Hikvision SDK.");
throw new CameraException("Failed to initialize Hikvision SDK.", ex);
}
}
}
}
+7
View File
@@ -7,12 +7,19 @@
<Nullable>enable</Nullable>
<RootNamespace>XP.Camera</RootNamespace>
<AssemblyName>XP.Camera</AssemblyName>
<AutoGenerateBindingRedirects>true</AutoGenerateBindingRedirects>
<GenerateBindingRedirectsOutputType>true</GenerateBindingRedirectsOutputType>
</PropertyGroup>
<ItemGroup>
<Reference Include="Basler.Pylon">
<HintPath>..\ExternalLibraries\Basler.Pylon.dll</HintPath>
</Reference>
<Reference Include="MvCameraControl.Net">
<HintPath>..\ExternalLibraries\MvCameraControl.Net.dll</HintPath>
<Private>true</Private>
<CopyLocal>true</CopyLocal>
</Reference>
<PackageReference Include="Emgu.CV" Version="4.10.0.5680" />
<PackageReference Include="Emgu.CV.Bitmap" Version="4.10.0.5680" />
<PackageReference Include="Prism.DryIoc" Version="9.0.537" />
+118
View File
@@ -1887,4 +1887,122 @@ Reprojection error: {1:F4} pixels</value>
<data name="ChessboardImageError" xml:space="preserve">
<value>Image{0}: {1:F4} pixels</value>
</data>
<!-- EdgeLineFitProcessor -->
<data name="EdgeLineFitProcessor_Name" xml:space="preserve">
<value>Edge Find Line Fit</value>
</data>
<data name="EdgeLineFitProcessor_Description" xml:space="preserve">
<value>Place calipers along a search line to detect edge points and fit a line (supports Least Squares and RANSAC)</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount" xml:space="preserve">
<value>Caliper Count</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>Number of calipers placed evenly along the search line</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth" xml:space="preserve">
<value>Caliper Width</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>Search length of each caliper (pixels), perpendicular to the search line</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity" xml:space="preserve">
<value>Edge Polarity</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>Edge direction: BrightToDark, DarkToBright, or Both</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold" xml:space="preserve">
<value>Edge Threshold</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>Gradient strength threshold; edges below this value are ignored</value>
</data>
<data name="EdgeLineFitProcessor_Sigma" xml:space="preserve">
<value>Smoothing Sigma</value>
</data>
<data name="EdgeLineFitProcessor_Sigma_Desc" xml:space="preserve">
<value>Gaussian smoothing standard deviation for noise suppression (larger = smoother)</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod" xml:space="preserve">
<value>Fit Method</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>Line fitting algorithm: LeastSquares or RANSAC (robust, rejects outliers)</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC Threshold</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC inlier distance threshold (pixels); points closer than this to the line are inliers</value>
</data>
<data name="EdgeLineFitProcessor_Thickness" xml:space="preserve">
<value>Line Thickness</value>
</data>
<data name="EdgeLineFitProcessor_Thickness_Desc" xml:space="preserve">
<value>Drawing thickness for result visualization</value>
</data>
<!-- EdgeCircleFitProcessor -->
<data name="EdgeCircleFitProcessor_Name" xml:space="preserve">
<value>Edge Find Circle Fit</value>
</data>
<data name="EdgeCircleFitProcessor_Description" xml:space="preserve">
<value>Place calipers along estimated circle to detect edge points and fit a circle (supports Least Squares and RANSAC)</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount" xml:space="preserve">
<value>Caliper Count</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>Number of calipers placed evenly around the circle</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth" xml:space="preserve">
<value>Caliper Width</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>Search length of each caliper along radial direction (pixels)</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity" xml:space="preserve">
<value>Edge Polarity</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>Edge direction: BrightToDark, DarkToBright, or Both</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold" xml:space="preserve">
<value>Edge Threshold</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>Gradient strength threshold; edges below this value are ignored</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma" xml:space="preserve">
<value>Smoothing Sigma</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma_Desc" xml:space="preserve">
<value>Gaussian smoothing standard deviation for noise suppression</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection" xml:space="preserve">
<value>Search Direction</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection_Desc" xml:space="preserve">
<value>Caliper search direction: Inward (toward center), Outward (away from center), Both</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod" xml:space="preserve">
<value>Fit Method</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>Circle fitting algorithm: LeastSquares or RANSAC (robust, rejects outliers)</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC Threshold</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC inlier distance threshold (pixels); points closer than this to the circle are inliers</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness" xml:space="preserve">
<value>Line Thickness</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness_Desc" xml:space="preserve">
<value>Drawing thickness for result visualization</value>
</data>
</root>
+118
View File
@@ -1920,4 +1920,122 @@
<data name="ChessboardImageError" xml:space="preserve">
<value>图像{0}: {1:F4} 像素</value>
</data>
<!-- EdgeLineFitProcessor -->
<data name="EdgeLineFitProcessor_Name" xml:space="preserve">
<value>边缘查找拟合直线</value>
</data>
<data name="EdgeLineFitProcessor_Description" xml:space="preserve">
<value>沿搜索线放置卡尺检测边缘点,拟合直线(支持最小二乘和RANSAC</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount" xml:space="preserve">
<value>卡尺数量</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>沿搜索线等间距放置的卡尺数量</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth" xml:space="preserve">
<value>卡尺宽度</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>每个卡尺的搜索长度(像素),沿垂直于搜索线方向</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity" xml:space="preserve">
<value>边缘极性</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>边缘方向:BrightToDark(亮到暗)、DarkToBright(暗到亮)、Both(双向)</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold" xml:space="preserve">
<value>边缘阈值</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>边缘梯度强度阈值,低于此值的边缘将被忽略</value>
</data>
<data name="EdgeLineFitProcessor_Sigma" xml:space="preserve">
<value>平滑Sigma</value>
</data>
<data name="EdgeLineFitProcessor_Sigma_Desc" xml:space="preserve">
<value>高斯平滑的标准差,用于抑制噪声(越大越平滑)</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod" xml:space="preserve">
<value>拟合方法</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>直线拟合算法:LeastSquares(最小二乘)、RANSAC(鲁棒拟合,可剔除异常点)</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC阈值</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC内点判定距离阈值(像素),点到直线距离小于此值视为内点</value>
</data>
<data name="EdgeLineFitProcessor_Thickness" xml:space="preserve">
<value>线条粗细</value>
</data>
<data name="EdgeLineFitProcessor_Thickness_Desc" xml:space="preserve">
<value>绘制结果的线条粗细</value>
</data>
<!-- EdgeCircleFitProcessor -->
<data name="EdgeCircleFitProcessor_Name" xml:space="preserve">
<value>边缘查找拟合圆</value>
</data>
<data name="EdgeCircleFitProcessor_Description" xml:space="preserve">
<value>沿预估圆周放置卡尺检测边缘点,拟合圆(支持最小二乘和RANSAC</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount" xml:space="preserve">
<value>卡尺数量</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>沿圆周等角度放置的卡尺数量</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth" xml:space="preserve">
<value>卡尺宽度</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>每个卡尺沿径向的搜索长度(像素)</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity" xml:space="preserve">
<value>边缘极性</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>边缘方向:BrightToDark(亮到暗)、DarkToBright(暗到亮)、Both(双向)</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold" xml:space="preserve">
<value>边缘阈值</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>边缘梯度强度阈值,低于此值的边缘将被忽略</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma" xml:space="preserve">
<value>平滑Sigma</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma_Desc" xml:space="preserve">
<value>高斯平滑的标准差,用于抑制噪声</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection" xml:space="preserve">
<value>搜索方向</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection_Desc" xml:space="preserve">
<value>卡尺搜索方向:Inward(向圆心)、Outward(背离圆心)、Both(双向)</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod" xml:space="preserve">
<value>拟合方法</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>圆拟合算法:LeastSquares(最小二乘)、RANSAC(鲁棒拟合)</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC阈值</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC内点判定距离阈值(像素),点到圆周距离小于此值视为内点</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness" xml:space="preserve">
<value>线条粗细</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness_Desc" xml:space="preserve">
<value>绘制结果的线条粗细</value>
</data>
</root>
+118
View File
@@ -1881,4 +1881,122 @@
<data name="ChessboardImageError" xml:space="preserve">
<value>图像{0}: {1:F4} 像素</value>
</data>
<!-- EdgeLineFitProcessor -->
<data name="EdgeLineFitProcessor_Name" xml:space="preserve">
<value>边缘查找拟合直线</value>
</data>
<data name="EdgeLineFitProcessor_Description" xml:space="preserve">
<value>沿搜索线放置卡尺检测边缘点,拟合直线(支持最小二乘和RANSAC</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount" xml:space="preserve">
<value>卡尺数量</value>
</data>
<data name="EdgeLineFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>沿搜索线等间距放置的卡尺数量</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth" xml:space="preserve">
<value>卡尺宽度</value>
</data>
<data name="EdgeLineFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>每个卡尺的搜索长度(像素),沿垂直于搜索线方向</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity" xml:space="preserve">
<value>边缘极性</value>
</data>
<data name="EdgeLineFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>边缘方向:BrightToDark(亮到暗)、DarkToBright(暗到亮)、Both(双向)</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold" xml:space="preserve">
<value>边缘阈值</value>
</data>
<data name="EdgeLineFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>边缘梯度强度阈值,低于此值的边缘将被忽略</value>
</data>
<data name="EdgeLineFitProcessor_Sigma" xml:space="preserve">
<value>平滑Sigma</value>
</data>
<data name="EdgeLineFitProcessor_Sigma_Desc" xml:space="preserve">
<value>高斯平滑的标准差,用于抑制噪声(越大越平滑)</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod" xml:space="preserve">
<value>拟合方法</value>
</data>
<data name="EdgeLineFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>直线拟合算法:LeastSquares(最小二乘)、RANSAC(鲁棒拟合,可剔除异常点)</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC阈值</value>
</data>
<data name="EdgeLineFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC内点判定距离阈值(像素),点到直线距离小于此值视为内点</value>
</data>
<data name="EdgeLineFitProcessor_Thickness" xml:space="preserve">
<value>线条粗细</value>
</data>
<data name="EdgeLineFitProcessor_Thickness_Desc" xml:space="preserve">
<value>绘制结果的线条粗细</value>
</data>
<!-- EdgeCircleFitProcessor -->
<data name="EdgeCircleFitProcessor_Name" xml:space="preserve">
<value>边缘查找拟合圆</value>
</data>
<data name="EdgeCircleFitProcessor_Description" xml:space="preserve">
<value>沿预估圆周放置卡尺检测边缘点,拟合圆(支持最小二乘和RANSAC</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount" xml:space="preserve">
<value>卡尺数量</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperCount_Desc" xml:space="preserve">
<value>沿圆周等角度放置的卡尺数量</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth" xml:space="preserve">
<value>卡尺宽度</value>
</data>
<data name="EdgeCircleFitProcessor_CaliperWidth_Desc" xml:space="preserve">
<value>每个卡尺沿径向的搜索长度(像素)</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity" xml:space="preserve">
<value>边缘极性</value>
</data>
<data name="EdgeCircleFitProcessor_EdgePolarity_Desc" xml:space="preserve">
<value>边缘方向:BrightToDark(亮到暗)、DarkToBright(暗到亮)、Both(双向)</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold" xml:space="preserve">
<value>边缘阈值</value>
</data>
<data name="EdgeCircleFitProcessor_EdgeThreshold_Desc" xml:space="preserve">
<value>边缘梯度强度阈值,低于此值的边缘将被忽略</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma" xml:space="preserve">
<value>平滑Sigma</value>
</data>
<data name="EdgeCircleFitProcessor_Sigma_Desc" xml:space="preserve">
<value>高斯平滑的标准差,用于抑制噪声</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection" xml:space="preserve">
<value>搜索方向</value>
</data>
<data name="EdgeCircleFitProcessor_SearchDirection_Desc" xml:space="preserve">
<value>卡尺搜索方向:Inward(向圆心)、Outward(背离圆心)、Both(双向)</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod" xml:space="preserve">
<value>拟合方法</value>
</data>
<data name="EdgeCircleFitProcessor_FitMethod_Desc" xml:space="preserve">
<value>圆拟合算法:LeastSquares(最小二乘)、RANSAC(鲁棒拟合)</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold" xml:space="preserve">
<value>RANSAC阈值</value>
</data>
<data name="EdgeCircleFitProcessor_RansacThreshold_Desc" xml:space="preserve">
<value>RANSAC内点判定距离阈值(像素),点到圆周距离小于此值视为内点</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness" xml:space="preserve">
<value>线条粗细</value>
</data>
<data name="EdgeCircleFitProcessor_Thickness_Desc" xml:space="preserve">
<value>绘制结果的线条粗细</value>
</data>
</root>
@@ -0,0 +1,21 @@
namespace XP.ImageProcessing.Core.Alignment;
/// <summary>
/// 示教阶段保存的对齐配方:基准位姿 + 示教图像素坐标下的检测 ROI。
/// </summary>
public sealed class AlignmentRecipe
{
/// <summary>示教图上的基准位姿(建议示教图自匹配得到,或与模板 ROI 中心 + 角度 0 一致)。</summary>
public Pose2D ReferencePose { get; set; }
/// <summary>示教图上的 ROI 多边形顶点(至少 3 点)。</summary>
public List<Point2D> RoiPoints { get; set; } = new();
/// <summary>将示教 ROI 变换到运行图坐标。</summary>
public Point2D[] TransformRoi(Pose2D measuredPose)
=> RoiAlignment.TransformPolygon(RoiPoints, ReferencePose, measuredPose);
/// <summary>变换为整型顶点,供检测算子注入。</summary>
public (int X, int Y)[] TransformRoiToInt(Pose2D measuredPose)
=> RoiAlignment.TransformPolygonToInt(RoiPoints, ReferencePose, measuredPose);
}
@@ -0,0 +1,4 @@
namespace XP.ImageProcessing.Core.Alignment;
/// <summary>图像像素平面上的点(与 WPF/Emgu 解耦)。</summary>
public readonly record struct Point2D(double X, double Y);
@@ -0,0 +1,18 @@
namespace XP.ImageProcessing.Core.Alignment;
/// <summary>
/// 图像平面上的刚体位姿:绕 <see cref="X"/>/<see cref="Y"/> 旋转 <see cref="AngleDegrees"/>(度)。
/// 与 TemplateMatchLib 的 CenterX/CenterY/Angle 约定一致。
/// </summary>
public readonly record struct Pose2D(double X, double Y, double AngleDegrees)
{
/// <summary>示教/标准姿态(角度 0,中心由调用方指定)。</summary>
public static Pose2D IdentityAt(double x, double y) => new(x, y, 0);
/// <summary>
/// 由模板学习 ROI 矩形估计示教位姿中心(pattern 几何中心),角度默认 0。
/// 更稳妥的做法是在示教图上自匹配得到 <see cref="Pose2D"/>。
/// </summary>
public static Pose2D FromTemplateRoiCenter(int roiX, int roiY, int roiWidth, int roiHeight, double angleDegrees = 0)
=> new(roiX + roiWidth * 0.5, roiY + roiHeight * 0.5, angleDegrees);
}
@@ -0,0 +1,96 @@
namespace XP.ImageProcessing.Core.Alignment;
/// <summary>
/// 将示教图(模板坐标系)上的 ROI 点变换到运行图坐标。
/// 旋转中心与模板匹配一致:绕 <see cref="Pose2D.X"/>/<see cref="Pose2D.Y"/>pattern 中心)。
/// </summary>
public static class RoiAlignment
{
/// <summary>
/// 刚体变换:示教图点 → 运行图点。
/// <paramref name="reference"/>:示教图上的基准位姿;
/// <paramref name="measured"/>:运行图匹配位姿。
/// </summary>
public static Point2D TransformPoint(Point2D point, Pose2D reference, Pose2D measured)
{
double dTheta = DegreesToRadians(measured.AngleDegrees - reference.AngleDegrees);
double cos = Math.Cos(dTheta);
double sin = Math.Sin(dTheta);
double dx = point.X - reference.X;
double dy = point.Y - reference.Y;
return new Point2D(
measured.X + cos * dx - sin * dy,
measured.Y + sin * dx + cos * dy);
}
public static Point2D TransformPoint(double x, double y, Pose2D reference, Pose2D measured)
=> TransformPoint(new Point2D(x, y), reference, measured);
/// <summary>变换多边形顶点(顺序不变)。</summary>
public static Point2D[] TransformPolygon(IReadOnlyList<Point2D> templatePoints, Pose2D reference, Pose2D measured)
{
if (templatePoints == null || templatePoints.Count == 0)
return Array.Empty<Point2D>();
var result = new Point2D[templatePoints.Count];
for (int i = 0; i < templatePoints.Count; i++)
result[i] = TransformPoint(templatePoints[i], reference, measured);
return result;
}
/// <summary>变换后四舍五入为整型顶点,供 BGA 等算子 PolyX/PolyY 注入。</summary>
public static (int X, int Y)[] TransformPolygonToInt(
IReadOnlyList<Point2D> templatePoints,
Pose2D reference,
Pose2D measured)
{
var transformed = TransformPolygon(templatePoints, reference, measured);
var result = new (int X, int Y)[transformed.Length];
for (int i = 0; i < transformed.Length; i++)
{
result[i] = (
(int)Math.Round(transformed[i].X, MidpointRounding.AwayFromZero),
(int)Math.Round(transformed[i].Y, MidpointRounding.AwayFromZero));
}
return result;
}
/// <summary>变换轴对齐矩形为四个顶点(左上、右上、右下、左下)。</summary>
public static Point2D[] TransformRect(double x, double y, double width, double height, Pose2D reference, Pose2D measured)
{
var corners = new[]
{
new Point2D(x, y),
new Point2D(x + width, y),
new Point2D(x + width, y + height),
new Point2D(x, y + height)
};
return TransformPolygon(corners, reference, measured);
}
/// <summary>
/// 校验匹配结果四角质心是否与 Center 一致(用于确认库的中心/角度约定)。
/// </summary>
public static bool IsMatchCenterConsistentWithCorners(
double centerX,
double centerY,
double ltX,
double ltY,
double rtX,
double rtY,
double rbX,
double rbY,
double lbX,
double lbY,
double tolerancePixels = 1.0)
{
double cx = (ltX + rtX + rbX + lbX) * 0.25;
double cy = (ltY + rtY + rbY + lbY) * 0.25;
double dx = cx - centerX;
double dy = cy - centerY;
return dx * dx + dy * dy <= tolerancePixels * tolerancePixels;
}
private static double DegreesToRadians(double degrees) => degrees * (Math.PI / 180.0);
}
@@ -0,0 +1,27 @@
using XP.ImageProcessing.Core.Alignment;
namespace XP.ImageProcessing.Processors;
/// <summary>
/// 将 TemplateMatchLib 匹配结果转换为对齐工具使用的 <see cref="Pose2D"/>。
/// </summary>
public static class TemplateMatchAlignmentExtensions
{
public static Pose2D ToPose2D(this TM_Result result)
=> new(result.CenterX, result.CenterY, result.Angle);
/// <summary>四角质心是否与 Center 一致(容差默认 1 像素)。</summary>
public static bool IsCenterConsistentWithCorners(this TM_Result result, double tolerancePixels = 1.0)
=> RoiAlignment.IsMatchCenterConsistentWithCorners(
result.CenterX,
result.CenterY,
result.LtX,
result.LtY,
result.RtX,
result.RtY,
result.RbX,
result.RbY,
result.LbX,
result.LbY,
tolerancePixels);
}
@@ -30,6 +30,12 @@ public struct TM_Params
/// <summary>是否亚像素估计 (1=是, 0=否)</summary>
public int UseSubPixel;
/// <summary>
/// 开启亚像素且角度容差绝对值超过该值时,托管封装会在调用原生库前关闭亚像素,
/// 以避免部分版本 TemplateMatchLib 在 Debug 下出现 vector 越界断言。
/// </summary>
public const double SubPixelAngleSafetyLimitDegrees = 90.0;
/// <summary>
/// 创建默认参数
/// </summary>
@@ -168,9 +174,33 @@ public sealed class TemplateMatcherHandle : IDisposable
public TM_Result[] Match(IntPtr srcData, int srcWidth, int srcHeight, int srcStep, TM_Params param)
{
ThrowIfDisposed();
var results = new TM_Result[param.MaxCount];
// 与库默认一致并对齐已知崩溃组合:Debug 下亚像素 + 大角度容差易触发 vector 越界断言;
// 金字塔最小面积过小也可能与内部层级假设不一致。
int tw = 0, th = 0, _pyramidLayers = 0;
_ = GetTemplateInfo(out tw, out th, out _pyramidLayers);
int templatePixels = Math.Max(0, tw) * Math.Max(0, th);
int maxCount = Math.Clamp(param.MaxCount, 1, 100);
int minReduce = (int)Math.Clamp(param.MinReduceArea, 64, 4096);
if (templatePixels >= 512)
minReduce = Math.Max(256, minReduce);
if (templatePixels > 0)
minReduce = Math.Min(minReduce, templatePixels);
minReduce = Math.Max(64, minReduce);
int useSubPixel = param.UseSubPixel;
if (useSubPixel != 0 && Math.Abs(param.ToleranceAngle) > TM_Params.SubPixelAngleSafetyLimitDegrees)
useSubPixel = 0;
var p = param;
p.MaxCount = maxCount;
p.MinReduceArea = minReduce;
p.UseSubPixel = useSubPixel;
var results = new TM_Result[p.MaxCount];
int count = TemplateMatchNative.TM_Match(_handle, srcData, srcWidth, srcHeight, srcStep,
ref param, results, param.MaxCount);
ref p, results, p.MaxCount);
if (count <= 0)
return Array.Empty<TM_Result>();
@@ -0,0 +1,146 @@
// ============================================================================
// Copyright © 2026 Hexagon Technology Center GmbH. All Rights Reserved.
// 文件名: BackgroundDefectAnalyzer.cs
// 描述: 白底/黑底对比下的缺陷斑点分析(仅 ROI 内计算,不接入流水线算子)
// 算法: Otsu 二值化 → 形态学开运算 → 外轮廓 → 面积过滤 → 轮廓顶点最远弦(物理长度与历史等效直径同一标定:mm/px → μm)
// 作者: 李伟 wei.lw.li@hexagon.com
// ============================================================================
using System.Collections.Generic;
using System.Drawing;
using Emgu.CV;
using Emgu.CV.CvEnum;
using Emgu.CV.Structure;
using Emgu.CV.Util;
namespace XP.ImageProcessing.Processors;
/// <summary>
/// 底色类型:决定 Otsu 后保留的前景是暗区还是亮区。
/// </summary>
public enum BackgroundDefectMode
{
/// <summary>白底图像上检测偏暗区域(BinaryInv + Otsu)。</summary>
WhiteBackground,
/// <summary>黑底图像上检测偏亮区域(Binary + Otsu)。</summary>
BlackBackground
}
/// <summary>
/// 单个斑点:轮廓顶点相对于 ROI 左上角;<see cref="MaxChordMicrometers"/> 为轮廓顶点间欧氏距离最大值(微米)。
/// </summary>
public sealed class BackgroundDefectBlob
{
public Point[] ContourInRoi { get; init; } = Array.Empty<Point>();
public double MaxChordMicrometers { get; init; }
public Point MaxChordEndAInRoi { get; init; }
public Point MaxChordEndBInRoi { get; init; }
}
/// <summary>
/// 在灰度 ROI 上执行底色缺陷斑点检测。调用方负责构造与释放 <paramref name="roiGray"/>。
/// </summary>
public static class BackgroundDefectAnalyzer
{
/// <summary>
/// 在 ROI 灰度图上检测斑点。
/// </summary>
/// <param name="roiGray">ROI 灰度图(单通道 8 位)。</param>
/// <param name="mode">白底或黑底模式。</param>
/// <param name="minAreaPixels">轮廓最小面积(像素²),小于此值的轮廓丢弃。</param>
/// <param name="mmPerPixel">像素物理尺寸(毫米/像素),用于轮廓最远弦换算为微米。</param>
/// <param name="morphKernelSize">形态学开运算核尺寸(奇数,默认 3)。</param>
public static List<BackgroundDefectBlob> DetectBlobs(
Image<Gray, byte> roiGray,
BackgroundDefectMode mode,
int minAreaPixels = 50,
double mmPerPixel = 0.139,
int morphKernelSize = 3)
{
if (roiGray == null) throw new ArgumentNullException(nameof(roiGray));
if (minAreaPixels < 1) minAreaPixels = 1;
if (mmPerPixel <= 0) mmPerPixel = 0.139;
if (morphKernelSize < 1) morphKernelSize = 1;
if ((morphKernelSize & 1) == 0) morphKernelSize++;
int rw = roiGray.Width;
int rh = roiGray.Height;
if (rw < 1 || rh < 1) return new List<BackgroundDefectBlob>();
var thresholdType = mode == BackgroundDefectMode.WhiteBackground
? ThresholdType.BinaryInv | ThresholdType.Otsu
: ThresholdType.Binary | ThresholdType.Otsu;
using var binary = new Image<Gray, byte>(rw, rh);
CvInvoke.Threshold(roiGray, binary, 0, 255, thresholdType);
using var kernel = CvInvoke.GetStructuringElement(
ElementShape.Ellipse, new Size(morphKernelSize, morphKernelSize), new Point(-1, -1));
CvInvoke.MorphologyEx(binary, binary, MorphOp.Open, kernel, new Point(-1, -1), 1,
BorderType.Default, new MCvScalar(0));
using var contours = new VectorOfVectorOfPoint();
using var hierarchy = new Mat();
CvInvoke.FindContours(binary, contours, hierarchy, RetrType.External, ChainApproxMethod.ChainApproxSimple);
var result = new List<BackgroundDefectBlob>();
for (int i = 0; i < contours.Size; i++)
{
double area = CvInvoke.ContourArea(contours[i]);
if (area < minAreaPixels) continue;
int n = contours[i].Size;
if (n < 2) continue;
var pts = new Point[n];
for (int j = 0; j < n; j++)
pts[j] = contours[i][j];
MaxChordInPixelSpace(pts, out double maxChordPx, out Point pa, out Point pb);
double maxChordMicrometers = maxChordPx * mmPerPixel * 1000.0;
result.Add(new BackgroundDefectBlob
{
ContourInRoi = pts,
MaxChordMicrometers = maxChordMicrometers,
MaxChordEndAInRoi = pa,
MaxChordEndBInRoi = pb
});
}
return result;
}
/// <summary>轮廓顶点集合上的最远点对(欧氏距离,像素)。</summary>
private static void MaxChordInPixelSpace(Point[] pts, out double maxChordPx, out Point a, out Point b)
{
maxChordPx = 0;
a = pts[0];
b = pts.Length > 1 ? pts[1] : pts[0];
long bestSq = 0;
int bestI = 0, bestJ = 1;
int n = pts.Length;
for (int i = 0; i < n; i++)
{
int iX = pts[i].X, iY = pts[i].Y;
for (int j = i + 1; j < n; j++)
{
long dx = iX - pts[j].X;
long dy = iY - pts[j].Y;
long sq = dx * dx + dy * dy;
if (sq > bestSq)
{
bestSq = sq;
bestI = i;
bestJ = j;
}
}
}
a = pts[bestI];
b = pts[bestJ];
maxChordPx = Math.Sqrt(bestSq);
}
}
@@ -0,0 +1,582 @@
// ============================================================================
// Copyright © 2026 Hexagon Technology Center GmbH. All Rights Reserved.
// 文件名: EdgeCircleFitProcessor.cs
// 描述: 边缘查找拟合圆算子
// 功能:
// - 沿预估圆周等角度放置卡尺,每个卡尺沿径向搜索边缘点
// - 支持亚像素精度(抛物线插值)
// - 支持边缘极性选择和搜索方向(向内/向外)
// - 使用最小二乘或RANSAC算法拟合圆
// - 输出拟合圆参数、边缘点、内点/外点、拟合误差
// 算法: 卡尺边缘检测 + 最小二乘/RANSAC圆拟合
// 作者: 李伟 wei.lw.li@hexagon.com
// ============================================================================
using Emgu.CV;
using Emgu.CV.Structure;
using XP.ImageProcessing.Core;
using Serilog;
using System.Drawing;
namespace XP.ImageProcessing.Processors;
/// <summary>
/// 圆拟合结果
/// </summary>
public class CircleFitResult
{
/// <summary>拟合是否成功</summary>
public bool Success { get; set; }
/// <summary>拟合圆心X</summary>
public double CenterX { get; set; }
/// <summary>拟合圆心Y</summary>
public double CenterY { get; set; }
/// <summary>拟合半径</summary>
public double Radius { get; set; }
/// <summary>所有检测到的边缘点</summary>
public List<EdgePointInfo> EdgePoints { get; set; } = new();
/// <summary>内点列表</summary>
public List<PointF> Inliers { get; set; } = new();
/// <summary>外点列表</summary>
public List<PointF> Outliers { get; set; } = new();
/// <summary>平均拟合误差(像素)</summary>
public double FitError { get; set; }
/// <summary>有效边缘点数</summary>
public int EdgePointCount { get; set; }
}
/// <summary>
/// 边缘查找拟合圆算子 - 沿预估圆周放置卡尺检测边缘点并拟合圆
/// </summary>
public class EdgeCircleFitProcessor : ImageProcessorBase
{
private static readonly ILogger _logger = Log.ForContext<EdgeCircleFitProcessor>();
private static readonly Random _random = new();
public EdgeCircleFitProcessor()
{
Name = LocalizationHelper.GetString("EdgeCircleFitProcessor_Name");
Description = LocalizationHelper.GetString("EdgeCircleFitProcessor_Description");
}
protected override void InitializeParameters()
{
// ── 预估圆参数(由UI交互注入,不可见) ──
Parameters.Add("CenterX", new ProcessorParameter(
"CenterX", "CenterX", typeof(int), 200, null, null, "") { IsVisible = false });
Parameters.Add("CenterY", new ProcessorParameter(
"CenterY", "CenterY", typeof(int), 200, null, null, "") { IsVisible = false });
Parameters.Add("Radius", new ProcessorParameter(
"Radius", "Radius", typeof(int), 100, null, null, "") { IsVisible = false });
// ── 卡尺参数 ──
Parameters.Add("CaliperCount", new ProcessorParameter(
"CaliperCount",
LocalizationHelper.GetString("EdgeCircleFitProcessor_CaliperCount"),
typeof(int), 36, 3, 360,
LocalizationHelper.GetString("EdgeCircleFitProcessor_CaliperCount_Desc")));
Parameters.Add("CaliperWidth", new ProcessorParameter(
"CaliperWidth",
LocalizationHelper.GetString("EdgeCircleFitProcessor_CaliperWidth"),
typeof(int), 40, 5, 500,
LocalizationHelper.GetString("EdgeCircleFitProcessor_CaliperWidth_Desc")));
// ── 边缘检测参数 ──
Parameters.Add("EdgePolarity", new ProcessorParameter(
"EdgePolarity",
LocalizationHelper.GetString("EdgeCircleFitProcessor_EdgePolarity"),
typeof(string), "Both", null, null,
LocalizationHelper.GetString("EdgeCircleFitProcessor_EdgePolarity_Desc"),
new string[] { "BrightToDark", "DarkToBright", "Both" }));
Parameters.Add("EdgeThreshold", new ProcessorParameter(
"EdgeThreshold",
LocalizationHelper.GetString("EdgeCircleFitProcessor_EdgeThreshold"),
typeof(int), 20, 1, 255,
LocalizationHelper.GetString("EdgeCircleFitProcessor_EdgeThreshold_Desc")));
Parameters.Add("Sigma", new ProcessorParameter(
"Sigma",
LocalizationHelper.GetString("EdgeCircleFitProcessor_Sigma"),
typeof(double), 1.0, 0.1, 10.0,
LocalizationHelper.GetString("EdgeCircleFitProcessor_Sigma_Desc")));
Parameters.Add("SearchDirection", new ProcessorParameter(
"SearchDirection",
LocalizationHelper.GetString("EdgeCircleFitProcessor_SearchDirection"),
typeof(string), "Both", null, null,
LocalizationHelper.GetString("EdgeCircleFitProcessor_SearchDirection_Desc"),
new string[] { "Inward", "Outward", "Both" }));
// ── 拟合参数 ──
Parameters.Add("FitMethod", new ProcessorParameter(
"FitMethod",
LocalizationHelper.GetString("EdgeCircleFitProcessor_FitMethod"),
typeof(string), "RANSAC", null, null,
LocalizationHelper.GetString("EdgeCircleFitProcessor_FitMethod_Desc"),
new string[] { "LeastSquares", "RANSAC" }));
Parameters.Add("RansacThreshold", new ProcessorParameter(
"RansacThreshold",
LocalizationHelper.GetString("EdgeCircleFitProcessor_RansacThreshold"),
typeof(double), 2.0, 0.5, 20.0,
LocalizationHelper.GetString("EdgeCircleFitProcessor_RansacThreshold_Desc")));
Parameters.Add("Thickness", new ProcessorParameter(
"Thickness",
LocalizationHelper.GetString("EdgeCircleFitProcessor_Thickness"),
typeof(int), 2, 1, 10,
LocalizationHelper.GetString("EdgeCircleFitProcessor_Thickness_Desc")));
}
public override Image<Gray, byte> Process(Image<Gray, byte> inputImage)
{
int centerX = GetParameter<int>("CenterX");
int centerY = GetParameter<int>("CenterY");
int radius = GetParameter<int>("Radius");
int caliperCount = GetParameter<int>("CaliperCount");
int caliperWidth = GetParameter<int>("CaliperWidth");
string edgePolarity = GetParameter<string>("EdgePolarity");
int edgeThreshold = GetParameter<int>("EdgeThreshold");
double sigma = GetParameter<double>("Sigma");
string searchDirection = GetParameter<string>("SearchDirection");
string fitMethod = GetParameter<string>("FitMethod");
double ransacThreshold = GetParameter<double>("RansacThreshold");
OutputData.Clear();
_logger.Debug(
"EdgeCircleFit started: Center=({CX},{CY}), R={R}, Calipers={Count}, Width={Width}",
centerX, centerY, radius, caliperCount, caliperWidth);
if (radius < 5)
{
_logger.Warning("Radius too small for circle fitting");
OutputData["CircleFitResult"] = new CircleFitResult { Success = false };
return inputImage.Clone();
}
// 沿圆周等角度放置卡尺
var edgePoints = new List<EdgePointInfo>();
double angleStep = 2.0 * Math.PI / caliperCount;
for (int i = 0; i < caliperCount; i++)
{
double angle = angleStep * i;
// 圆周上的采样点
double sampleX = centerX + radius * Math.Cos(angle);
double sampleY = centerY + radius * Math.Sin(angle);
// 径向方向(从圆心指向外)
double dirX = Math.Cos(angle);
double dirY = Math.Sin(angle);
// 根据搜索方向确定卡尺搜索方向
double searchDirX, searchDirY;
if (searchDirection == "Inward")
{
searchDirX = -dirX;
searchDirY = -dirY;
}
else if (searchDirection == "Outward")
{
searchDirX = dirX;
searchDirY = dirY;
}
else // Both: 搜索方向为径向(从内到外),卡尺中心在圆周上
{
searchDirX = dirX;
searchDirY = dirY;
}
var edgePoint = FindEdgeInCaliper(
inputImage, sampleX, sampleY, searchDirX, searchDirY,
caliperWidth, edgePolarity, edgeThreshold, sigma, i);
if (edgePoint != null)
{
edgePoints.Add(edgePoint);
}
}
_logger.Debug("Found {Count} edge points from {Total} calipers", edgePoints.Count, caliperCount);
// 拟合圆
var result = FitCircle(edgePoints, fitMethod, ransacThreshold);
// 存储输出
OutputData["CircleFitResult"] = result;
OutputData["EdgePoints"] = edgePoints.Select(p => p.Position).ToArray();
OutputData["EdgePointCount"] = edgePoints.Count;
OutputData["Thickness"] = GetParameter<int>("Thickness");
if (result.Success)
{
OutputData["FittedCenterX"] = result.CenterX;
OutputData["FittedCenterY"] = result.CenterY;
OutputData["FittedRadius"] = result.Radius;
OutputData["InlierPoints"] = result.Inliers.ToArray();
OutputData["OutlierPoints"] = result.Outliers.ToArray();
OutputData["FitError"] = result.FitError;
_logger.Information(
"EdgeCircleFit completed: Center=({CX:F2},{CY:F2}), R={R:F2}, Inliers={Inliers}/{Total}, Error={Error:F3}px",
result.CenterX, result.CenterY, result.Radius,
result.Inliers.Count, edgePoints.Count, result.FitError);
}
else
{
_logger.Warning("EdgeCircleFit failed: insufficient edge points");
}
return inputImage.Clone();
}
// ══════════════════════════════════════════════════════════════
// 卡尺边缘检测(复用直线拟合中的逻辑)
// ══════════════════════════════════════════════════════════════
private EdgePointInfo? FindEdgeInCaliper(
Image<Gray, byte> image,
double centerX, double centerY,
double dirX, double dirY,
int caliperWidth, string polarity,
int threshold, double sigma, int caliperIndex)
{
int halfWidth = caliperWidth / 2;
int profileLength = caliperWidth;
var profile = new double[profileLength];
int validCount = 0;
for (int i = 0; i < profileLength; i++)
{
double offset = i - halfWidth;
double px = centerX + dirX * offset;
double py = centerY + dirY * offset;
int ix = (int)Math.Round(px);
int iy = (int)Math.Round(py);
if (ix >= 0 && ix < image.Width && iy >= 0 && iy < image.Height)
{
profile[i] = image.Data[iy, ix, 0];
validCount++;
}
else
{
profile[i] = 0;
}
}
if (validCount < profileLength * 0.5)
return null;
if (sigma > 0.1)
profile = GaussianSmooth1D(profile, sigma);
var derivative = new double[profileLength];
for (int i = 1; i < profileLength - 1; i++)
derivative[i] = (profile[i + 1] - profile[i - 1]) / 2.0;
int bestIdx = -1;
double bestStrength = 0;
for (int i = 2; i < profileLength - 2; i++)
{
double strength = derivative[i];
bool validPolarity = polarity switch
{
"BrightToDark" => strength < 0,
"DarkToBright" => strength > 0,
_ => true
};
if (!validPolarity) continue;
double absStrength = Math.Abs(strength);
if (absStrength >= threshold && absStrength > bestStrength)
{
bestStrength = absStrength;
bestIdx = i;
}
}
if (bestIdx < 0)
return null;
// 亚像素插值
double subPixelOffset = 0;
if (bestIdx > 0 && bestIdx < profileLength - 1)
{
double left = Math.Abs(derivative[bestIdx - 1]);
double center = Math.Abs(derivative[bestIdx]);
double right = Math.Abs(derivative[bestIdx + 1]);
double denom = 2.0 * (2.0 * center - left - right);
if (Math.Abs(denom) > 1e-6)
{
subPixelOffset = (left - right) / denom;
subPixelOffset = Math.Clamp(subPixelOffset, -0.5, 0.5);
}
}
double edgeOffset = (bestIdx + subPixelOffset) - halfWidth;
float edgeX = (float)(centerX + dirX * edgeOffset);
float edgeY = (float)(centerY + dirY * edgeOffset);
return new EdgePointInfo
{
Position = new PointF(edgeX, edgeY),
Strength = bestStrength,
CaliperIndex = caliperIndex,
IsInlier = true
};
}
private static double[] GaussianSmooth1D(double[] data, double sigma)
{
int kernelRadius = (int)Math.Ceiling(sigma * 3);
int kernelSize = kernelRadius * 2 + 1;
var kernel = new double[kernelSize];
double sum = 0;
for (int i = 0; i < kernelSize; i++)
{
double x = i - kernelRadius;
kernel[i] = Math.Exp(-x * x / (2.0 * sigma * sigma));
sum += kernel[i];
}
for (int i = 0; i < kernelSize; i++)
kernel[i] /= sum;
var result = new double[data.Length];
for (int i = 0; i < data.Length; i++)
{
double val = 0, wSum = 0;
for (int k = 0; k < kernelSize; k++)
{
int idx = i + k - kernelRadius;
if (idx >= 0 && idx < data.Length)
{
val += data[idx] * kernel[k];
wSum += kernel[k];
}
}
result[i] = wSum > 0 ? val / wSum : data[i];
}
return result;
}
// ══════════════════════════════════════════════════════════════
// 圆拟合
// ══════════════════════════════════════════════════════════════
private CircleFitResult FitCircle(List<EdgePointInfo> edgePoints, string method, double ransacThreshold)
{
var result = new CircleFitResult();
if (edgePoints.Count < 3)
{
result.Success = false;
return result;
}
if (method == "RANSAC" && edgePoints.Count >= 4)
return FitCircleRANSAC(edgePoints, ransacThreshold);
else
return FitCircleLeastSquares(edgePoints);
}
/// <summary>
/// 最小二乘拟合圆(Kasa方法)
/// 将 (x-a)² + (y-b)² = r² 展开为: x² + y² = 2ax + 2by + (r²-a²-b²)
/// 令 c = r²-a²-b², 线性方程: 2ax + 2by + c = x² + y²
/// </summary>
private CircleFitResult FitCircleLeastSquares(List<EdgePointInfo> edgePoints)
{
var points = edgePoints.Select(p => p.Position).ToArray();
var (cx, cy, r) = KasaFit(points);
var result = new CircleFitResult
{
Success = true,
CenterX = cx,
CenterY = cy,
Radius = r,
Inliers = points.ToList(),
Outliers = new List<PointF>(),
EdgePointCount = edgePoints.Count,
EdgePoints = edgePoints
};
foreach (var ep in edgePoints)
ep.IsInlier = true;
result.FitError = ComputeCircleFitError(points, cx, cy, r);
return result;
}
/// <summary>
/// RANSAC 圆拟合
/// </summary>
private CircleFitResult FitCircleRANSAC(List<EdgePointInfo> edgePoints, double threshold)
{
var result = new CircleFitResult();
var points = edgePoints.Select(p => p.Position).ToArray();
int n = points.Length;
int maxIterations = Math.Min(2000, n * (n - 1) * (n - 2) / 6);
int bestInlierCount = 0;
double bestCx = 0, bestCy = 0, bestR = 0;
List<int> bestInlierIndices = new();
for (int iter = 0; iter < maxIterations; iter++)
{
// 随机选3个点
int i1 = _random.Next(n), i2 = _random.Next(n), i3 = _random.Next(n);
if (i1 == i2 || i1 == i3 || i2 == i3) continue;
var (cx, cy, r) = FitCircleFrom3Points(points[i1], points[i2], points[i3]);
if (r <= 0 || double.IsNaN(r)) continue;
// 统计内点
var inlierIndices = new List<int>();
for (int i = 0; i < n; i++)
{
double dist = Math.Abs(Distance(points[i], cx, cy) - r);
if (dist <= threshold)
inlierIndices.Add(i);
}
if (inlierIndices.Count > bestInlierCount)
{
bestInlierCount = inlierIndices.Count;
bestInlierIndices = inlierIndices;
// 用所有内点重新拟合
var inlierPoints = inlierIndices.Select(i => points[i]).ToArray();
(bestCx, bestCy, bestR) = KasaFit(inlierPoints);
}
if (bestInlierCount > n * 0.95)
break;
}
if (bestInlierCount < 3)
{
result.Success = false;
return result;
}
result.Success = true;
result.CenterX = bestCx;
result.CenterY = bestCy;
result.Radius = bestR;
var inlierSet = new HashSet<int>(bestInlierIndices);
for (int i = 0; i < n; i++)
{
if (inlierSet.Contains(i))
{
result.Inliers.Add(points[i]);
edgePoints[i].IsInlier = true;
}
else
{
result.Outliers.Add(points[i]);
edgePoints[i].IsInlier = false;
}
}
result.FitError = ComputeCircleFitError(result.Inliers.ToArray(), bestCx, bestCy, bestR);
result.EdgePointCount = edgePoints.Count;
result.EdgePoints = edgePoints;
return result;
}
/// <summary>
/// Kasa 最小二乘圆拟合
/// </summary>
private static (double cx, double cy, double r) KasaFit(PointF[] points)
{
int n = points.Length;
if (n < 3) return (0, 0, 0);
// 构建线性方程组: A * [a, b, c]^T = B
// 其中 2*a*xi + 2*b*yi + c = xi² + yi²
double sumX = 0, sumY = 0, sumX2 = 0, sumY2 = 0;
double sumXY = 0, sumX3 = 0, sumY3 = 0, sumX2Y = 0, sumXY2 = 0;
for (int i = 0; i < n; i++)
{
double x = points[i].X, y = points[i].Y;
double x2 = x * x, y2 = y * y;
sumX += x; sumY += y;
sumX2 += x2; sumY2 += y2;
sumXY += x * y;
sumX3 += x2 * x; sumY3 += y2 * y;
sumX2Y += x2 * y; sumXY2 += x * y2;
}
double A = n * sumX2 - sumX * sumX;
double B = n * sumXY - sumX * sumY;
double C = n * sumY2 - sumY * sumY;
double D = 0.5 * (n * (sumX3 + sumXY2) - sumX * (sumX2 + sumY2));
double E = 0.5 * (n * (sumX2Y + sumY3) - sumY * (sumX2 + sumY2));
double denom = A * C - B * B;
if (Math.Abs(denom) < 1e-10)
return (0, 0, 0);
double cx = (D * C - B * E) / denom;
double cy = (A * E - B * D) / denom;
double r = Math.Sqrt((sumX2 + sumY2 - 2 * cx * sumX - 2 * cy * sumY) / n + cx * cx + cy * cy);
return (cx, cy, r);
}
/// <summary>
/// 3点拟合圆
/// </summary>
private static (double cx, double cy, double r) FitCircleFrom3Points(PointF p1, PointF p2, PointF p3)
{
double ax = p1.X, ay = p1.Y;
double bx = p2.X, by = p2.Y;
double cx = p3.X, cy = p3.Y;
double d = 2 * (ax * (by - cy) + bx * (cy - ay) + cx * (ay - by));
if (Math.Abs(d) < 1e-10)
return (0, 0, -1);
double ux = ((ax * ax + ay * ay) * (by - cy) + (bx * bx + by * by) * (cy - ay) + (cx * cx + cy * cy) * (ay - by)) / d;
double uy = ((ax * ax + ay * ay) * (cx - bx) + (bx * bx + by * by) * (ax - cx) + (cx * cx + cy * cy) * (bx - ax)) / d;
double r = Math.Sqrt((ax - ux) * (ax - ux) + (ay - uy) * (ay - uy));
return (ux, uy, r);
}
private static double Distance(PointF p, double cx, double cy)
{
double dx = p.X - cx, dy = p.Y - cy;
return Math.Sqrt(dx * dx + dy * dy);
}
private static double ComputeCircleFitError(PointF[] points, double cx, double cy, double r)
{
if (points.Length == 0) return 0;
double total = 0;
foreach (var p in points)
total += Math.Abs(Distance(p, cx, cy) - r);
return total / points.Length;
}
}
@@ -0,0 +1,638 @@
// ============================================================================
// Copyright © 2026 Hexagon Technology Center GmbH. All Rights Reserved.
// 文件名: EdgeLineFitProcessor.cs
// 描述: 边缘查找拟合直线算子
// 功能:
// - 沿用户定义的搜索线等间距放置多个卡尺(Caliper)
// - 在每个卡尺内沿垂直方向提取灰度投影并求导,定位边缘点
// - 支持亚像素精度(抛物线插值)
// - 支持边缘极性选择(亮到暗/暗到亮/双向)
// - 使用最小二乘或RANSAC算法拟合直线
// - 输出拟合直线参数、边缘点、内点/外点、拟合误差
// 算法: 卡尺边缘检测 + 最小二乘/RANSAC直线拟合
// 作者: 李伟 wei.lw.li@hexagon.com
// ============================================================================
using Emgu.CV;
using Emgu.CV.CvEnum;
using Emgu.CV.Structure;
using Emgu.CV.Util;
using XP.ImageProcessing.Core;
using Serilog;
using System.Drawing;
namespace XP.ImageProcessing.Processors;
/// <summary>
/// 边缘点信息
/// </summary>
public class EdgePointInfo
{
/// <summary>边缘点坐标(亚像素)</summary>
public PointF Position { get; set; }
/// <summary>边缘强度(梯度绝对值)</summary>
public double Strength { get; set; }
/// <summary>卡尺索引</summary>
public int CaliperIndex { get; set; }
/// <summary>是否为拟合内点</summary>
public bool IsInlier { get; set; } = true;
}
/// <summary>
/// 直线拟合结果
/// </summary>
public class LineFitResult
{
/// <summary>拟合是否成功</summary>
public bool Success { get; set; }
/// <summary>直线方向向量 (vx, vy)</summary>
public PointF Direction { get; set; }
/// <summary>直线上一点 (x0, y0)</summary>
public PointF PointOnLine { get; set; }
/// <summary>直线角度(度,相对于X轴)</summary>
public double AngleDegrees { get; set; }
/// <summary>直线端点1(用于绘制)</summary>
public PointF Endpoint1 { get; set; }
/// <summary>直线端点2(用于绘制)</summary>
public PointF Endpoint2 { get; set; }
/// <summary>所有检测到的边缘点</summary>
public List<EdgePointInfo> EdgePoints { get; set; } = new();
/// <summary>内点列表</summary>
public List<PointF> Inliers { get; set; } = new();
/// <summary>外点列表</summary>
public List<PointF> Outliers { get; set; } = new();
/// <summary>平均拟合误差(像素)</summary>
public double FitError { get; set; }
/// <summary>有效边缘点数</summary>
public int EdgePointCount { get; set; }
}
/// <summary>
/// 边缘查找拟合直线算子 - 使用卡尺法检测边缘点并拟合直线
/// </summary>
public class EdgeLineFitProcessor : ImageProcessorBase
{
private static readonly ILogger _logger = Log.ForContext<EdgeLineFitProcessor>();
private static readonly Random _random = new();
public EdgeLineFitProcessor()
{
Name = LocalizationHelper.GetString("EdgeLineFitProcessor_Name");
Description = LocalizationHelper.GetString("EdgeLineFitProcessor_Description");
}
protected override void InitializeParameters()
{
// ── 搜索线起止点(由UI交互控件注入,不可见) ──
Parameters.Add("StartX", new ProcessorParameter(
"StartX", "StartX", typeof(int), 100, null, null, "") { IsVisible = false });
Parameters.Add("StartY", new ProcessorParameter(
"StartY", "StartY", typeof(int), 200, null, null, "") { IsVisible = false });
Parameters.Add("EndX", new ProcessorParameter(
"EndX", "EndX", typeof(int), 400, null, null, "") { IsVisible = false });
Parameters.Add("EndY", new ProcessorParameter(
"EndY", "EndY", typeof(int), 200, null, null, "") { IsVisible = false });
// ── 卡尺参数 ──
Parameters.Add("CaliperCount", new ProcessorParameter(
"CaliperCount",
LocalizationHelper.GetString("EdgeLineFitProcessor_CaliperCount"),
typeof(int), 20, 3, 200,
LocalizationHelper.GetString("EdgeLineFitProcessor_CaliperCount_Desc")));
Parameters.Add("CaliperWidth", new ProcessorParameter(
"CaliperWidth",
LocalizationHelper.GetString("EdgeLineFitProcessor_CaliperWidth"),
typeof(int), 40, 5, 500,
LocalizationHelper.GetString("EdgeLineFitProcessor_CaliperWidth_Desc")));
// ── 边缘检测参数 ──
Parameters.Add("EdgePolarity", new ProcessorParameter(
"EdgePolarity",
LocalizationHelper.GetString("EdgeLineFitProcessor_EdgePolarity"),
typeof(string), "Both", null, null,
LocalizationHelper.GetString("EdgeLineFitProcessor_EdgePolarity_Desc"),
new string[] { "BrightToDark", "DarkToBright", "Both" }));
Parameters.Add("EdgeThreshold", new ProcessorParameter(
"EdgeThreshold",
LocalizationHelper.GetString("EdgeLineFitProcessor_EdgeThreshold"),
typeof(int), 30, 1, 255,
LocalizationHelper.GetString("EdgeLineFitProcessor_EdgeThreshold_Desc")));
Parameters.Add("Sigma", new ProcessorParameter(
"Sigma",
LocalizationHelper.GetString("EdgeLineFitProcessor_Sigma"),
typeof(double), 1.0, 0.1, 10.0,
LocalizationHelper.GetString("EdgeLineFitProcessor_Sigma_Desc")));
// ── 拟合参数 ──
Parameters.Add("FitMethod", new ProcessorParameter(
"FitMethod",
LocalizationHelper.GetString("EdgeLineFitProcessor_FitMethod"),
typeof(string), "RANSAC", null, null,
LocalizationHelper.GetString("EdgeLineFitProcessor_FitMethod_Desc"),
new string[] { "LeastSquares", "RANSAC" }));
Parameters.Add("RansacThreshold", new ProcessorParameter(
"RansacThreshold",
LocalizationHelper.GetString("EdgeLineFitProcessor_RansacThreshold"),
typeof(double), 2.0, 0.5, 20.0,
LocalizationHelper.GetString("EdgeLineFitProcessor_RansacThreshold_Desc")));
Parameters.Add("Thickness", new ProcessorParameter(
"Thickness",
LocalizationHelper.GetString("EdgeLineFitProcessor_Thickness"),
typeof(int), 2, 1, 10,
LocalizationHelper.GetString("EdgeLineFitProcessor_Thickness_Desc")));
}
public override Image<Gray, byte> Process(Image<Gray, byte> inputImage)
{
// 读取参数
int startX = GetParameter<int>("StartX");
int startY = GetParameter<int>("StartY");
int endX = GetParameter<int>("EndX");
int endY = GetParameter<int>("EndY");
int caliperCount = GetParameter<int>("CaliperCount");
int caliperWidth = GetParameter<int>("CaliperWidth");
string edgePolarity = GetParameter<string>("EdgePolarity");
int edgeThreshold = GetParameter<int>("EdgeThreshold");
double sigma = GetParameter<double>("Sigma");
string fitMethod = GetParameter<string>("FitMethod");
double ransacThreshold = GetParameter<double>("RansacThreshold");
int thickness = GetParameter<int>("Thickness");
OutputData.Clear();
_logger.Debug(
"EdgeLineFit started: Search({StartX},{StartY})->({EndX},{EndY}), Calipers={Count}, Width={Width}, Polarity={Polarity}",
startX, startY, endX, endY, caliperCount, caliperWidth, edgePolarity);
// 计算搜索线方向和垂直方向
double searchDx = endX - startX;
double searchDy = endY - startY;
double searchLen = Math.Sqrt(searchDx * searchDx + searchDy * searchDy);
if (searchLen < 1.0)
{
_logger.Warning("Search line too short, cannot perform edge detection");
OutputData["LineFitResult"] = new LineFitResult { Success = false };
return inputImage.Clone();
}
// 搜索线单位方向
double ux = searchDx / searchLen;
double uy = searchDy / searchLen;
// 垂直于搜索线的方向(卡尺搜索方向)
double perpX = -uy;
double perpY = ux;
// 沿搜索线等间距放置卡尺
var edgePoints = new List<EdgePointInfo>();
double step = searchLen / (caliperCount + 1);
for (int i = 0; i < caliperCount; i++)
{
// 卡尺中心点
double cx = startX + ux * step * (i + 1);
double cy = startY + uy * step * (i + 1);
// 在卡尺内沿垂直方向提取灰度剖面
var edgePoint = FindEdgeInCaliper(
inputImage, cx, cy, perpX, perpY,
caliperWidth, edgePolarity, edgeThreshold, sigma, i);
if (edgePoint != null)
{
edgePoints.Add(edgePoint);
}
}
_logger.Debug("Found {Count} edge points from {Total} calipers", edgePoints.Count, caliperCount);
// 拟合直线
var result = FitLine(edgePoints, fitMethod, ransacThreshold, inputImage.Size);
// 存储输出数据
OutputData["LineFitResult"] = result;
OutputData["EdgePoints"] = edgePoints.Select(p => p.Position).ToArray();
OutputData["EdgePointCount"] = edgePoints.Count;
OutputData["Thickness"] = thickness;
if (result.Success)
{
OutputData["FittedLineDirection"] = result.Direction;
OutputData["FittedLinePoint"] = result.PointOnLine;
OutputData["LineAngle"] = result.AngleDegrees;
OutputData["LineEndpoint1"] = result.Endpoint1;
OutputData["LineEndpoint2"] = result.Endpoint2;
OutputData["InlierPoints"] = result.Inliers.ToArray();
OutputData["OutlierPoints"] = result.Outliers.ToArray();
OutputData["FitError"] = result.FitError;
_logger.Information(
"EdgeLineFit completed: Angle={Angle:F2}°, Inliers={Inliers}/{Total}, Error={Error:F3}px",
result.AngleDegrees, result.Inliers.Count, edgePoints.Count, result.FitError);
}
else
{
_logger.Warning("EdgeLineFit failed: insufficient edge points for line fitting");
}
// 搜索区域信息(供UI绘制)
OutputData["SearchStart"] = new PointF(startX, startY);
OutputData["SearchEnd"] = new PointF(endX, endY);
OutputData["CaliperWidth"] = caliperWidth;
OutputData["CaliperCount"] = caliperCount;
OutputData["PerpDirection"] = new PointF((float)perpX, (float)perpY);
return inputImage.Clone();
}
/// <summary>
/// 在单个卡尺内查找边缘点
/// </summary>
private EdgePointInfo? FindEdgeInCaliper(
Image<Gray, byte> image,
double centerX, double centerY,
double perpX, double perpY,
int caliperWidth, string polarity,
int threshold, double sigma, int caliperIndex)
{
int halfWidth = caliperWidth / 2;
int profileLength = caliperWidth;
// 提取灰度剖面
var profile = new double[profileLength];
int validCount = 0;
for (int i = 0; i < profileLength; i++)
{
double offset = i - halfWidth;
double px = centerX + perpX * offset;
double py = centerY + perpY * offset;
int ix = (int)Math.Round(px);
int iy = (int)Math.Round(py);
if (ix >= 0 && ix < image.Width && iy >= 0 && iy < image.Height)
{
profile[i] = image.Data[iy, ix, 0];
validCount++;
}
else
{
profile[i] = 0;
}
}
if (validCount < profileLength * 0.5)
return null;
// 高斯平滑
if (sigma > 0.1)
{
profile = GaussianSmooth1D(profile, sigma);
}
// 求一阶导数
var derivative = new double[profileLength];
for (int i = 1; i < profileLength - 1; i++)
{
derivative[i] = (profile[i + 1] - profile[i - 1]) / 2.0;
}
// 根据极性查找最强边缘
int bestIdx = -1;
double bestStrength = 0;
for (int i = 2; i < profileLength - 2; i++)
{
double strength = derivative[i];
bool validPolarity = polarity switch
{
"BrightToDark" => strength < 0, // 亮到暗:导数为负
"DarkToBright" => strength > 0, // 暗到亮:导数为正
_ => true // Both:任意方向
};
if (!validPolarity) continue;
double absStrength = Math.Abs(strength);
if (absStrength >= threshold && absStrength > bestStrength)
{
bestStrength = absStrength;
bestIdx = i;
}
}
if (bestIdx < 0)
return null;
// 亚像素精度:抛物线插值
double subPixelOffset = 0;
if (bestIdx > 0 && bestIdx < profileLength - 1)
{
double left = Math.Abs(derivative[bestIdx - 1]);
double center = Math.Abs(derivative[bestIdx]);
double right = Math.Abs(derivative[bestIdx + 1]);
double denom = 2.0 * (2.0 * center - left - right);
if (Math.Abs(denom) > 1e-6)
{
subPixelOffset = (left - right) / denom;
subPixelOffset = Math.Clamp(subPixelOffset, -0.5, 0.5);
}
}
double edgeOffset = (bestIdx + subPixelOffset) - halfWidth;
float edgeX = (float)(centerX + perpX * edgeOffset);
float edgeY = (float)(centerY + perpY * edgeOffset);
return new EdgePointInfo
{
Position = new PointF(edgeX, edgeY),
Strength = bestStrength,
CaliperIndex = caliperIndex,
IsInlier = true
};
}
/// <summary>
/// 一维高斯平滑
/// </summary>
private static double[] GaussianSmooth1D(double[] data, double sigma)
{
int kernelRadius = (int)Math.Ceiling(sigma * 3);
int kernelSize = kernelRadius * 2 + 1;
var kernel = new double[kernelSize];
double sum = 0;
for (int i = 0; i < kernelSize; i++)
{
double x = i - kernelRadius;
kernel[i] = Math.Exp(-x * x / (2.0 * sigma * sigma));
sum += kernel[i];
}
for (int i = 0; i < kernelSize; i++)
kernel[i] /= sum;
var result = new double[data.Length];
for (int i = 0; i < data.Length; i++)
{
double val = 0;
double wSum = 0;
for (int k = 0; k < kernelSize; k++)
{
int idx = i + k - kernelRadius;
if (idx >= 0 && idx < data.Length)
{
val += data[idx] * kernel[k];
wSum += kernel[k];
}
}
result[i] = wSum > 0 ? val / wSum : data[i];
}
return result;
}
/// <summary>
/// 拟合直线
/// </summary>
private LineFitResult FitLine(List<EdgePointInfo> edgePoints, string method,
double ransacThreshold, Size imageSize)
{
var result = new LineFitResult();
if (edgePoints.Count < 2)
{
result.Success = false;
return result;
}
if (method == "RANSAC" && edgePoints.Count >= 3)
{
return FitLineRANSAC(edgePoints, ransacThreshold, imageSize);
}
else
{
return FitLineLeastSquares(edgePoints, imageSize);
}
}
/// <summary>
/// 最小二乘直线拟合(使用OpenCV FitLine
/// </summary>
private LineFitResult FitLineLeastSquares(List<EdgePointInfo> edgePoints, Size imageSize)
{
var result = new LineFitResult();
var points = edgePoints.Select(p => p.Position).ToArray();
using var pointVector = new VectorOfPointF(points);
using var lineMat = new Mat();
CvInvoke.FitLine(pointVector, lineMat, DistType.L2, 0, 0.01, 0.01);
var lineParams = new float[4];
System.Runtime.InteropServices.Marshal.Copy(lineMat.DataPointer, lineParams, 0, 4);
float vx = lineParams[0], vy = lineParams[1];
float x0 = lineParams[2], y0 = lineParams[3];
result.Success = true;
result.Direction = new PointF(vx, vy);
result.PointOnLine = new PointF(x0, y0);
result.AngleDegrees = Math.Atan2(vy, vx) * 180.0 / Math.PI;
// 计算端点(延伸到图像边界或搜索范围)
ComputeLineEndpoints(result, points, imageSize);
// 所有点都是内点
result.Inliers = points.ToList();
result.Outliers = new List<PointF>();
foreach (var ep in edgePoints)
ep.IsInlier = true;
// 计算拟合误差
result.FitError = ComputeFitError(points, vx, vy, x0, y0);
result.EdgePointCount = edgePoints.Count;
result.EdgePoints = edgePoints;
return result;
}
/// <summary>
/// RANSAC直线拟合
/// </summary>
private LineFitResult FitLineRANSAC(List<EdgePointInfo> edgePoints, double threshold, Size imageSize)
{
var result = new LineFitResult();
var points = edgePoints.Select(p => p.Position).ToArray();
int n = points.Length;
// RANSAC参数
int maxIterations = Math.Min(1000, n * (n - 1) / 2);
int bestInlierCount = 0;
float bestVx = 0, bestVy = 0, bestX0 = 0, bestY0 = 0;
List<int> bestInlierIndices = new();
for (int iter = 0; iter < maxIterations; iter++)
{
// 随机选择2个点
int idx1 = _random.Next(n);
int idx2 = _random.Next(n);
if (idx1 == idx2) continue;
PointF p1 = points[idx1], p2 = points[idx2];
float dx = p2.X - p1.X, dy = p2.Y - p1.Y;
float len = (float)Math.Sqrt(dx * dx + dy * dy);
if (len < 1e-6f) continue;
float vx = dx / len, vy = dy / len;
// 统计内点
var inlierIndices = new List<int>();
for (int i = 0; i < n; i++)
{
double dist = PointToLineDistance(points[i], p1, vx, vy);
if (dist <= threshold)
{
inlierIndices.Add(i);
}
}
if (inlierIndices.Count > bestInlierCount)
{
bestInlierCount = inlierIndices.Count;
bestInlierIndices = inlierIndices;
// 用所有内点重新拟合
var inlierPoints = inlierIndices.Select(i => points[i]).ToArray();
using var pv = new VectorOfPointF(inlierPoints);
using var lpMat = new Mat();
CvInvoke.FitLine(pv, lpMat, DistType.L2, 0, 0.01, 0.01);
var lp = new float[4];
System.Runtime.InteropServices.Marshal.Copy(lpMat.DataPointer, lp, 0, 4);
bestVx = lp[0]; bestVy = lp[1]; bestX0 = lp[2]; bestY0 = lp[3];
}
// 如果内点比例已经很高,提前退出
if (bestInlierCount > n * 0.95)
break;
}
if (bestInlierCount < 2)
{
result.Success = false;
return result;
}
result.Success = true;
result.Direction = new PointF(bestVx, bestVy);
result.PointOnLine = new PointF(bestX0, bestY0);
result.AngleDegrees = Math.Atan2(bestVy, bestVx) * 180.0 / Math.PI;
// 分类内点/外点
var inliers = new List<PointF>();
var outliers = new List<PointF>();
var inlierSet = new HashSet<int>(bestInlierIndices);
for (int i = 0; i < n; i++)
{
if (inlierSet.Contains(i))
{
inliers.Add(points[i]);
edgePoints[i].IsInlier = true;
}
else
{
outliers.Add(points[i]);
edgePoints[i].IsInlier = false;
}
}
result.Inliers = inliers;
result.Outliers = outliers;
// 计算端点
ComputeLineEndpoints(result, inliers.ToArray(), imageSize);
// 计算拟合误差(仅内点)
result.FitError = ComputeFitError(inliers.ToArray(), bestVx, bestVy, bestX0, bestY0);
result.EdgePointCount = edgePoints.Count;
result.EdgePoints = edgePoints;
return result;
}
/// <summary>
/// 计算点到直线的距离
/// </summary>
private static double PointToLineDistance(PointF point, PointF linePoint, float vx, float vy)
{
// 直线法向量 (-vy, vx)
double dx = point.X - linePoint.X;
double dy = point.Y - linePoint.Y;
return Math.Abs(-vy * dx + vx * dy);
}
/// <summary>
/// 计算直线端点(基于边缘点的投影范围)
/// </summary>
private static void ComputeLineEndpoints(LineFitResult result, PointF[] points, Size imageSize)
{
float vx = result.Direction.X, vy = result.Direction.Y;
float x0 = result.PointOnLine.X, y0 = result.PointOnLine.Y;
// 将所有点投影到直线方向上,找最小和最大投影值
double minT = double.MaxValue, maxT = double.MinValue;
foreach (var p in points)
{
double t = (p.X - x0) * vx + (p.Y - y0) * vy;
if (t < minT) minT = t;
if (t > maxT) maxT = t;
}
// 稍微延伸一点
double extend = (maxT - minT) * 0.05;
minT -= extend;
maxT += extend;
result.Endpoint1 = new PointF(
(float)(x0 + vx * minT),
(float)(y0 + vy * minT));
result.Endpoint2 = new PointF(
(float)(x0 + vx * maxT),
(float)(y0 + vy * maxT));
}
/// <summary>
/// 计算平均拟合误差
/// </summary>
private static double ComputeFitError(PointF[] points, float vx, float vy, float x0, float y0)
{
if (points.Length == 0) return 0;
double totalError = 0;
foreach (var p in points)
{
double dx = p.X - x0;
double dy = p.Y - y0;
double dist = Math.Abs(-vy * dx + vx * dy);
totalError += dist;
}
return totalError / points.Length;
}
}
@@ -504,6 +504,12 @@ namespace XP.ImageProcessing.RoiControl.Controls
private Point? _bgaPendingCenter; // 等待第二次点击定半径
private Ellipse _bgaPendingDot;
// 边缘查找拟合直线临时状态
private int _elfClickCount;
private Ellipse _elfTempDot1;
private Line _elfTempLine;
private Point? _elfTempStart;
// 气泡测量状态
public enum BubbleSubTool { Roi, RoiCircle, RoiPolygon, Wand, Brush, Eraser }
private BubbleSubTool _bubbleTool = BubbleSubTool.Roi;
@@ -690,6 +696,8 @@ namespace XP.ImageProcessing.RoiControl.Controls
HandleFillRateClick(pos);
else if (CurrentMeasureMode == Models.MeasureMode.BgaVoid)
HandleBgaVoidClick(pos);
else if (CurrentMeasureMode == Models.MeasureMode.EdgeLineFit)
HandleEdgeLineFitClick(pos);
// BubbleMeasure 的点击在 MouseDown/Move/Up 中处理(拖拽画 ROI 和画笔)
}
@@ -870,6 +878,49 @@ namespace XP.ImageProcessing.RoiControl.Controls
return g;
}
// ── 边缘查找拟合直线 ──
private void HandleEdgeLineFitClick(Point pos)
{
_elfClickCount++;
if (_elfClickCount == 1)
{
_elfTempStart = pos;
_elfTempDot1 = CreateMDot(Brushes.Cyan);
_measureOverlay.Children.Add(_elfTempDot1);
SetDotPos(_elfTempDot1, pos);
RaiseMeasureStatusChanged($"直线拟合 - 搜索线起点: ({pos.X:F0}, {pos.Y:F0}),请点击搜索线终点");
}
else if (_elfClickCount == 2)
{
// 绘制搜索线
_elfTempLine = new Line
{
Stroke = Brushes.Cyan,
StrokeThickness = 1,
StrokeDashArray = new DoubleCollection { 4, 2 },
IsHitTestVisible = false,
X1 = _elfTempStart.Value.X,
Y1 = _elfTempStart.Value.Y,
X2 = pos.X,
Y2 = pos.Y
};
_measureOverlay.Children.Add(_elfTempLine);
// 触发完成事件,传递搜索线起止点
RaiseMeasureCompleted(_elfTempStart.Value, pos, 0, MeasureCount, "EdgeLineFit");
RaiseMeasureStatusChanged($"直线拟合 - 搜索线已定义: ({_elfTempStart.Value.X:F0},{_elfTempStart.Value.Y:F0}) → ({pos.X:F0},{pos.Y:F0})");
// 清理临时状态
if (_elfTempDot1 != null) _measureOverlay.Children.Remove(_elfTempDot1);
_elfTempDot1 = null;
_elfTempStart = null;
_elfClickCount = 0;
CurrentMeasureMode = Models.MeasureMode.None;
}
}
// ── 角度测量 ──
private void HandleAngleClick(Point pos)
@@ -8,6 +8,7 @@ namespace XP.ImageProcessing.RoiControl.Models
Angle,
FillRate,
BgaVoid,
BubbleMeasure
BubbleMeasure,
EdgeLineFit
}
}
+47 -19
View File
@@ -55,6 +55,7 @@ using XplorePlane.Services.Storage;
using XplorePlane.ViewModels;
using XplorePlane.ViewModels.Cnc;
using XplorePlane.ViewModels.Debug;
using XplorePlane.ViewModels.ImageProcessing;
using XplorePlane.Views;
using XplorePlane.Views.Cnc;
using XplorePlane.Views.Debug;
@@ -174,12 +175,12 @@ namespace XplorePlane
{
var cameraVm = bootstrapper.Container.Resolve<NavigationPropertyPanelViewModel>();
cameraVm?.Dispose();
Log.Information("导航相机 ViewModel 已释放");
Log.Information("Navigation camera ViewModel has been released");
}
}
catch (Exception ex)
{
Log.Error(ex, "导航相机 ViewModel 释放失败");
Log.Error(ex, "Navigation camera ViewModel release failed");
}
// 释放导航相机服务资源
@@ -190,12 +191,12 @@ namespace XplorePlane
{
var cameraService = bootstrapper.Container.Resolve<ICameraService>();
cameraService?.Dispose();
Log.Information("导航相机服务资源已释放");
Log.Information("Navigation camera service resources have been released");
}
}
catch (Exception ex)
{
Log.Error(ex, "导航相机服务资源释放失败");
Log.Error(ex, "Navigation camera service resource release failed");
}
// 释放主界面探测器帧流水线资源
@@ -350,9 +351,6 @@ namespace XplorePlane
// 主窗体加载完成后再连接相机,确保所有模块和原生 DLL 已完成初始化
shell.Loaded += async (s, e) =>
{
// [DEV] 导航相机连接已屏蔽,开发阶段跳过以加快启动速度
// TryConnectCamera();
// 初始化主界面探测器帧流水线,开始接收探测器图像事件
try
{
@@ -376,11 +374,22 @@ namespace XplorePlane
// {
// Log.Error(ex, "通知相机 ViewModel 失败");
// }
// [DEV] 导航相机连接已屏蔽,开发阶段跳过以加快启动速度
//TryConnectCamera();
//try
//{
// var cameraVm = Container.Resolve<NavigationPropertyPanelViewModel>();
// cameraVm.OnCameraReady();
//}
//catch (Exception ex)
//{
// Log.Error(ex, "Failed to notify the camera ViewModel");
//}
// if (_cameraError != null)
// {
// HexMessageBox.Show(_cameraError, MessageBoxButton.OK, MessageBoxImage.Error);
// }
//if (_cameraError != null)
//{
// HexMessageBox.Show(_cameraError, MessageBoxButton.OK, MessageBoxImage.Error);
//}
};
return shell;
@@ -531,17 +540,17 @@ namespace XplorePlane
try
{
var info = camera.Open();
Log.Information("导航相机已连接: {ModelName} (SN: {SerialNumber})", info.ModelName, info.SerialNumber);
Log.Information("Navigation camera connected: {ModelName} (SN: {SerialNumber})", info.ModelName, info.SerialNumber);
}
catch (DeviceNotFoundException)
{
Log.Warning("未检测到导航相机");
_cameraError = "未检测到导航相机,请检查连接后重启软件。";
Log.Warning("Navigation camera not detected");
_cameraError = "Navigation camera not detectedPlease check the connection and restart the software.。";
}
catch (Exception ex)
{
Log.Warning(ex, "导航相机自动连接失败: {Message}", ex.Message);
_cameraError = $"导航相机连接失败: {ex.Message}";
Log.Warning(ex, "Automatic connection of navigation camera failed: {Message}", ex.Message);
_cameraError = $"Navigation camera connection failed: {ex.Message}";
}
}
@@ -575,7 +584,6 @@ namespace XplorePlane
containerRegistry.RegisterSingleton<IImageProcessingService, ImageProcessingService>();
containerRegistry.Register<ImageProcessingViewModel>();
// 注册流水线服务(单例,共享 IImageProcessingService
containerRegistry.RegisterSingleton<IPipelineExecutionService, PipelineExecutionService>();
containerRegistry.RegisterSingleton<IPipelinePersistenceService, PipelinePersistenceService>();
@@ -593,6 +601,7 @@ namespace XplorePlane
// 注册流水线 ViewModel(每次解析创建新实例)
containerRegistry.Register<PipelineEditorViewModel>();
containerRegistry.Register<OperatorToolboxViewModel>();
containerRegistry.Register<TemplateMatchAssistantViewModel>();
// 注册硬件库的 ViewModel(供 ViewModelLocator 自动装配)
containerRegistry.Register<XP.Hardware.RaySource.ViewModels.RaySourceOperateViewModel>();
@@ -634,7 +643,26 @@ namespace XplorePlane
// ── 导航相机服务(单例)──
containerRegistry.RegisterSingleton<ICameraFactory, CameraFactory>();
containerRegistry.RegisterSingleton<ICameraController>(() =>
new CameraFactory().CreateController("Basler"));
{
string cameraType = "Hikvision"; // 默认值
try
{
var configPath = Path.Combine(AppContext.BaseDirectory, "config.json");
if (File.Exists(configPath))
{
var json = File.ReadAllText(configPath);
var match = System.Text.RegularExpressions.Regex.Match(json, "\"CameraType\"\\s*:\\s*\"([^\"]+)\"");
if (match.Success)
cameraType = match.Groups[1].Value;
}
}
catch (Exception ex)
{
Log.Warning(ex, "Failed to read CameraType configuration, using default value Hikvision");
}
Log.Information("Camera Type: {CameraType}", cameraType);
return new CameraFactory().CreateController(cameraType);
});
containerRegistry.RegisterSingleton<ICameraService, CameraService>();
// ── 录制服务(单例)──
@@ -662,4 +690,4 @@ namespace XplorePlane
base.ConfigureModuleCatalog(moduleCatalog);
}
}
}
}
Binary file not shown.

After

Width:  |  Height:  |  Size: 778 B

Binary file not shown.

After

Width:  |  Height:  |  Size: 583 B

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 KiB

+63 -1
View File
@@ -1,7 +1,21 @@
using System.Collections.Generic;
using System.Drawing;
using Prism.Events;
namespace XplorePlane.Events
{
/// <summary>
/// 白底/黑底检测单条结果:全局图像坐标下的轮廓与最远弦(微米与既有展示规则一致)。
/// </summary>
public class BackgroundDefectDetectionItem
{
public List<Point> Contour { get; set; } = new();
/// <summary>轮廓顶点间最远距离(微米)。</summary>
public double SizeMicrometers { get; set; }
public Point ChordP1 { get; set; }
public Point ChordP2 { get; set; }
}
/// <summary>
/// 测量工具模式
/// </summary>
@@ -13,7 +27,8 @@ namespace XplorePlane.Events
Angle,
ThroughHoleFillRate,
BgaVoid,
BubbleMeasure
BubbleMeasure,
EdgeLineFit
}
/// <summary>
@@ -25,4 +40,51 @@ namespace XplorePlane.Events
/// 十字辅助线切换事件
/// </summary>
public class ToggleCrosshairEvent : PubSubEvent { }
/// <summary>
/// 行灰度分布切换事件
/// </summary>
public class ToggleLineProfileEvent : PubSubEvent { }
/// <summary>
/// 白底检测事件(进入ROI绘制模式)
/// </summary>
public class WhiteBackgroundDetectionEvent : PubSubEvent { }
/// <summary>
/// 白底检测ROI绘制完成事件
/// </summary>
public class WhiteBackgroundRoiDrawnEvent : PubSubEvent<System.Windows.Int32Rect> { }
/// <summary>
/// 白底检测结果事件
/// </summary>
public class WhiteBackgroundResultEvent : PubSubEvent<WhiteBackgroundResultPayload> { }
public class WhiteBackgroundResultPayload
{
public System.Drawing.Rectangle RoiRect { get; set; }
public List<BackgroundDefectDetectionItem> Detections { get; set; } = new();
}
/// <summary>
/// 黑底检测事件(进入ROI绘制模式)
/// </summary>
public class BlackBackgroundDetectionEvent : PubSubEvent { }
/// <summary>
/// 黑底检测ROI绘制完成事件
/// </summary>
public class BlackBackgroundRoiDrawnEvent : PubSubEvent<System.Windows.Int32Rect> { }
/// <summary>
/// 黑底检测结果事件
/// </summary>
public class BlackBackgroundResultEvent : PubSubEvent<BlackBackgroundResultPayload> { }
public class BlackBackgroundResultPayload
{
public System.Drawing.Rectangle RoiRect { get; set; }
public List<BackgroundDefectDetectionItem> Detections { get; set; } = new();
}
}
@@ -0,0 +1,47 @@
using System.Collections.Generic;
using System.Windows;
using Prism.Events;
namespace XplorePlane.Events;
/// <summary>
/// 进入「在视口上框选模板 ROI」模式(与主画布 Preview 鼠标逻辑配合)。
/// </summary>
public class TemplateMatchEnterRoiModeEvent : PubSubEvent { }
/// <summary>
/// 模板 ROI 框选完成(图像/画布像素坐标,与白底检测 ROI 约定一致)。仅表示区域已确定,不表示已训练。
/// </summary>
public class TemplateMatchRoiDrawnEvent : PubSubEvent<Int32Rect> { }
/// <summary>
/// 清除视口上的模板助手持久 ROI 框(例如加载模型后或重置时)。
/// </summary>
public class TemplateMatchClearRoiOverlayEvent : PubSubEvent { }
/// <summary>
/// 单次模板匹配试跑结果,供主视图叠加层绘制。
/// </summary>
public class TemplateMatchHitDto
{
public double CenterX { get; set; }
public double CenterY { get; set; }
public double Angle { get; set; }
public double Score { get; set; }
public double LtX { get; set; }
public double LtY { get; set; }
public double RtX { get; set; }
public double RtY { get; set; }
public double RbX { get; set; }
public double RbY { get; set; }
public double LbX { get; set; }
public double LbY { get; set; }
}
public class TemplateMatchPreviewPayload
{
public List<TemplateMatchHitDto> Hits { get; set; } = new();
public double MatchTimeMs { get; set; }
}
public class TemplateMatchPreviewResultEvent : PubSubEvent<TemplateMatchPreviewPayload> { }
@@ -0,0 +1,497 @@
using System;
using System.Collections.Generic;
using System.Linq;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Media;
using System.Windows.Media.Imaging;
using System.Windows.Shapes;
using Emgu.CV;
using Emgu.CV.Structure;
using Prism.Commands;
using Prism.Mvvm;
using XP.ImageProcessing.Processors;
using XP.ImageProcessing.RoiControl.Controls;
using XplorePlane.Services.MainViewport;
using Brushes = System.Windows.Media.Brushes;
using Ellipse = System.Windows.Shapes.Ellipse;
using Point = System.Windows.Point;
namespace XplorePlane.ViewModels.ImageProcessing
{
/// <summary>
/// 边缘查找拟合圆 ViewModel
/// 交互:3点定义预估圆,手柄可调整圆心和半径,点击拟合执行
/// </summary>
public class EdgeCircleFitViewModel : BindableBase
{
private readonly IMainViewportService _viewportService;
private PolygonRoiCanvas _canvas;
private Canvas _mainCanvas;
// 预估圆
private Point _center;
private double _radius;
private bool _circleDefined;
// 可视化
private readonly List<UIElement> _tempOverlays = new();
private readonly List<UIElement> _committedOverlays = new();
// 手柄位置
private Point _handleCenterPos;
private Point _handleRadiusPos; // 圆周上0°位置
// 交互
private enum DragTarget { None, Center, Radius }
private DragTarget _dragging = DragTarget.None;
private bool _isDrawing;
private int _fitCount;
private const double HandleSize = 12;
private const double HitRadius = 10;
private static readonly SolidColorBrush CaliperStroke;
private static readonly SolidColorBrush CaliperFill;
private static readonly SolidColorBrush FitCircleBrush;
private static readonly SolidColorBrush HandleFill;
static EdgeCircleFitViewModel()
{
CaliperStroke = new SolidColorBrush(Color.FromRgb(0, 255, 0));
CaliperStroke.Freeze();
CaliperFill = new SolidColorBrush(Color.FromArgb(15, 0, 255, 0));
CaliperFill.Freeze();
FitCircleBrush = new SolidColorBrush(Color.FromRgb(30, 144, 255));
FitCircleBrush.Freeze();
HandleFill = new SolidColorBrush(Color.FromArgb(220, 255, 255, 255));
HandleFill.Freeze();
}
public EdgeCircleFitViewModel(IMainViewportService viewportService)
{
_viewportService = viewportService;
FitCommand = new DelegateCommand(ExecuteFit, () => _circleDefined);
ClearAllCommand = new DelegateCommand(ExecuteClearAll);
DrawCircleCommand = new DelegateCommand(ExecuteDrawCircle);
}
// ── 命令 ──
public DelegateCommand FitCommand { get; }
public DelegateCommand ClearAllCommand { get; }
public DelegateCommand DrawCircleCommand { get; }
// ── 参数 ──
private int _caliperCount = 36;
public int CaliperCount { get => _caliperCount; set { if (SetProperty(ref _caliperCount, value)) RedrawTemp(); } }
private int _caliperWidth = 40;
public int CaliperWidth { get => _caliperWidth; set { if (SetProperty(ref _caliperWidth, value)) RedrawTemp(); } }
private string _edgePolarity = "Both";
public string EdgePolarity { get => _edgePolarity; set { if (SetProperty(ref _edgePolarity, value)) RedrawTemp(); } }
private int _edgeThreshold = 20;
public int EdgeThreshold { get => _edgeThreshold; set => SetProperty(ref _edgeThreshold, value); }
private double _sigma = 1.0;
public double Sigma { get => _sigma; set => SetProperty(ref _sigma, value); }
private string _searchDirection = "Both";
public string SearchDirection { get => _searchDirection; set => SetProperty(ref _searchDirection, value); }
private string _fitMethod = "RANSAC";
public string FitMethod { get => _fitMethod; set => SetProperty(ref _fitMethod, value); }
private double _ransacThreshold = 2.0;
public double RansacThreshold { get => _ransacThreshold; set => SetProperty(ref _ransacThreshold, value); }
private string _resultText = "Ready - click Draw Circle";
public string ResultText { get => _resultText; set => SetProperty(ref _resultText, value); }
// ── 初始化 ──
public void SetCanvas(PolygonRoiCanvas canvas)
{
_canvas = canvas;
_mainCanvas = FindChild<Canvas>(canvas, "mainCanvas");
}
public void OnPanelClosed()
{
UnregisterAll();
ClearTempOverlays();
}
// ══════════════════════════════════════════════════════════════
// 命令
// ══════════════════════════════════════════════════════════════
private void ExecuteDrawCircle()
{
ClearTempOverlays();
UnregisterAll();
_circleDefined = false;
_dragging = DragTarget.None;
FitCommand.RaiseCanExecuteChanged();
_isDrawing = true;
ResultText = "Press and drag to define circle (center → radius)";
RegisterInteraction();
}
private void ExecuteFit()
{
if (!_circleDefined) return;
// 清除上一次拟合结果
ClearCommitted();
var imageSource = _viewportService?.CurrentDisplayImage as BitmapSource;
if (imageSource == null) { ResultText = "Error: no image"; return; }
try
{
BitmapSource source = imageSource;
if (imageSource.Format != PixelFormats.Gray8)
source = new FormatConvertedBitmap(imageSource, PixelFormats.Gray8, null, 0);
int w = source.PixelWidth, h = source.PixelHeight;
int stride = w;
byte[] px = new byte[h * stride];
source.CopyPixels(px, stride, 0);
using var img = new Image<Gray, byte>(w, h);
for (int y = 0; y < h; y++)
for (int x = 0; x < w; x++)
img.Data[y, x, 0] = px[y * stride + x];
var proc = new EdgeCircleFitProcessor();
proc.SetParameter("CenterX", (int)_center.X);
proc.SetParameter("CenterY", (int)_center.Y);
proc.SetParameter("Radius", (int)_radius);
proc.SetParameter("CaliperCount", CaliperCount);
proc.SetParameter("CaliperWidth", CaliperWidth);
proc.SetParameter("EdgePolarity", EdgePolarity);
proc.SetParameter("EdgeThreshold", EdgeThreshold);
proc.SetParameter("Sigma", Sigma);
proc.SetParameter("SearchDirection", SearchDirection);
proc.SetParameter("FitMethod", FitMethod);
proc.SetParameter("RansacThreshold", RansacThreshold);
var result = proc.Process(img);
var od = proc.OutputData;
if (od.ContainsKey("CircleFitResult"))
{
var fr = od["CircleFitResult"] as CircleFitResult;
if (fr != null && fr.Success)
{
_fitCount++;
DrawFitResult(fr);
ResultText = $"[#{_fitCount}] Fit OK\nCenter: ({fr.CenterX:F1}, {fr.CenterY:F1})\n" +
$"Radius: {fr.Radius:F2} px\n" +
$"Inliers: {fr.Inliers.Count}/{fr.EdgePointCount}\n" +
$"Error: {fr.FitError:F3} px\n\nAdjust and fit again, or draw new";
}
else
{
int ec = od.ContainsKey("EdgePointCount") ? (int)od["EdgePointCount"] : 0;
ResultText = $"Fit failed\nEdge points: {ec}\nAdjust params or circle";
}
}
result.Dispose();
}
catch (Exception ex) { ResultText = $"Exception: {ex.Message}"; }
}
private void ExecuteClearAll()
{
ClearTempOverlays();
if (_mainCanvas != null)
foreach (var el in _committedOverlays) _mainCanvas.Children.Remove(el);
_committedOverlays.Clear();
_fitCount = 0;
UnregisterAll();
_circleDefined = false;
FitCommand.RaiseCanExecuteChanged();
ResultText = "Cleared";
}
// ══════════════════════════════════════════════════════════════
// 拟合结果绘制(永久)
// ══════════════════════════════════════════════════════════════
private void DrawFitResult(CircleFitResult fr)
{
if (_mainCanvas == null) return;
// 拟合圆(蓝色)
var circle = new Ellipse
{
Width = fr.Radius * 2, Height = fr.Radius * 2,
Stroke = FitCircleBrush, StrokeThickness = 2, Fill = Brushes.Transparent,
IsHitTestVisible = false
};
Canvas.SetLeft(circle, fr.CenterX - fr.Radius);
Canvas.SetTop(circle, fr.CenterY - fr.Radius);
AddCommitted(circle);
// 圆心十字
double cs = 6;
AddCommitted(new Line { X1 = fr.CenterX - cs, Y1 = fr.CenterY, X2 = fr.CenterX + cs, Y2 = fr.CenterY, Stroke = FitCircleBrush, StrokeThickness = 1.5, IsHitTestVisible = false });
AddCommitted(new Line { X1 = fr.CenterX, Y1 = fr.CenterY - cs, X2 = fr.CenterX, Y2 = fr.CenterY + cs, Stroke = FitCircleBrush, StrokeThickness = 1.5, IsHitTestVisible = false });
// 标注
var lbl = new TextBlock
{
Text = $"R:{fr.Radius:F1} C:({fr.CenterX:F1},{fr.CenterY:F1})",
Foreground = FitCircleBrush, FontSize = 11, FontWeight = FontWeights.Bold, IsHitTestVisible = false
};
Canvas.SetLeft(lbl, fr.CenterX + 5); Canvas.SetTop(lbl, fr.CenterY - fr.Radius - 18);
AddCommitted(lbl);
}
private void AddCommitted(UIElement el) { _mainCanvas.Children.Add(el); _committedOverlays.Add(el); }
private void ClearCommitted()
{
if (_mainCanvas == null) return;
foreach (var el in _committedOverlays) _mainCanvas.Children.Remove(el);
_committedOverlays.Clear();
}
// ══════════════════════════════════════════════════════════════
// 临时卡尺可视化
// ══════════════════════════════════════════════════════════════
private void RedrawTemp()
{
if (!_circleDefined || _mainCanvas == null) return;
ClearTempOverlays();
DrawTempCaliper();
}
private void DrawTempCaliper()
{
if (_mainCanvas == null || _radius < 5) return;
// 预估圆(虚线)
var previewCircle = new Ellipse
{
Width = _radius * 2, Height = _radius * 2,
Stroke = CaliperStroke, StrokeThickness = 1,
StrokeDashArray = new DoubleCollection { 4, 3 },
Fill = CaliperFill, IsHitTestVisible = false
};
Canvas.SetLeft(previewCircle, _center.X - _radius);
Canvas.SetTop(previewCircle, _center.Y - _radius);
AddTemp(previewCircle);
// 卡尺径向线
int count = CaliperCount;
double halfW = CaliperWidth / 2.0;
double angleStep = 2.0 * Math.PI / count;
for (int i = 0; i < count; i++)
{
double angle = angleStep * i;
double dirX = Math.Cos(angle), dirY = Math.Sin(angle);
double cx = _center.X + _radius * dirX;
double cy = _center.Y + _radius * dirY;
AddTemp(new Line
{
X1 = cx - dirX * halfW, Y1 = cy - dirY * halfW,
X2 = cx + dirX * halfW, Y2 = cy + dirY * halfW,
Stroke = CaliperStroke, StrokeThickness = 1, Opacity = 0.4, IsHitTestVisible = false
});
}
// 手柄
_handleCenterPos = _center;
_handleRadiusPos = new Point(_center.X + _radius, _center.Y);
AddTemp(MakeHandle(_handleCenterPos));
AddTemp(MakeHandle(_handleRadiusPos));
}
private void CommitCurrentCaliper()
{
if (_mainCanvas == null) return;
foreach (var el in _tempOverlays) _mainCanvas.Children.Remove(el);
_tempOverlays.Clear();
// 绘制永久卡尺(半透明)
var circle = new Ellipse
{
Width = _radius * 2, Height = _radius * 2,
Stroke = CaliperStroke, StrokeThickness = 1, Opacity = 0.4,
Fill = Brushes.Transparent, IsHitTestVisible = false
};
Canvas.SetLeft(circle, _center.X - _radius);
Canvas.SetTop(circle, _center.Y - _radius);
AddCommitted(circle);
}
private Ellipse MakeHandle(Point pos)
{
var h = new Ellipse
{
Width = HandleSize, Height = HandleSize,
Fill = HandleFill, Stroke = CaliperStroke, StrokeThickness = 2,
IsHitTestVisible = false
};
Canvas.SetLeft(h, pos.X - HandleSize / 2);
Canvas.SetTop(h, pos.Y - HandleSize / 2);
return h;
}
private void AddTemp(UIElement el) { _mainCanvas.Children.Add(el); _tempOverlays.Add(el); }
private void ClearTempOverlays()
{
if (_mainCanvas == null) return;
foreach (var el in _tempOverlays) _mainCanvas.Children.Remove(el);
_tempOverlays.Clear();
}
// ══════════════════════════════════════════════════════════════
// 鼠标交互
// ══════════════════════════════════════════════════════════════
private bool _interactionRegistered;
private void RegisterInteraction()
{
if (_canvas == null || _interactionRegistered) return;
_canvas.PreviewMouseLeftButtonDown += OnMouseDown;
_canvas.PreviewMouseMove += OnMouseMove;
_canvas.PreviewMouseLeftButtonUp += OnMouseUp;
_interactionRegistered = true;
}
private void UnregisterAll()
{
if (_canvas == null || !_interactionRegistered) return;
_canvas.PreviewMouseLeftButtonDown -= OnMouseDown;
_canvas.PreviewMouseMove -= OnMouseMove;
_canvas.PreviewMouseLeftButtonUp -= OnMouseUp;
_interactionRegistered = false;
_isDrawing = false;
_dragging = DragTarget.None;
}
private void OnMouseDown(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (_mainCanvas == null) return;
var pos = e.GetPosition(_mainCanvas);
// 绘制模式:按下鼠标确定圆心,拖拽确定半径
if (_isDrawing)
{
_center = pos;
_radius = 0;
_dragging = DragTarget.Radius; // 复用 Radius 拖拽逻辑
_canvas.CaptureMouse();
e.Handled = true;
return;
}
// 拖拽手柄
if (_circleDefined)
{
var target = HitTest(pos);
if (target != DragTarget.None)
{
_dragging = target;
_canvas.CaptureMouse();
e.Handled = true;
}
}
}
private void OnMouseMove(object sender, System.Windows.Input.MouseEventArgs e)
{
if (_dragging == DragTarget.None || _mainCanvas == null) return;
var pos = e.GetPosition(_mainCanvas);
if (_dragging == DragTarget.Center)
{
_center = pos;
}
else if (_dragging == DragTarget.Radius)
{
_radius = Math.Max(5, Dist(pos, _center));
}
// 实时预览
if (_radius >= 5)
{
_circleDefined = true;
ClearTempOverlays();
DrawTempCaliper();
}
e.Handled = true;
}
private void OnMouseUp(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (_dragging == DragTarget.None) return;
// 绘制模式完成
if (_isDrawing)
{
_isDrawing = false;
_dragging = DragTarget.None;
_canvas.ReleaseMouseCapture();
if (_radius >= 5)
{
_circleDefined = true;
FitCommand.RaiseCanExecuteChanged();
RedrawTemp();
ResultText = $"Circle defined: R={_radius:F0}px\nDrag handles to adjust\nClick Fit to execute";
}
else
{
_circleDefined = false;
ResultText = "Circle too small, try again";
}
e.Handled = true;
return;
}
_dragging = DragTarget.None;
_canvas.ReleaseMouseCapture();
ResultText = $"Circle: R={_radius:F0}px\nClick Fit to execute";
e.Handled = true;
}
private DragTarget HitTest(Point pos)
{
if (Dist(pos, _handleCenterPos) <= HitRadius) return DragTarget.Center;
if (Dist(pos, _handleRadiusPos) <= HitRadius) return DragTarget.Radius;
return DragTarget.None;
}
// ══════════════════════════════════════════════════════════════
// 辅助
// ══════════════════════════════════════════════════════════════
private static double Dist(Point a, Point b)
{
double dx = a.X - b.X, dy = a.Y - b.Y;
return Math.Sqrt(dx * dx + dy * dy);
}
private static T FindChild<T>(DependencyObject parent, string name) where T : FrameworkElement
{
int count = VisualTreeHelper.GetChildrenCount(parent);
for (int i = 0; i < count; i++)
{
var child = VisualTreeHelper.GetChild(parent, i);
if (child is T t && t.Name == name) return t;
var r = FindChild<T>(child, name);
if (r != null) return r;
}
return null;
}
}
}
@@ -0,0 +1,620 @@
using System;
using System.Collections.Generic;
using System.Linq;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Media;
using System.Windows.Media.Imaging;
using System.Windows.Shapes;
using Emgu.CV;
using Emgu.CV.Structure;
using Prism.Commands;
using Prism.Mvvm;
using XP.ImageProcessing.Processors;
using XP.ImageProcessing.RoiControl.Controls;
using XplorePlane.Services.MainViewport;
using Brushes = System.Windows.Media.Brushes;
using Ellipse = System.Windows.Shapes.Ellipse;
using Point = System.Windows.Point;
namespace XplorePlane.ViewModels.ImageProcessing
{
/// <summary>
/// 边缘查找拟合直线 ViewModel
/// 支持多次拟合,每次点击"画卡尺"开始一次新的测量,结果累积保留
/// 关闭面板时保留所有结果,仅清除当前正在编辑的临时卡尺
/// </summary>
public class EdgeLineFitViewModel : BindableBase
{
private readonly IMainViewportService _viewportService;
private PolygonRoiCanvas _canvas;
private Canvas _mainCanvas;
// 当前正在编辑的搜索线
private Point _lineStart;
private Point _lineEnd;
private double _halfWidth = 30;
private bool _lineDefined;
// 当前编辑中的临时可视化(卡尺框+手柄,拟合前可调整)
private readonly List<UIElement> _tempOverlays = new();
// 已完成的拟合结果(永久保留在画布上)
// 不由本类管理生命周期,关闭面板后仍保留
private readonly List<UIElement> _committedOverlays = new();
// 手柄位置
private Point _handleStartPos, _handleEndPos, _handleTopPos, _handleBottomPos;
// 交互状态
private enum DragTarget { None, Start, End, Top, Bottom }
private DragTarget _dragging = DragTarget.None;
private bool _isDrawingLine;
private int _drawClickCount;
private int _fitCount;
private const double HandleSize = 12;
private const double HitRadius = 10;
private static readonly SolidColorBrush CaliperStroke;
private static readonly SolidColorBrush CaliperFill;
private static readonly SolidColorBrush FitLineBrush;
private static readonly SolidColorBrush HandleFill;
static EdgeLineFitViewModel()
{
CaliperStroke = new SolidColorBrush(Color.FromRgb(0, 255, 0));
CaliperStroke.Freeze();
CaliperFill = new SolidColorBrush(Color.FromArgb(20, 0, 255, 0));
CaliperFill.Freeze();
FitLineBrush = new SolidColorBrush(Color.FromRgb(30, 144, 255));
FitLineBrush.Freeze();
HandleFill = new SolidColorBrush(Color.FromArgb(220, 255, 255, 255));
HandleFill.Freeze();
}
public EdgeLineFitViewModel(IMainViewportService viewportService)
{
_viewportService = viewportService;
FitCommand = new DelegateCommand(ExecuteFit, () => _lineDefined);
ClearAllCommand = new DelegateCommand(ExecuteClearAll);
DrawCaliperCommand = new DelegateCommand(ExecuteDrawCaliper);
}
// ── 命令 ──
public DelegateCommand FitCommand { get; }
public DelegateCommand ClearAllCommand { get; }
public DelegateCommand DrawCaliperCommand { get; }
// ── 参数 ──
private int _caliperCount = 20;
public int CaliperCount
{
get => _caliperCount;
set { if (SetProperty(ref _caliperCount, value)) RedrawTemp(); }
}
private int _displayWidth = 60;
public int DisplayWidth
{
get => _displayWidth;
set
{
if (SetProperty(ref _displayWidth, Math.Max(10, value)))
{
_halfWidth = _displayWidth / 2.0;
RedrawTemp();
}
}
}
private string _edgePolarity = "Both";
public string EdgePolarity
{
get => _edgePolarity;
set { if (SetProperty(ref _edgePolarity, value)) RedrawTemp(); }
}
private int _edgeThreshold = 20;
public int EdgeThreshold { get => _edgeThreshold; set => SetProperty(ref _edgeThreshold, value); }
private double _sigma = 1.0;
public double Sigma { get => _sigma; set => SetProperty(ref _sigma, value); }
private string _fitMethod = "RANSAC";
public string FitMethod { get => _fitMethod; set => SetProperty(ref _fitMethod, value); }
private double _ransacThreshold = 2.0;
public double RansacThreshold { get => _ransacThreshold; set => SetProperty(ref _ransacThreshold, value); }
private string _resultText = "就绪 - 点击「画卡尺」开始";
public string ResultText { get => _resultText; set => SetProperty(ref _resultText, value); }
// ── 初始化 ──
public void SetCanvas(PolygonRoiCanvas canvas)
{
_canvas = canvas;
_mainCanvas = FindChild<Canvas>(canvas, "mainCanvas");
}
/// <summary>面板关闭时调用:仅清除临时编辑状态,保留已拟合结果</summary>
public void OnPanelClosed()
{
UnregisterAll();
ClearTempOverlays(); // 清除正在编辑的卡尺手柄
// _committedOverlays 保留在画布上不清除
}
// ══════════════════════════════════════════════════════════════
// 命令实现
// ══════════════════════════════════════════════════════════════
/// <summary>开始一次新的卡尺绘制(不影响已有结果)</summary>
private void ExecuteDrawCaliper()
{
// 清除当前临时编辑
ClearTempOverlays();
UnregisterAll();
_lineDefined = false;
_dragging = DragTarget.None;
FitCommand.RaiseCanExecuteChanged();
_drawClickCount = 0;
_isDrawingLine = true;
ResultText = "请在图像上点击搜索线起点";
RegisterInteraction();
}
/// <summary>执行拟合,将结果提交为永久显示</summary>
private void ExecuteFit()
{
if (!_lineDefined) return;
// 清除上一次拟合结果
ClearCommitted();
var imageSource = _viewportService?.CurrentDisplayImage as BitmapSource;
if (imageSource == null) { ResultText = "错误:无可用图像"; return; }
try
{
BitmapSource source = imageSource;
if (imageSource.Format != PixelFormats.Gray8)
source = new FormatConvertedBitmap(imageSource, PixelFormats.Gray8, null, 0);
int w = source.PixelWidth, h = source.PixelHeight;
int stride = w;
byte[] px = new byte[h * stride];
source.CopyPixels(px, stride, 0);
using var img = new Image<Gray, byte>(w, h);
for (int y = 0; y < h; y++)
for (int x = 0; x < w; x++)
img.Data[y, x, 0] = px[y * stride + x];
var proc = new EdgeLineFitProcessor();
proc.SetParameter("StartX", (int)_lineStart.X);
proc.SetParameter("StartY", (int)_lineStart.Y);
proc.SetParameter("EndX", (int)_lineEnd.X);
proc.SetParameter("EndY", (int)_lineEnd.Y);
proc.SetParameter("CaliperCount", CaliperCount);
proc.SetParameter("CaliperWidth", (int)(_halfWidth * 2));
proc.SetParameter("EdgePolarity", EdgePolarity);
proc.SetParameter("EdgeThreshold", EdgeThreshold);
proc.SetParameter("Sigma", Sigma);
proc.SetParameter("FitMethod", FitMethod);
proc.SetParameter("RansacThreshold", RansacThreshold);
var result = proc.Process(img);
var od = proc.OutputData;
if (od.ContainsKey("LineFitResult"))
{
var fr = od["LineFitResult"] as LineFitResult;
if (fr != null && fr.Success)
{
_fitCount++;
DrawFitResult(fr);
ResultText = $"[#{_fitCount}] 拟合成功\n角度: {fr.AngleDegrees:F2}°\n" +
$"内点: {fr.Inliers.Count}/{fr.EdgePointCount}\n" +
$"误差: {fr.FitError:F3} px\n\n可继续调整后再次拟合";
}
else
{
int ec = od.ContainsKey("EdgePointCount") ? (int)od["EdgePointCount"] : 0;
ResultText = $"拟合失败\n边缘点: {ec}\n请调整参数或拖拽手柄";
}
}
result.Dispose();
}
catch (Exception ex) { ResultText = $"异常: {ex.Message}"; }
}
/// <summary>清除所有(包括已拟合的结果)</summary>
private void ExecuteClearAll()
{
ClearTempOverlays();
// 清除所有已提交的结果
if (_mainCanvas != null)
{
foreach (var el in _committedOverlays)
_mainCanvas.Children.Remove(el);
}
_committedOverlays.Clear();
_fitCount = 0;
UnregisterAll();
_lineDefined = false;
_dragging = DragTarget.None;
FitCommand.RaiseCanExecuteChanged();
ResultText = "已清除所有结果";
}
// ══════════════════════════════════════════════════════════════
// 提交当前卡尺为永久显示
// ══════════════════════════════════════════════════════════════
/// <summary>将当前临时卡尺可视化转为永久(去掉手柄,保留边框和等分线)</summary>
private void CommitCurrentCaliper()
{
if (_mainCanvas == null) return;
// 移除临时元素
foreach (var el in _tempOverlays)
_mainCanvas.Children.Remove(el);
_tempOverlays.Clear();
// 重新绘制卡尺(无手柄,作为永久元素)
double dx = _lineEnd.X - _lineStart.X, dy = _lineEnd.Y - _lineStart.Y;
double len = Math.Sqrt(dx * dx + dy * dy);
if (len < 2) return;
double ux = dx / len, uy = dy / len;
double px = -uy, py = ux;
double hw = _halfWidth;
var c1 = new Point(_lineStart.X + px * hw, _lineStart.Y + py * hw);
var c2 = new Point(_lineEnd.X + px * hw, _lineEnd.Y + py * hw);
var c3 = new Point(_lineEnd.X - px * hw, _lineEnd.Y - py * hw);
var c4 = new Point(_lineStart.X - px * hw, _lineStart.Y - py * hw);
// 矩形边框(半透明,不抢眼)
var border = new Polygon
{
Points = new System.Windows.Media.PointCollection { c1, c2, c3, c4 },
Stroke = CaliperStroke, StrokeThickness = 1, Opacity = 0.5,
Fill = System.Windows.Media.Brushes.Transparent, IsHitTestVisible = false
};
_mainCanvas.Children.Add(border);
_committedOverlays.Add(border);
// 等分线
int count = CaliperCount;
double step = len / (count + 1);
for (int i = 1; i <= count; i++)
{
double cx = _lineStart.X + ux * step * i, cy = _lineStart.Y + uy * step * i;
var line = new Line
{
X1 = cx + px * hw, Y1 = cy + py * hw,
X2 = cx - px * hw, Y2 = cy - py * hw,
Stroke = CaliperStroke, StrokeThickness = 1, Opacity = 0.3, IsHitTestVisible = false
};
_mainCanvas.Children.Add(line);
_committedOverlays.Add(line);
}
}
// ══════════════════════════════════════════════════════════════
// 绘制拟合结果(永久)
// ══════════════════════════════════════════════════════════════
private void DrawFitResult(LineFitResult fr)
{
if (_mainCanvas == null) return;
// 拟合直线(蓝色)
AddCommitted(new Line
{
X1 = fr.Endpoint1.X, Y1 = fr.Endpoint1.Y,
X2 = fr.Endpoint2.X, Y2 = fr.Endpoint2.Y,
Stroke = FitLineBrush, StrokeThickness = 2, IsHitTestVisible = false
});
// 标注
var lbl = new TextBlock
{
Text = $"∠{fr.AngleDegrees:F2}°",
Foreground = FitLineBrush, FontSize = 11, FontWeight = FontWeights.Bold, IsHitTestVisible = false
};
Canvas.SetLeft(lbl, (fr.Endpoint1.X + fr.Endpoint2.X) / 2 + 5);
Canvas.SetTop(lbl, (fr.Endpoint1.Y + fr.Endpoint2.Y) / 2 - 18);
AddCommitted(lbl);
}
private void AddCommitted(UIElement el)
{
_mainCanvas.Children.Add(el);
_committedOverlays.Add(el);
}
private void ClearCommitted()
{
if (_mainCanvas == null) return;
foreach (var el in _committedOverlays) _mainCanvas.Children.Remove(el);
_committedOverlays.Clear();
}
// ══════════════════════════════════════════════════════════════
// 临时卡尺可视化(编辑中,带手柄)
// ══════════════════════════════════════════════════════════════
private void RedrawTemp()
{
if (!_lineDefined || _mainCanvas == null) return;
ClearTempOverlays();
DrawTempCaliper();
}
private void DrawTempCaliper()
{
if (_mainCanvas == null) return;
double dx = _lineEnd.X - _lineStart.X, dy = _lineEnd.Y - _lineStart.Y;
double len = Math.Sqrt(dx * dx + dy * dy);
if (len < 2) return;
double ux = dx / len, uy = dy / len;
double px = -uy, py = ux;
double hw = _halfWidth;
var c1 = new Point(_lineStart.X + px * hw, _lineStart.Y + py * hw);
var c2 = new Point(_lineEnd.X + px * hw, _lineEnd.Y + py * hw);
var c3 = new Point(_lineEnd.X - px * hw, _lineEnd.Y - py * hw);
var c4 = new Point(_lineStart.X - px * hw, _lineStart.Y - py * hw);
// 矩形
AddTemp(new Polygon
{
Points = new System.Windows.Media.PointCollection { c1, c2, c3, c4 },
Stroke = CaliperStroke, StrokeThickness = 1, Fill = CaliperFill, IsHitTestVisible = false
});
// 搜索线虚线
AddTemp(new Line
{
X1 = _lineStart.X, Y1 = _lineStart.Y, X2 = _lineEnd.X, Y2 = _lineEnd.Y,
Stroke = CaliperStroke, StrokeThickness = 1,
StrokeDashArray = new DoubleCollection { 4, 3 }, IsHitTestVisible = false
});
// 等分线
int count = CaliperCount;
double step = len / (count + 1);
for (int i = 1; i <= count; i++)
{
double cx = _lineStart.X + ux * step * i, cy = _lineStart.Y + uy * step * i;
AddTemp(new Line
{
X1 = cx + px * hw, Y1 = cy + py * hw,
X2 = cx - px * hw, Y2 = cy - py * hw,
Stroke = CaliperStroke, StrokeThickness = 1, Opacity = 0.4, IsHitTestVisible = false
});
}
// 极性箭头
DrawPolarityArrow(px, py);
// 手柄位置
double midX = (_lineStart.X + _lineEnd.X) / 2, midY = (_lineStart.Y + _lineEnd.Y) / 2;
_handleStartPos = _lineStart;
_handleEndPos = _lineEnd;
_handleTopPos = new Point(midX + px * hw, midY + py * hw);
_handleBottomPos = new Point(midX - px * hw, midY - py * hw);
// 绘制手柄
AddTemp(MakeHandleVisual(_handleStartPos));
AddTemp(MakeHandleVisual(_handleEndPos));
AddTemp(MakeHandleVisual(_handleTopPos));
AddTemp(MakeHandleVisual(_handleBottomPos));
}
private void DrawPolarityArrow(double px, double py)
{
double midX = (_lineStart.X + _lineEnd.X) / 2, midY = (_lineStart.Y + _lineEnd.Y) / 2;
double arrowLen = Math.Min(_halfWidth * 0.6, 16);
if (EdgePolarity == "Both")
{
DrawArrow(midX, midY, px, py, arrowLen);
DrawArrow(midX, midY, -px, -py, arrowLen);
}
else if (EdgePolarity == "DarkToBright")
DrawArrow(midX, midY, px, py, arrowLen);
else
DrawArrow(midX, midY, -px, -py, arrowLen);
string txt = EdgePolarity switch { "BrightToDark" => "B→D", "DarkToBright" => "D→B", _ => "↔" };
var tb = new TextBlock { Text = txt, Foreground = CaliperStroke, FontSize = 10, IsHitTestVisible = false };
Canvas.SetLeft(tb, midX + px * (_halfWidth + 12));
Canvas.SetTop(tb, midY + py * (_halfWidth + 12) - 7);
AddTemp(tb);
}
private void DrawArrow(double fx, double fy, double dx, double dy, double length)
{
double tx = fx + dx * length, ty = fy + dy * length;
AddTemp(new Line { X1 = fx, Y1 = fy, X2 = tx, Y2 = ty, Stroke = CaliperStroke, StrokeThickness = 1.5, IsHitTestVisible = false });
double ang = Math.Atan2(dy, dx), hl = 5;
double a1 = ang + 2.5, a2 = ang - 2.5;
AddTemp(new Line { X1 = tx, Y1 = ty, X2 = tx + Math.Cos(a1) * hl, Y2 = ty + Math.Sin(a1) * hl, Stroke = CaliperStroke, StrokeThickness = 1.5, IsHitTestVisible = false });
AddTemp(new Line { X1 = tx, Y1 = ty, X2 = tx + Math.Cos(a2) * hl, Y2 = ty + Math.Sin(a2) * hl, Stroke = CaliperStroke, StrokeThickness = 1.5, IsHitTestVisible = false });
}
private Ellipse MakeHandleVisual(Point pos)
{
var h = new Ellipse
{
Width = HandleSize, Height = HandleSize,
Fill = HandleFill, Stroke = CaliperStroke, StrokeThickness = 2,
IsHitTestVisible = false
};
Canvas.SetLeft(h, pos.X - HandleSize / 2);
Canvas.SetTop(h, pos.Y - HandleSize / 2);
return h;
}
private void AddTemp(UIElement el) { _mainCanvas.Children.Add(el); _tempOverlays.Add(el); }
private void ClearTempOverlays()
{
if (_mainCanvas == null) return;
foreach (var el in _tempOverlays) _mainCanvas.Children.Remove(el);
_tempOverlays.Clear();
}
// ══════════════════════════════════════════════════════════════
// 统一鼠标交互
// ══════════════════════════════════════════════════════════════
private bool _interactionRegistered;
private void RegisterInteraction()
{
if (_canvas == null || _interactionRegistered) return;
_canvas.PreviewMouseLeftButtonDown += OnMouseDown;
_canvas.PreviewMouseMove += OnMouseMove;
_canvas.PreviewMouseLeftButtonUp += OnMouseUp;
_interactionRegistered = true;
}
private void UnregisterAll()
{
if (_canvas == null || !_interactionRegistered) return;
_canvas.PreviewMouseLeftButtonDown -= OnMouseDown;
_canvas.PreviewMouseMove -= OnMouseMove;
_canvas.PreviewMouseLeftButtonUp -= OnMouseUp;
_interactionRegistered = false;
_isDrawingLine = false;
_dragging = DragTarget.None;
}
private void OnMouseDown(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (_mainCanvas == null) return;
var pos = e.GetPosition(_mainCanvas);
// 绘制模式
if (_isDrawingLine)
{
_drawClickCount++;
if (_drawClickCount == 1)
{
_lineStart = pos;
ResultText = "请点击搜索线终点";
}
else if (_drawClickCount == 2)
{
_lineEnd = pos;
_isDrawingLine = false;
_lineDefined = true;
_halfWidth = DisplayWidth / 2.0;
FitCommand.RaiseCanExecuteChanged();
RedrawTemp();
ResultText = $"搜索线已定义 ({Len():F0}px)\n拖拽手柄调整,点击「拟合」执行";
}
e.Handled = true;
return;
}
// 拖拽模式
if (_lineDefined)
{
var target = HitTestHandle(pos);
if (target != DragTarget.None)
{
_dragging = target;
_canvas.CaptureMouse();
e.Handled = true;
}
}
}
private void OnMouseMove(object sender, System.Windows.Input.MouseEventArgs e)
{
if (_dragging == DragTarget.None || _mainCanvas == null) return;
var pos = e.GetPosition(_mainCanvas);
switch (_dragging)
{
case DragTarget.Start:
_lineStart = pos;
break;
case DragTarget.End:
_lineEnd = pos;
break;
case DragTarget.Top:
case DragTarget.Bottom:
double dist = PointToLineDist(pos, _lineStart, _lineEnd);
_halfWidth = Math.Max(5, dist);
SetProperty(ref _displayWidth, (int)(_halfWidth * 2), nameof(DisplayWidth));
break;
}
ClearTempOverlays();
DrawTempCaliper();
e.Handled = true;
}
private void OnMouseUp(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (_dragging == DragTarget.None) return;
_dragging = DragTarget.None;
_canvas.ReleaseMouseCapture();
ResultText = $"搜索线: {Len():F0}px, 宽度: {(int)(_halfWidth * 2)}px\n点击「拟合」执行";
e.Handled = true;
}
private DragTarget HitTestHandle(Point pos)
{
if (Dist(pos, _handleStartPos) <= HitRadius) return DragTarget.Start;
if (Dist(pos, _handleEndPos) <= HitRadius) return DragTarget.End;
if (Dist(pos, _handleTopPos) <= HitRadius) return DragTarget.Top;
if (Dist(pos, _handleBottomPos) <= HitRadius) return DragTarget.Bottom;
return DragTarget.None;
}
// ══════════════════════════════════════════════════════════════
// 辅助
// ══════════════════════════════════════════════════════════════
private double Len()
{
double dx = _lineEnd.X - _lineStart.X, dy = _lineEnd.Y - _lineStart.Y;
return Math.Sqrt(dx * dx + dy * dy);
}
private static double Dist(Point a, Point b)
{
double dx = a.X - b.X, dy = a.Y - b.Y;
return Math.Sqrt(dx * dx + dy * dy);
}
private static double PointToLineDist(Point p, Point a, Point b)
{
double abx = b.X - a.X, aby = b.Y - a.Y;
double len2 = abx * abx + aby * aby;
if (len2 < 1e-6) return Dist(p, a);
return Math.Abs(abx * (a.Y - p.Y) - aby * (a.X - p.X)) / Math.Sqrt(len2);
}
private static T FindChild<T>(DependencyObject parent, string name) where T : FrameworkElement
{
int count = VisualTreeHelper.GetChildrenCount(parent);
for (int i = 0; i < count; i++)
{
var child = VisualTreeHelper.GetChild(parent, i);
if (child is T t && t.Name == name) return t;
var r = FindChild<T>(child, name);
if (r != null) return r;
}
return null;
}
}
}
@@ -0,0 +1,507 @@
using System;
using System.Collections.Generic;
using System.IO;
using System.Windows;
using System.Windows.Media;
using System.Windows.Media.Imaging;
using Emgu.CV;
using Emgu.CV.Structure;
using Microsoft.Win32;
using Prism.Commands;
using Prism.Events;
using Prism.Ioc;
using Prism.Mvvm;
using XP.Common.Logging.Interfaces;
using XP.ImageProcessing.Processors;
using XplorePlane.Events;
using XplorePlane.Services.MainViewport;
using XplorePlane.ViewModels;
using Serilog;
namespace XplorePlane.ViewModels.ImageProcessing;
/// <summary>
/// 旋转模板匹配助手:框选 ROI、从 ROI 训练、参数、加载/保存模型、在当页试匹配;批量测试见 <see cref="Batch"/>。
/// </summary>
public class TemplateMatchAssistantViewModel : BindableBase, IDisposable
{
private static readonly Serilog.ILogger Log = Serilog.Log.ForContext<TemplateMatchAssistantViewModel>();
private readonly IEventAggregator _eventAggregator;
private readonly IContainerProvider _containerProvider;
private readonly ILoggerService _logger;
private readonly object _matcherLock = new();
private readonly TemplateMatchBatchViewModel _batch;
private SubscriptionToken? _roiDrawnToken;
private TemplateMatcherHandle? _matcher;
private bool _disposed;
private Int32Rect _pendingTemplateRoi;
private bool _hasPendingTemplateRoi;
private string _statusMessage = "可先加载模型,或点击「框选模板 ROI」后在主视图框选,再点「从 ROI 训练模板」。批量测试请切换到「批量测试」选项卡。";
private double _matchThreshold = 0.75;
private double _toleranceAngle;
private double _maxMatchCount = 5;
private double _maxOverlap = 0.3;
private double _minReduceArea = 256;
private bool _useSimd = true;
private bool _useSubPixel;
private bool _isModelReady;
public TemplateMatchAssistantViewModel(
IEventAggregator eventAggregator,
IContainerProvider containerProvider,
ILoggerService logger)
{
_eventAggregator = eventAggregator;
_containerProvider = containerProvider;
_logger = logger;
SelectTemplateRoiCommand = new DelegateCommand(ExecuteSelectTemplateRoi);
LearnFromRoiCommand = new DelegateCommand(ExecuteLearnFromRoi, () => _hasPendingTemplateRoi);
LoadModelCommand = new DelegateCommand(ExecuteLoadModel);
SaveModelCommand = new DelegateCommand(ExecuteSaveModel, () => _isModelReady && _matcher != null);
RunMatchCommand = new DelegateCommand(ExecuteRunMatch, () => _isModelReady && _matcher != null);
_batch = new TemplateMatchBatchViewModel(
this,
_eventAggregator,
_containerProvider.Resolve<IMainViewportService>());
_roiDrawnToken = _eventAggregator.GetEvent<TemplateMatchRoiDrawnEvent>()
.Subscribe(OnTemplateRoiDrawn, ThreadOption.UIThread);
}
/// <summary>批量测试子页(与助手共用同一模型与参数)。</summary>
public TemplateMatchBatchViewModel Batch => _batch;
public DelegateCommand SelectTemplateRoiCommand { get; }
public DelegateCommand LearnFromRoiCommand { get; }
public DelegateCommand LoadModelCommand { get; }
public DelegateCommand SaveModelCommand { get; }
public DelegateCommand RunMatchCommand { get; }
public string StatusMessage
{
get => _statusMessage;
set => SetProperty(ref _statusMessage, value);
}
public double MatchThreshold
{
get => _matchThreshold;
set => SetProperty(ref _matchThreshold, Math.Clamp(value, 0, 1));
}
public double ToleranceAngle
{
get => _toleranceAngle;
set => SetProperty(ref _toleranceAngle, Math.Clamp(value, 0, 180));
}
/// <summary>最大匹配数(滑块 1~100,运行匹配时取整)。</summary>
public double MaxMatchCount
{
get => _maxMatchCount;
set => SetProperty(ref _maxMatchCount, Math.Clamp(value, 1, 100));
}
public double MaxOverlap
{
get => _maxOverlap;
set => SetProperty(ref _maxOverlap, Math.Clamp(value, 0, 1));
}
/// <summary>金字塔最小面积(滑块 64~4096,运行匹配时取整)。</summary>
public double MinReduceArea
{
get => _minReduceArea;
set => SetProperty(ref _minReduceArea, Math.Clamp(value, 64, 4096));
}
public bool UseSimd
{
get => _useSimd;
set => SetProperty(ref _useSimd, value);
}
public bool UseSubPixel
{
get => _useSubPixel;
set => SetProperty(ref _useSubPixel, value);
}
public bool IsModelReady
{
get => _isModelReady;
private set
{
if (SetProperty(ref _isModelReady, value))
{
SaveModelCommand.RaiseCanExecuteChanged();
RunMatchCommand.RaiseCanExecuteChanged();
}
}
}
/// <summary>是否已有框选完成的模板 ROI(与是否已训练无关)。</summary>
public bool HasPendingTemplateRoi
{
get => _hasPendingTemplateRoi;
private set
{
if (SetProperty(ref _hasPendingTemplateRoi, value))
LearnFromRoiCommand.RaiseCanExecuteChanged();
}
}
/// <summary>
/// 清除主视图上框选的模板学习 ROI 叠加,并重置待训练 ROI 状态(运行匹配或开始批量测试前调用)。
/// </summary>
public void ClearTemplateLearningRoiOnViewport()
{
_eventAggregator.GetEvent<TemplateMatchClearRoiOverlayEvent>().Publish();
HasPendingTemplateRoi = false;
}
private void ExecuteSelectTemplateRoi()
{
_eventAggregator.GetEvent<TemplateMatchEnterRoiModeEvent>().Publish();
StatusMessage = "请在主视图图像上拖拽框选模板区域;框选完成后 ROI 会保留在图上,再点击「从 ROI 训练模板」。";
}
private void OnTemplateRoiDrawn(Int32Rect roi)
{
var viewportVm = _containerProvider.Resolve<ViewportPanelViewModel>();
var imageSource = viewportVm?.ImageSource as BitmapSource;
if (imageSource == null)
{
StatusMessage = "当前无主视图图像,无法记录模板 ROI。";
HasPendingTemplateRoi = false;
return;
}
int imgW = imageSource.PixelWidth;
int imgH = imageSource.PixelHeight;
int rx = Math.Clamp(roi.X, 0, Math.Max(0, imgW - 1));
int ry = Math.Clamp(roi.Y, 0, Math.Max(0, imgH - 1));
int rw = Math.Clamp(roi.Width, 1, Math.Max(1, imgW - rx));
int rh = Math.Clamp(roi.Height, 1, Math.Max(1, imgH - ry));
_pendingTemplateRoi = new Int32Rect(rx, ry, rw, rh);
HasPendingTemplateRoi = true;
StatusMessage = $"已框选模板区域 {rw}×{rh} 像素。请点击「从 ROI 训练模板」。";
}
private void ExecuteLearnFromRoi()
{
if (!HasPendingTemplateRoi) return;
try
{
var viewportVm = _containerProvider.Resolve<ViewportPanelViewModel>();
var imageSource = viewportVm?.ImageSource as BitmapSource;
if (imageSource == null)
{
StatusMessage = "当前无主视图图像,无法学习模板。";
return;
}
BitmapSource gray = imageSource.Format == PixelFormats.Gray8
? imageSource
: new FormatConvertedBitmap(imageSource, PixelFormats.Gray8, null, 0);
int imgW = gray.PixelWidth;
int imgH = gray.PixelHeight;
int rx = Math.Clamp(_pendingTemplateRoi.X, 0, imgW - 1);
int ry = Math.Clamp(_pendingTemplateRoi.Y, 0, imgH - 1);
int rw = Math.Clamp(_pendingTemplateRoi.Width, 1, imgW - rx);
int rh = Math.Clamp(_pendingTemplateRoi.Height, 1, imgH - ry);
var pixels = new byte[rw * rh];
gray.CopyPixels(new Int32Rect(rx, ry, rw, rh), pixels, rw, 0);
using var roiImage = new Image<Gray, byte>(rw, rh);
for (int y = 0; y < rh; y++)
for (int x = 0; x < rw; x++)
roiImage.Data[y, x, 0] = pixels[y * rw + x];
lock (_matcherLock)
{
_matcher?.Dispose();
_matcher = new TemplateMatcherHandle();
IntPtr p = roiImage.Mat.DataPointer;
int step = (int)roiImage.Mat.Step;
if (!_matcher.LearnPattern(p, rw, rh, step))
{
_matcher.Dispose();
_matcher = null;
IsModelReady = false;
StatusMessage = "模板学习失败。";
_logger.Warn("Template assistant: LearnPattern failed for ROI {0},{1},{2},{3}", rx, ry, rw, rh);
return;
}
}
IsModelReady = true;
StatusMessage = $"模板已从 ROI 学习成功({rw}×{rh} 像素)。可保存模型或运行匹配。";
_logger.Info("Template assistant: learned from ROI {0},{1},{2},{3}", rx, ry, rw, rh);
}
catch (Exception ex)
{
Log.Error(ex, "Template assistant ROI learn failed");
StatusMessage = $"学习失败: {ex.Message}";
IsModelReady = false;
}
}
private void ExecuteLoadModel()
{
var dlg = new OpenFileDialog
{
Title = "加载模板模型",
Filter = "模板模型|*.tmmodel;*.tm|所有文件|*.*"
};
if (dlg.ShowDialog() != true) return;
try
{
lock (_matcherLock)
{
_matcher?.Dispose();
_matcher = new TemplateMatcherHandle();
if (!_matcher.LoadModel(dlg.FileName))
{
_matcher.Dispose();
_matcher = null;
IsModelReady = false;
StatusMessage = "模型加载失败。";
return;
}
}
HasPendingTemplateRoi = false;
_eventAggregator.GetEvent<TemplateMatchClearRoiOverlayEvent>().Publish();
IsModelReady = true;
StatusMessage = $"已加载模型: {Path.GetFileName(dlg.FileName)}";
}
catch (Exception ex)
{
Log.Error(ex, "LoadModel failed");
StatusMessage = $"加载失败: {ex.Message}";
IsModelReady = false;
}
}
private void ExecuteSaveModel()
{
if (_matcher == null || !IsModelReady) return;
var dlg = new SaveFileDialog
{
Title = "保存模板模型",
Filter = "模板模型|*.tmmodel|所有文件|*.*",
DefaultExt = ".tmmodel"
};
if (dlg.ShowDialog() != true) return;
try
{
bool ok;
lock (_matcherLock)
ok = _matcher != null && _matcher.SaveModel(dlg.FileName);
if (ok)
StatusMessage = $"模型已保存: {dlg.FileName}";
else
StatusMessage = "模型保存失败。";
}
catch (Exception ex)
{
Log.Error(ex, "SaveModel failed");
StatusMessage = $"保存失败: {ex.Message}";
}
}
private void ExecuteRunMatch()
{
if (_matcher == null || !IsModelReady) return;
ClearTemplateLearningRoiOnViewport();
try
{
var viewportVm = _containerProvider.Resolve<ViewportPanelViewModel>();
var imageSource = viewportVm?.ImageSource as BitmapSource;
if (imageSource == null)
{
StatusMessage = "当前无主视图图像。";
return;
}
using Image<Gray, byte>? full = BitmapSourceToGrayImage(imageSource);
if (full == null) return;
bool forcedSubPixelOff = UseSubPixel &&
Math.Abs(ToleranceAngle) > TM_Params.SubPixelAngleSafetyLimitDegrees;
int templatePixels = 0;
lock (_matcherLock)
{
if (_matcher != null && _matcher.GetTemplateInfo(out int tw, out int th, out _))
templatePixels = Math.Max(0, tw) * Math.Max(0, th);
}
bool bumpedMinReduce = templatePixels >= 512 &&
(int)Math.Clamp(Math.Round(MinReduceArea), 64, 4096) < 256;
if (!TryMatchGrayImage(full, out IReadOnlyList<TemplateMatchHitDto> hitsReadonly, out double t, out string? matchErr))
{
if (matchErr != null && matchErr.Contains("TemplateMatchLib", StringComparison.OrdinalIgnoreCase))
StatusMessage = "未找到 TemplateMatchLib.dll,请确认已复制到输出目录。";
else
StatusMessage = string.IsNullOrEmpty(matchErr) ? "匹配失败。" : matchErr;
_eventAggregator.GetEvent<TemplateMatchPreviewResultEvent>().Publish(new TemplateMatchPreviewPayload());
return;
}
var hits = new List<TemplateMatchHitDto>(hitsReadonly);
_eventAggregator.GetEvent<TemplateMatchPreviewResultEvent>().Publish(
new TemplateMatchPreviewPayload { Hits = hits, MatchTimeMs = t });
StatusMessage = hits.Count == 0
? $"未找到匹配(耗时 {t:F1} ms)。可调低阈值或角度范围。"
: $"匹配到 {hits.Count} 个目标,耗时 {t:F1} ms。";
if (forcedSubPixelOff)
StatusMessage += $" 已自动关闭亚像素(角度容差>{TM_Params.SubPixelAngleSafetyLimitDegrees:F0}° 时匹配库易崩溃)。";
if (bumpedMinReduce)
StatusMessage += " 已将金字塔最小面积提升至不低于 256(与库默认一致)。";
}
catch (Exception ex)
{
Log.Error(ex, "RunMatch failed");
StatusMessage = $"匹配失败: {ex.Message}";
_eventAggregator.GetEvent<TemplateMatchPreviewResultEvent>().Publish(new TemplateMatchPreviewPayload());
}
}
private TM_Params BuildCurrentMatchParams() => new TM_Params
{
Score = MatchThreshold,
ToleranceAngle = ToleranceAngle,
MaxOverlap = MaxOverlap,
MaxCount = (int)Math.Clamp(Math.Round(MaxMatchCount), 1, 100),
MinReduceArea = (int)Math.Clamp(Math.Round(MinReduceArea), 64, 4096),
UseSIMD = UseSimd ? 1 : 0,
UseSubPixel = UseSubPixel ? 1 : 0
};
/// <summary>
/// 使用当前助手参数对灰度图做模板匹配(与单张「运行匹配」一致)。供批量测试在后台线程调用,内部已加锁。
/// </summary>
public bool TryMatchGrayImage(Image<Gray, byte> fullImage, out IReadOnlyList<TemplateMatchHitDto> hits,
out double matchTimeMs, out string? errorMessage)
{
hits = Array.Empty<TemplateMatchHitDto>();
matchTimeMs = 0;
errorMessage = null;
if (fullImage == null)
{
errorMessage = "图像为空";
return false;
}
lock (_matcherLock)
{
if (_matcher == null || !IsModelReady)
{
errorMessage = "模型未就绪";
return false;
}
try
{
var param = BuildCurrentMatchParams();
var results = _matcher.Match(
fullImage.Mat.DataPointer,
fullImage.Width,
fullImage.Height,
(int)fullImage.Mat.Step,
param);
matchTimeMs = _matcher.LastMatchTime;
var list = new List<TemplateMatchHitDto>();
foreach (var r in results)
{
list.Add(new TemplateMatchHitDto
{
CenterX = r.CenterX,
CenterY = r.CenterY,
Angle = r.Angle,
Score = r.Score,
LtX = r.LtX, LtY = r.LtY,
RtX = r.RtX, RtY = r.RtY,
RbX = r.RbX, RbY = r.RbY,
LbX = r.LbX, LbY = r.LbY
});
}
hits = list;
return true;
}
catch (DllNotFoundException)
{
errorMessage = "TemplateMatchLib.dll 未找到";
return false;
}
catch (Exception ex)
{
Log.Error(ex, "TryMatchGrayImage failed");
errorMessage = ex.Message;
return false;
}
}
}
internal static Image<Gray, byte>? BitmapSourceToGrayImage(BitmapSource bitmapSource)
{
BitmapSource source = bitmapSource.Format == PixelFormats.Gray8
? bitmapSource
: new FormatConvertedBitmap(bitmapSource, PixelFormats.Gray8, null, 0);
int width = source.PixelWidth;
int height = source.PixelHeight;
if (width < 1 || height < 1) return null;
int stride = width;
var pixels = new byte[width * height];
source.CopyPixels(pixels, stride, 0);
var image = new Image<Gray, byte>(width, height);
for (int y = 0; y < height; y++)
for (int x = 0; x < width; x++)
image.Data[y, x, 0] = pixels[y * stride + x];
return image;
}
public void Dispose()
{
if (_disposed) return;
_disposed = true;
if (_roiDrawnToken != null)
{
_eventAggregator.GetEvent<TemplateMatchRoiDrawnEvent>().Unsubscribe(_roiDrawnToken);
_roiDrawnToken = null;
}
_batch.Dispose();
lock (_matcherLock)
{
_matcher?.Dispose();
_matcher = null;
}
}
}
@@ -0,0 +1,340 @@
using System;
using System.Collections.Generic;
using System.Collections.ObjectModel;
using System.ComponentModel;
using System.IO;
using System.Linq;
using System.Threading;
using System.Threading.Tasks;
using System.Windows.Media.Imaging;
using Emgu.CV;
using Emgu.CV.CvEnum;
using Emgu.CV.Structure;
using Microsoft.Win32;
using Prism.Commands;
using Prism.Events;
using Prism.Mvvm;
using XplorePlane.Events;
using XplorePlane.Services.MainViewport;
namespace XplorePlane.ViewModels.ImageProcessing;
/// <summary>单张批量匹配结果行(绑定到 DataGrid)。</summary>
public sealed class TemplateMatchBatchRow
{
public string FileName { get; init; } = "";
public string FullPath { get; init; } = "";
/// <summary>简要结果:命中 / 未找到 / 失败原因摘要。</summary>
public string Result { get; init; } = "";
public int MatchCount { get; init; }
public double BestScore { get; init; }
public double TimeMs { get; init; }
public string? ErrorDetail { get; init; }
public IReadOnlyList<TemplateMatchHitDto> Hits { get; init; } = Array.Empty<TemplateMatchHitDto>();
}
/// <summary>
/// 模板匹配批量测试:扫描文件夹、逐张匹配(与助手当前参数一致)、在主视口打开选中结果。
/// </summary>
public class TemplateMatchBatchViewModel : BindableBase, IDisposable
{
private static readonly Serilog.ILogger Log = Serilog.Log.ForContext<TemplateMatchBatchViewModel>();
private static readonly HashSet<string> ImageExtensions = new(StringComparer.OrdinalIgnoreCase)
{
".bmp", ".png", ".jpg", ".jpeg", ".tif", ".tiff"
};
private readonly TemplateMatchAssistantViewModel _assistant;
private readonly IEventAggregator _eventAggregator;
private readonly IMainViewportService _mainViewportService;
private readonly List<string> _imagePaths = new();
private CancellationTokenSource? _batchCts;
private bool _disposed;
private bool _isRunning;
private string _folderPath = "";
private string _batchStatusText = "请选择文件夹后点击「开始批量匹配」。";
private TemplateMatchBatchRow? _selectedRow;
private int _imageFileCount;
public TemplateMatchBatchViewModel(
TemplateMatchAssistantViewModel assistant,
IEventAggregator eventAggregator,
IMainViewportService mainViewportService)
{
_assistant = assistant ?? throw new ArgumentNullException(nameof(assistant));
_eventAggregator = eventAggregator ?? throw new ArgumentNullException(nameof(eventAggregator));
_mainViewportService = mainViewportService ?? throw new ArgumentNullException(nameof(mainViewportService));
_assistant.PropertyChanged += OnAssistantPropertyChanged;
PickFolderCommand = new DelegateCommand(ExecutePickFolder);
StartBatchCommand = new DelegateCommand(ExecuteStartBatch, CanStartBatch);
StopBatchCommand = new DelegateCommand(ExecuteStopBatch, () => _isRunning);
OpenSelectedInMainViewportCommand = new DelegateCommand(ExecuteOpenSelectedInMainViewport, () => SelectedRow != null);
}
private void OnAssistantPropertyChanged(object? sender, PropertyChangedEventArgs e)
{
if (e.PropertyName == nameof(TemplateMatchAssistantViewModel.IsModelReady))
StartBatchCommand.RaiseCanExecuteChanged();
}
public ObservableCollection<TemplateMatchBatchRow> Rows { get; } = new();
public DelegateCommand PickFolderCommand { get; }
public DelegateCommand StartBatchCommand { get; }
public DelegateCommand StopBatchCommand { get; }
public DelegateCommand OpenSelectedInMainViewportCommand { get; }
public string FolderPath
{
get => _folderPath;
private set => SetProperty(ref _folderPath, value);
}
public string BatchStatusText
{
get => _batchStatusText;
private set => SetProperty(ref _batchStatusText, value);
}
public int ImageFileCount
{
get => _imageFileCount;
private set => SetProperty(ref _imageFileCount, value);
}
public bool IsRunning
{
get => _isRunning;
private set
{
if (SetProperty(ref _isRunning, value))
{
StartBatchCommand.RaiseCanExecuteChanged();
StopBatchCommand.RaiseCanExecuteChanged();
}
}
}
public TemplateMatchBatchRow? SelectedRow
{
get => _selectedRow;
set
{
if (SetProperty(ref _selectedRow, value))
OpenSelectedInMainViewportCommand.RaiseCanExecuteChanged();
}
}
/// <summary>供宿主窗口 DataGrid 双击调用。</summary>
public void OpenSelectedFromDoubleClick() => OpenSelectedInMainViewportCommand.Execute();
private bool CanStartBatch() =>
!_isRunning && _imagePaths.Count > 0 && _assistant.IsModelReady;
private void ExecutePickFolder()
{
var dlg = new OpenFolderDialog
{
Title = "选择待测试图像所在文件夹(仅当前目录,不含子文件夹)",
InitialDirectory = Directory.Exists(_folderPath) ? _folderPath : Environment.GetFolderPath(Environment.SpecialFolder.MyPictures)
};
if (dlg.ShowDialog() != true)
return;
ScanFolder(dlg.FolderName);
}
private void ScanFolder(string folder)
{
_imagePaths.Clear();
if (!Directory.Exists(folder))
{
FolderPath = "";
ImageFileCount = 0;
BatchStatusText = "文件夹不存在。";
StartBatchCommand.RaiseCanExecuteChanged();
return;
}
foreach (var path in Directory.EnumerateFiles(folder, "*.*", SearchOption.TopDirectoryOnly)
.OrderBy(p => p, StringComparer.OrdinalIgnoreCase))
{
if (ImageExtensions.Contains(Path.GetExtension(path)))
_imagePaths.Add(path);
}
FolderPath = folder;
ImageFileCount = _imagePaths.Count;
BatchStatusText = ImageFileCount == 0
? "该文件夹下没有支持的图像文件(bmp/png/jpg/tif…)。"
: $"已扫描 {ImageFileCount} 个图像文件,可开始批量匹配。";
StartBatchCommand.RaiseCanExecuteChanged();
}
private async void ExecuteStartBatch()
{
if (!CanStartBatch()) return;
_assistant.ClearTemplateLearningRoiOnViewport();
Rows.Clear();
IsRunning = true;
_batchCts = new CancellationTokenSource();
var token = _batchCts.Token;
int total = _imagePaths.Count;
int index = 0;
try
{
foreach (var path in _imagePaths)
{
token.ThrowIfCancellationRequested();
index++;
BatchStatusText = $"正在处理 {index}/{total}{Path.GetFileName(path)}";
var row = await Task.Run(() => ProcessOneFile(path), token).ConfigureAwait(true);
Rows.Add(row);
}
BatchStatusText = $"完成,共处理 {Rows.Count} 张。";
}
catch (OperationCanceledException)
{
BatchStatusText = $"已停止(已处理 {Rows.Count}/{total} 张)。";
}
catch (Exception ex)
{
Log.Error(ex, "Batch template match failed");
BatchStatusText = $"批量过程异常:{ex.Message}";
}
finally
{
IsRunning = false;
_batchCts?.Dispose();
_batchCts = null;
}
}
private void ExecuteStopBatch() => _batchCts?.Cancel();
private TemplateMatchBatchRow ProcessOneFile(string path)
{
var fileName = Path.GetFileName(path);
try
{
using var mat = CvInvoke.Imread(path, ImreadModes.Grayscale);
if (mat == null || mat.IsEmpty)
{
return new TemplateMatchBatchRow
{
FileName = fileName,
FullPath = path,
Result = "无法读取",
ErrorDetail = "Imread 为空或失败"
};
}
using var gray = mat.ToImage<Gray, byte>();
if (!_assistant.TryMatchGrayImage(gray, out var hits, out var t, out var err))
{
return new TemplateMatchBatchRow
{
FileName = fileName,
FullPath = path,
Result = "失败",
ErrorDetail = err,
TimeMs = t
};
}
int c = hits.Count;
double best = c == 0 ? 0 : hits.Max(h => h.Score);
return new TemplateMatchBatchRow
{
FileName = fileName,
FullPath = path,
Result = c > 0 ? "命中" : "未找到",
MatchCount = c,
BestScore = best,
TimeMs = t,
Hits = CloneHits(hits)
};
}
catch (Exception ex)
{
Log.Warning(ex, "ProcessOneFile: {Path}", path);
return new TemplateMatchBatchRow
{
FileName = fileName,
FullPath = path,
Result = "异常",
ErrorDetail = ex.Message
};
}
}
private static List<TemplateMatchHitDto> CloneHits(IReadOnlyList<TemplateMatchHitDto> src)
{
var list = new List<TemplateMatchHitDto>(src.Count);
foreach (var h in src)
{
list.Add(new TemplateMatchHitDto
{
CenterX = h.CenterX,
CenterY = h.CenterY,
Angle = h.Angle,
Score = h.Score,
LtX = h.LtX, LtY = h.LtY,
RtX = h.RtX, RtY = h.RtY,
RbX = h.RbX, RbY = h.RbY,
LbX = h.LbX, LbY = h.LbY
});
}
return list;
}
private void ExecuteOpenSelectedInMainViewport()
{
if (SelectedRow == null || string.IsNullOrWhiteSpace(SelectedRow.FullPath) || !File.Exists(SelectedRow.FullPath))
return;
try
{
var bitmap = new BitmapImage();
bitmap.BeginInit();
bitmap.UriSource = new Uri(SelectedRow.FullPath, UriKind.Absolute);
bitmap.CacheOption = BitmapCacheOption.OnLoad;
bitmap.EndInit();
bitmap.Freeze();
_mainViewportService.SetManualImage(bitmap, SelectedRow.FullPath);
_eventAggregator.GetEvent<ManualImageLoadedEvent>()
.Publish(new ManualImageLoadedPayload(bitmap, SelectedRow.FullPath));
var hits = CloneHits(SelectedRow.Hits);
_eventAggregator.GetEvent<TemplateMatchPreviewResultEvent>().Publish(
new TemplateMatchPreviewPayload { Hits = hits, MatchTimeMs = SelectedRow.TimeMs });
BatchStatusText = $"已在主视图打开:{SelectedRow.FileName}";
}
catch (Exception ex)
{
Log.Error(ex, "OpenSelectedInMainViewport");
BatchStatusText = $"打开失败:{ex.Message}";
}
}
public void Dispose()
{
if (_disposed) return;
_disposed = true;
_assistant.PropertyChanged -= OnAssistantPropertyChanged;
_batchCts?.Cancel();
_batchCts?.Dispose();
_batchCts = null;
}
}
+192 -6
View File
@@ -36,6 +36,13 @@ using XP.Common.PdfViewer.Interfaces;
using XP.Hardware.MotionControl.Abstractions;
using XP.Hardware.MotionControl.Services;
using System.Windows.Threading;
using XP.ImageProcessing.Processors;
using XplorePlane.Services.Storage;
using XplorePlane.ViewModels.Cnc;
using XplorePlane.ViewModels.ImageProcessing;
using XplorePlane.Views;
using XplorePlane.Views.Cnc;
using XplorePlane.Views.ImageProcessing;
namespace XplorePlane.ViewModels
{
@@ -180,9 +187,12 @@ namespace XplorePlane.ViewModels
public DelegateCommand WhiteBackgroundDetectionCommand { get; }
public DelegateCommand BlackBackgroundDetectionCommand { get; }
public DelegateCommand OpenTemplateMatchAssistantCommand { get; }
public DelegateCommand GrayscaleCommand { get; }
public DelegateCommand SharpenCommand { get; }
public DelegateCommand EnhanceCommand { get; }
public DelegateCommand EdgeLineFitCommand { get; }
public DelegateCommand EdgeCircleFitCommand { get; }
// 设置命令
public DelegateCommand OpenLanguageSwitcherCommand { get; }
@@ -277,6 +287,7 @@ namespace XplorePlane.ViewModels
private Window _reportConfigWindow;
private Window _inspectionReportViewerWindow;
private Window _debugPanelWindow;
private Window _templateMatchAssistantWindow;
private object _imagePanelContent;
private GridLength _viewportPanelWidth = new(1, GridUnitType.Star);
private GridLength _imagePanelWidth = new(320);
@@ -332,6 +343,12 @@ namespace XplorePlane.ViewModels
_eventAggregator.GetEvent<PipelinePreviewUpdatedEvent>()
.Subscribe(OnPipelinePreviewUpdated, ThreadOption.UIThread);
_eventAggregator.GetEvent<WhiteBackgroundRoiDrawnEvent>()
.Subscribe(OnWhiteBackgroundRoiDrawn, ThreadOption.UIThread);
_eventAggregator.GetEvent<BlackBackgroundRoiDrawnEvent>()
.Subscribe(OnBlackBackgroundRoiDrawn, ThreadOption.UIThread);
NavigationTree = new ObservableCollection<object>();
NavigateHomeCommand = new DelegateCommand(OnNavigateHome);
@@ -394,9 +411,12 @@ namespace XplorePlane.ViewModels
// 图像处理命令
WhiteBackgroundDetectionCommand = new DelegateCommand(ExecuteWhiteBackgroundDetection);
BlackBackgroundDetectionCommand = new DelegateCommand(ExecuteBlackBackgroundDetection);
OpenTemplateMatchAssistantCommand = new DelegateCommand(ExecuteOpenTemplateMatchAssistant);
GrayscaleCommand = new DelegateCommand(ExecuteGrayscale);
SharpenCommand = new DelegateCommand(ExecuteSharpen);
EnhanceCommand = new DelegateCommand(ExecuteEnhance);
EdgeLineFitCommand = new DelegateCommand(ExecuteEdgeLineFit);
EdgeCircleFitCommand = new DelegateCommand(ExecuteEdgeCircleFit);
AxisResetCommand = new DelegateCommand(ExecuteAxisReset);
OpenDoorCommand = new DelegateCommand(ExecuteOpenDoor);
@@ -1045,6 +1065,8 @@ namespace XplorePlane.ViewModels
}
private Window _bgaDetectionPanel;
private Window _edgeLineFitPanel;
private Window _edgeCircleFitPanel;
private void ExecuteBgaDetection()
{
@@ -1116,22 +1138,148 @@ namespace XplorePlane.ViewModels
private void ExecuteWhiteBackgroundDetection()
{
if (!CheckImageLoaded()) return;
_logger.Info("White background detection triggered.");
// TODO: 实现白底检测逻辑
_logger.Info("White background detection: entering ROI draw mode.");
_eventAggregator.GetEvent<WhiteBackgroundDetectionEvent>().Publish();
StatusMessage = "白底检测:请在图像上拖拽绘制矩形ROI";
}
private void OnWhiteBackgroundRoiDrawn(System.Windows.Int32Rect roi) =>
RunBackgroundRoiDetection(roi, BackgroundDefectMode.WhiteBackground);
private void ExecuteBlackBackgroundDetection()
{
if (!CheckImageLoaded()) return;
_logger.Info("Black background detection triggered.");
// TODO: 实现黑底检测逻辑
_logger.Info("Black background detection: entering ROI draw mode.");
_eventAggregator.GetEvent<BlackBackgroundDetectionEvent>().Publish();
StatusMessage = "黑底检测:请在图像上拖拽绘制矩形ROI";
}
private void ExecuteOpenTemplateMatchAssistant()
{
try
{
if (!CheckImageLoaded())
{
StatusMessage = "请先加载图像再使用模板助手。";
return;
}
if (_templateMatchAssistantWindow != null)
{
if (_templateMatchAssistantWindow.IsLoaded)
{
_templateMatchAssistantWindow.Activate();
return;
}
_templateMatchAssistantWindow = null;
}
var vm = _containerProvider.Resolve<TemplateMatchAssistantViewModel>();
var w = new TemplateMatchAssistantWindow
{
DataContext = vm,
Owner = Application.Current?.MainWindow
};
w.Closed += (_, _) => { _templateMatchAssistantWindow = null; };
_templateMatchAssistantWindow = w;
w.Show();
_logger.Info("Template match assistant opened.");
StatusMessage = "已打开模板匹配助手";
}
catch (Exception ex)
{
_logger.Error(ex, "Failed to open template match assistant");
StatusMessage = $"打开模板助手失败: {ex.Message}";
}
}
private void OnBlackBackgroundRoiDrawn(System.Windows.Int32Rect roi) =>
RunBackgroundRoiDetection(roi, BackgroundDefectMode.BlackBackground);
/// <summary>
/// 从视口灰度图取 ROI,调用 <see cref="BackgroundDefectAnalyzer"/>,再发布结果事件(全局坐标)。
/// </summary>
private void RunBackgroundRoiDetection(System.Windows.Int32Rect roi, BackgroundDefectMode mode)
{
try
{
var viewportVm = _containerProvider.Resolve<ViewportPanelViewModel>();
var imageSource = viewportVm?.ImageSource as System.Windows.Media.Imaging.BitmapSource;
if (imageSource == null) return;
System.Windows.Media.Imaging.BitmapSource gray8;
if (imageSource.Format != System.Windows.Media.PixelFormats.Gray8)
gray8 = new System.Windows.Media.Imaging.FormatConvertedBitmap(
imageSource, System.Windows.Media.PixelFormats.Gray8, null, 0);
else
gray8 = imageSource;
int imgW = gray8.PixelWidth;
int imgH = gray8.PixelHeight;
int rx = Math.Clamp(roi.X, 0, imgW - 1);
int ry = Math.Clamp(roi.Y, 0, imgH - 1);
int rw = Math.Clamp(roi.Width, 1, imgW - rx);
int rh = Math.Clamp(roi.Height, 1, imgH - ry);
byte[] roiPixels = new byte[rw * rh];
gray8.CopyPixels(new System.Windows.Int32Rect(rx, ry, rw, rh), roiPixels, rw, 0);
using var roiImage = new Image<Gray, byte>(rw, rh);
for (int y = 0; y < rh; y++)
for (int x = 0; x < rw; x++)
roiImage.Data[y, x, 0] = roiPixels[y * rw + x];
const int minArea = 50;
const double mmPerPixel = 0.139;
var blobs = BackgroundDefectAnalyzer.DetectBlobs(roiImage, mode, minArea, mmPerPixel);
var detections = new System.Collections.Generic.List<BackgroundDefectDetectionItem>(blobs.Count);
foreach (var b in blobs)
{
var item = new BackgroundDefectDetectionItem
{
SizeMicrometers = b.MaxChordMicrometers,
ChordP1 = new System.Drawing.Point(b.MaxChordEndAInRoi.X + rx, b.MaxChordEndAInRoi.Y + ry),
ChordP2 = new System.Drawing.Point(b.MaxChordEndBInRoi.X + rx, b.MaxChordEndBInRoi.Y + ry)
};
foreach (var p in b.ContourInRoi)
item.Contour.Add(new System.Drawing.Point(p.X + rx, p.Y + ry));
detections.Add(item);
}
var roiRect = new System.Drawing.Rectangle(rx, ry, rw, rh);
if (mode == BackgroundDefectMode.WhiteBackground)
{
_eventAggregator.GetEvent<WhiteBackgroundResultEvent>().Publish(
new WhiteBackgroundResultPayload { RoiRect = roiRect, Detections = detections });
StatusMessage = $"白底检测完成:检测到 {detections.Count} 个黑色区域";
_logger.Info("White background detection: found {Count} dark regions in ROI ({X},{Y},{W},{H})",
detections.Count, rx, ry, rw, rh);
}
else
{
_eventAggregator.GetEvent<BlackBackgroundResultEvent>().Publish(
new BlackBackgroundResultPayload { RoiRect = roiRect, Detections = detections });
StatusMessage = $"黑底检测完成:检测到 {detections.Count} 个亮色区域";
_logger.Info("Black background detection: found {Count} bright regions in ROI ({X},{Y},{W},{H})",
detections.Count, rx, ry, rw, rh);
}
}
catch (Exception ex)
{
string label = mode == BackgroundDefectMode.WhiteBackground ? "白底" : "黑底";
_logger.Error(ex, "{Label} background detection failed", label);
StatusMessage = $"{label}检测失败: {ex.Message}";
}
}
private void ExecuteGrayscale()
{
if (!CheckImageLoaded()) return;
_logger.Info("Grayscale conversion triggered.");
// TODO: 实现灰度转换逻辑
_logger.Info("Line profile toggled.");
_eventAggregator.GetEvent<ToggleLineProfileEvent>().Publish();
}
private void ExecuteSharpen()
@@ -1203,6 +1351,44 @@ namespace XplorePlane.ViewModels
}
}
private void ExecuteEdgeLineFit()
{
if (!CheckImageLoaded()) return;
_logger.Info("边缘查找拟合直线功能已触发");
if (_edgeLineFitPanel != null && _edgeLineFitPanel.IsVisible)
{
_edgeLineFitPanel.Activate();
return;
}
_edgeLineFitPanel = new Views.ImageProcessing.EdgeLineFitPanel
{
Owner = System.Windows.Application.Current.MainWindow
};
_edgeLineFitPanel.Closed += (_, _) => { _edgeLineFitPanel = null; };
_edgeLineFitPanel.Show();
}
private void ExecuteEdgeCircleFit()
{
if (!CheckImageLoaded()) return;
_logger.Info("边缘查找拟合圆功能已触发");
if (_edgeCircleFitPanel != null && _edgeCircleFitPanel.IsVisible)
{
_edgeCircleFitPanel.Activate();
return;
}
_edgeCircleFitPanel = new Views.ImageProcessing.EdgeCircleFitPanel
{
Owner = System.Windows.Application.Current.MainWindow
};
_edgeCircleFitPanel.Closed += (_, _) => { _edgeCircleFitPanel = null; };
_edgeCircleFitPanel.Show();
}
private Image<Gray, byte>? BitmapSourceToImage(BitmapSource bitmapSource)
{
// 转换为可用的图像格式
@@ -55,9 +55,6 @@ namespace XplorePlane.ViewModels
StopGrabCommand.RaiseCanExecuteChanged();
ApplyExposureCommand.RaiseCanExecuteChanged();
ApplyGainCommand.RaiseCanExecuteChanged();
ApplyWidthCommand.RaiseCanExecuteChanged();
ApplyHeightCommand.RaiseCanExecuteChanged();
ApplyPixelFormatCommand.RaiseCanExecuteChanged();
RefreshCameraParamsCommand.RaiseCanExecuteChanged();
OpenCameraSettingsCommand.RaiseCanExecuteChanged();
}
@@ -79,7 +76,7 @@ namespace XplorePlane.ViewModels
}
}
private string _cameraStatusText = "未连接";
private string _cameraStatusText = "Disconnected";
public string CameraStatusText
{
@@ -152,8 +149,6 @@ namespace XplorePlane.ViewModels
set => SetProperty(ref _selectedPixelFormat, value);
}
public ObservableCollection<string> PixelFormatOptions { get; } = new() { "Mono8", "BGR8", "BGRA8" };
#endregion Properties
#region Commands
@@ -164,9 +159,6 @@ namespace XplorePlane.ViewModels
public DelegateCommand StopGrabCommand { get; }
public DelegateCommand ApplyExposureCommand { get; }
public DelegateCommand ApplyGainCommand { get; }
public DelegateCommand ApplyWidthCommand { get; }
public DelegateCommand ApplyHeightCommand { get; }
public DelegateCommand ApplyPixelFormatCommand { get; }
public DelegateCommand RefreshCameraParamsCommand { get; }
public DelegateCommand OpenCameraSettingsCommand { get; }
@@ -183,9 +175,6 @@ namespace XplorePlane.ViewModels
StopGrabCommand = new DelegateCommand(StopGrab, () => IsCameraGrabbing);
ApplyExposureCommand = new DelegateCommand(() => ApplyCameraParam(() => _camera.SetExposureTime(ExposureTime)), () => IsCameraConnected);
ApplyGainCommand = new DelegateCommand(() => ApplyCameraParam(() => _camera.SetGain(GainValue)), () => IsCameraConnected);
ApplyWidthCommand = new DelegateCommand(() => ApplyCameraParam(() => _camera.SetWidth(ImageWidth)), () => IsCameraConnected);
ApplyHeightCommand = new DelegateCommand(() => ApplyCameraParam(() => _camera.SetHeight(ImageHeight)), () => IsCameraConnected);
ApplyPixelFormatCommand = new DelegateCommand(() => ApplyCameraParam(() => _camera.SetPixelFormat(SelectedPixelFormat)), () => IsCameraConnected);
RefreshCameraParamsCommand = new DelegateCommand(RefreshCameraParams, () => IsCameraConnected);
OpenCameraSettingsCommand = new DelegateCommand(OpenCameraSettings, () => IsCameraConnected);
@@ -193,7 +182,7 @@ namespace XplorePlane.ViewModels
_defaultImageSource = new BitmapImage(new Uri("pack://application:,,,/Assets/Icons/NoCamera.png"));
_cameraImageSource = _defaultImageSource;
CameraStatusText = "正在检索相机...";
CameraStatusText = "Searching camera...";
}
/// <summary>
@@ -203,7 +192,7 @@ namespace XplorePlane.ViewModels
{
if (!_camera.IsConnected)
{
CameraStatusText = "未检测到相机";
CameraStatusText = "No camera detected";
return;
}
@@ -212,7 +201,7 @@ namespace XplorePlane.ViewModels
_camera.ConnectionLost += OnCameraConnectionLost;
IsCameraConnected = true;
CameraStatusText = "已连接";
CameraStatusText = "Connected";
RefreshCameraParams();
SyncCameraStateToAppState();
StartGrab();
@@ -231,7 +220,7 @@ namespace XplorePlane.ViewModels
var info = _camera.Open();
IsCameraConnected = true;
CameraStatusText = $"已连接: {info.ModelName} (SN: {info.SerialNumber})";
CameraStatusText = $"Connected: {info.ModelName} (SN: {info.SerialNumber})";
_logger.Information("Camera connected: {ModelName}", info.ModelName);
RefreshCameraParams();
SyncCameraStateToAppState();
@@ -239,7 +228,7 @@ namespace XplorePlane.ViewModels
catch (Exception ex)
{
_logger.Error(ex, "Failed to connect camera");
CameraStatusText = $"连接失败: {ex.Message}";
CameraStatusText = $"Connection failed: {ex.Message}";
IsCameraConnected = false;
SyncCameraStateToAppState();
}
@@ -263,7 +252,7 @@ namespace XplorePlane.ViewModels
_camera.ConnectionLost -= OnCameraConnectionLost;
IsCameraConnected = false;
IsCameraGrabbing = false;
CameraStatusText = "未连接";
CameraStatusText = "Disconnected";
CameraImageSource = null;
SyncCameraStateToAppState();
_logger.Information("Camera disconnected");
@@ -276,7 +265,7 @@ namespace XplorePlane.ViewModels
{
_camera.StartGrabbing();
IsCameraGrabbing = true;
CameraStatusText = "采集中...";
CameraStatusText = "Grabbing...";
SyncCameraStateToAppState();
// 如果已勾选实时,自动启动 Live View
@@ -288,7 +277,7 @@ namespace XplorePlane.ViewModels
catch (Exception ex)
{
_logger.Error(ex, "Failed to start grabbing");
CameraStatusText = $"采集失败: {ex.Message}";
CameraStatusText = $"Grab failed: {ex.Message}";
}
}
@@ -299,7 +288,7 @@ namespace XplorePlane.ViewModels
IsLiveViewEnabled = false;
_camera.StopGrabbing();
IsCameraGrabbing = false;
CameraStatusText = "已停止采集";
CameraStatusText = "Grab stopped";
SyncCameraStateToAppState();
}
catch (Exception ex)
@@ -313,7 +302,7 @@ namespace XplorePlane.ViewModels
if (!IsCameraGrabbing) return;
_liveViewRunning = true;
CameraStatusText = "实时采集中...";
CameraStatusText = "Live...";
try { _camera.ExecuteSoftwareTrigger(); }
catch (Exception ex) { _logger.Error(ex, "Live view trigger failed"); }
@@ -323,7 +312,7 @@ namespace XplorePlane.ViewModels
{
_liveViewRunning = false;
if (IsCameraGrabbing)
CameraStatusText = "采集中...";
CameraStatusText = "Grabbing...";
}
private void RefreshCameraParams()
@@ -334,13 +323,16 @@ namespace XplorePlane.ViewModels
GainValue = _camera.GetGain();
ImageWidth = _camera.GetWidth();
ImageHeight = _camera.GetHeight();
SelectedPixelFormat = _camera.GetPixelFormat();
var currentFormat = _camera.GetPixelFormat();
SelectedPixelFormat = currentFormat;
_logger.Information("Camera parameters refreshed");
}
catch (Exception ex)
{
_logger.Error(ex, "Failed to read camera parameters");
CameraStatusText = $"读取参数失败: {ex.Message}";
CameraStatusText = $"Read params failed: {ex.Message}";
}
}
@@ -354,7 +346,7 @@ namespace XplorePlane.ViewModels
catch (Exception ex)
{
_logger.Error(ex, "Failed to apply camera parameter");
CameraStatusText = $"设置参数失败: {ex.Message}";
CameraStatusText = $"Set param failed: {ex.Message}";
}
}
@@ -385,17 +377,21 @@ namespace XplorePlane.ViewModels
if (!_disposed)
CameraImageSource = bitmap;
});
if (_liveViewRunning)
{
_camera.ExecuteSoftwareTrigger();
}
}
catch (Exception ex)
{
if (!_disposed)
_logger.Error(ex, "Failed to process camera image");
}
finally
{
// 无论图像处理是否成功,都继续触发下一帧,保持采集链不断
if (_liveViewRunning && !_disposed)
{
try { _camera.ExecuteSoftwareTrigger(); }
catch { /* 忽略触发失败 */ }
}
}
}
private void OnCameraGrabError(object? sender, GrabErrorEventArgs e)
@@ -407,7 +403,7 @@ namespace XplorePlane.ViewModels
app.Dispatcher.BeginInvoke(() =>
{
if (!_disposed)
CameraStatusText = $"采集错误: {e.ErrorDescription}";
CameraStatusText = $"Grab error: {e.ErrorDescription}";
});
}
@@ -422,7 +418,7 @@ namespace XplorePlane.ViewModels
if (_disposed) return;
IsCameraConnected = false;
IsCameraGrabbing = false;
CameraStatusText = "连接已断开";
CameraStatusText = "Connection lost";
CameraImageSource = null;
SyncCameraStateToAppState();
});
@@ -0,0 +1,128 @@
<Window
x:Class="XplorePlane.Views.ImageProcessing.EdgeCircleFitPanel"
xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation"
xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml"
Title="边缘查找拟合圆" Width="300" Height="600"
ResizeMode="CanResize" WindowStartupLocation="CenterOwner"
Topmost="True" ShowInTaskbar="False"
Background="#F5F5F5" FontFamily="Microsoft YaHei UI">
<Window.Resources>
<Style x:Key="IconBtnStyle" TargetType="ButtonBase">
<Setter Property="Cursor" Value="Hand" />
<Setter Property="Template">
<Setter.Value>
<ControlTemplate TargetType="ButtonBase">
<Border x:Name="Bd" Background="#FFFFFF" BorderBrush="#E0E0E0"
BorderThickness="1" CornerRadius="6" Padding="8,6">
<ContentPresenter HorizontalAlignment="Center" VerticalAlignment="Center" />
</Border>
<ControlTemplate.Triggers>
<Trigger Property="IsMouseOver" Value="True">
<Setter TargetName="Bd" Property="Background" Value="#EAF2FB" />
<Setter TargetName="Bd" Property="BorderBrush" Value="#B0D4F1" />
</Trigger>
</ControlTemplate.Triggers>
</ControlTemplate>
</Setter.Value>
</Setter>
</Style>
<Style x:Key="CardStyle" TargetType="Border">
<Setter Property="Background" Value="White" />
<Setter Property="BorderBrush" Value="#E8E8E8" />
<Setter Property="BorderThickness" Value="1" />
<Setter Property="CornerRadius" Value="8" />
<Setter Property="Padding" Value="12,10" />
<Setter Property="Margin" Value="0,0,0,8" />
</Style>
<Style x:Key="ParamLabel" TargetType="TextBlock">
<Setter Property="FontSize" Value="11" />
<Setter Property="Foreground" Value="#555" />
<Setter Property="Margin" Value="0,0,0,3" />
</Style>
</Window.Resources>
<Grid Margin="10">
<ScrollViewer VerticalScrollBarVisibility="Auto">
<StackPanel>
<!-- 工具栏 -->
<StackPanel Orientation="Horizontal" Margin="0,0,0,10">
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding DrawCircleCommand}" ToolTip="画圆(3点)" Margin="0,0,6,0">
<TextBlock Text="⊙ 画圆" FontSize="11" />
</Button>
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding FitCommand}" ToolTip="执行拟合" Margin="0,0,6,0">
<TextBlock Text="▶ 拟合" FontSize="11" FontWeight="SemiBold" Foreground="#005FB8" />
</Button>
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding ClearAllCommand}" ToolTip="清除全部">
<TextBlock Text="✕ 清除全部" FontSize="11" />
</Button>
</StackPanel>
<!-- 卡尺参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="卡尺参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="卡尺数量" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,6">
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding CaliperCount, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="3" Maximum="180" Value="{Binding CaliperCount}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
<TextBlock Text="卡尺宽度 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding CaliperWidth, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="10" Maximum="200" Value="{Binding CaliperWidth}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
</StackPanel>
</Border>
<!-- 边缘检测参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="边缘检测参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="边缘极性" Style="{StaticResource ParamLabel}" />
<ComboBox SelectedValue="{Binding EdgePolarity}" Margin="0,0,0,6">
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">Both</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">BrightToDark</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">DarkToBright</sys:String>
</ComboBox>
<TextBlock Text="搜索方向" Style="{StaticResource ParamLabel}" />
<ComboBox SelectedValue="{Binding SearchDirection}" Margin="0,0,0,6">
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">Both</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">Inward</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">Outward</sys:String>
</ComboBox>
<TextBlock Text="边缘阈值" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,6">
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding EdgeThreshold, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="1" Maximum="200" Value="{Binding EdgeThreshold}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
<TextBlock Text="平滑 Sigma" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding Sigma, StringFormat=F1, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="0.1" Maximum="5.0" Value="{Binding Sigma}" SmallChange="0.1" LargeChange="0.5" VerticalAlignment="Center" />
</DockPanel>
</StackPanel>
</Border>
<!-- 拟合参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="拟合参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="RANSAC 阈值 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding RansacThreshold, StringFormat=F1, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="0.5" Maximum="20" Value="{Binding RansacThreshold}" SmallChange="0.5" LargeChange="2" VerticalAlignment="Center" />
</DockPanel>
</StackPanel>
</Border>
<!-- 结果 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="结果" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,6" />
<TextBlock Text="{Binding ResultText}" FontSize="11.5" Foreground="#333" TextWrapping="Wrap" />
</StackPanel>
</Border>
</StackPanel>
</ScrollViewer>
</Grid>
</Window>
@@ -0,0 +1,52 @@
using System.Windows;
using Prism.Ioc;
using XP.ImageProcessing.RoiControl.Controls;
using XplorePlane.Services.MainViewport;
using XplorePlane.ViewModels.ImageProcessing;
namespace XplorePlane.Views.ImageProcessing
{
public partial class EdgeCircleFitPanel : Window
{
public EdgeCircleFitPanel()
{
InitializeComponent();
var viewportService = ContainerLocator.Current?.Resolve<IMainViewportService>();
DataContext = new EdgeCircleFitViewModel(viewportService);
Loaded += (s, e) =>
{
var mainWin = Owner as MainWindow;
if (mainWin != null)
{
var canvas = FindChild<PolygonRoiCanvas>(mainWin);
if (DataContext is EdgeCircleFitViewModel vm)
{
vm.SetCanvas(canvas);
vm.DrawCircleCommand.Execute();
}
}
};
Closed += (s, e) =>
{
if (DataContext is EdgeCircleFitViewModel vm)
vm.OnPanelClosed();
};
}
private static T FindChild<T>(DependencyObject parent) where T : DependencyObject
{
int count = System.Windows.Media.VisualTreeHelper.GetChildrenCount(parent);
for (int i = 0; i < count; i++)
{
var child = System.Windows.Media.VisualTreeHelper.GetChild(parent, i);
if (child is T t) return t;
var result = FindChild<T>(child);
if (result != null) return result;
}
return null;
}
}
}
@@ -0,0 +1,122 @@
<Window
x:Class="XplorePlane.Views.ImageProcessing.EdgeLineFitPanel"
xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation"
xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml"
Title="边缘查找拟合直线" Width="300" Height="560"
ResizeMode="CanResize" WindowStartupLocation="CenterOwner"
Topmost="True" ShowInTaskbar="False"
Background="#F5F5F5" FontFamily="Microsoft YaHei UI">
<Window.Resources>
<Style x:Key="IconBtnStyle" TargetType="ButtonBase">
<Setter Property="Cursor" Value="Hand" />
<Setter Property="Template">
<Setter.Value>
<ControlTemplate TargetType="ButtonBase">
<Border x:Name="Bd" Background="#FFFFFF" BorderBrush="#E0E0E0"
BorderThickness="1" CornerRadius="6" Padding="8,6">
<ContentPresenter HorizontalAlignment="Center" VerticalAlignment="Center" />
</Border>
<ControlTemplate.Triggers>
<Trigger Property="IsMouseOver" Value="True">
<Setter TargetName="Bd" Property="Background" Value="#EAF2FB" />
<Setter TargetName="Bd" Property="BorderBrush" Value="#B0D4F1" />
</Trigger>
</ControlTemplate.Triggers>
</ControlTemplate>
</Setter.Value>
</Setter>
</Style>
<Style x:Key="CardStyle" TargetType="Border">
<Setter Property="Background" Value="White" />
<Setter Property="BorderBrush" Value="#E8E8E8" />
<Setter Property="BorderThickness" Value="1" />
<Setter Property="CornerRadius" Value="8" />
<Setter Property="Padding" Value="12,10" />
<Setter Property="Margin" Value="0,0,0,8" />
</Style>
<Style x:Key="ParamLabel" TargetType="TextBlock">
<Setter Property="FontSize" Value="11" />
<Setter Property="Foreground" Value="#555" />
<Setter Property="Margin" Value="0,0,0,3" />
</Style>
</Window.Resources>
<Grid Margin="10">
<ScrollViewer VerticalScrollBarVisibility="Auto">
<StackPanel>
<!-- 工具栏 -->
<StackPanel Orientation="Horizontal" Margin="0,0,0,10">
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding DrawCaliperCommand}" ToolTip="绘制卡尺" Margin="0,0,6,0">
<TextBlock Text="✏ 画卡尺" FontSize="11" />
</Button>
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding FitCommand}" ToolTip="执行拟合" Margin="0,0,6,0">
<TextBlock Text="▶ 拟合" FontSize="11" FontWeight="SemiBold" Foreground="#005FB8" />
</Button>
<Button Style="{StaticResource IconBtnStyle}" Command="{Binding ClearAllCommand}" ToolTip="清除全部">
<TextBlock Text="✕ 清除全部" FontSize="11" />
</Button>
</StackPanel>
<!-- 卡尺参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="卡尺参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="卡尺数量" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,6">
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding CaliperCount, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="3" Maximum="100" Value="{Binding CaliperCount}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
<TextBlock Text="卡尺宽度 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding DisplayWidth, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="10" Maximum="300" Value="{Binding DisplayWidth}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
</StackPanel>
</Border>
<!-- 边缘检测参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="边缘检测参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="边缘极性" Style="{StaticResource ParamLabel}" />
<ComboBox SelectedValue="{Binding EdgePolarity}" Margin="0,0,0,6">
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">Both</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">BrightToDark</sys:String>
<sys:String xmlns:sys="clr-namespace:System;assembly=mscorlib">DarkToBright</sys:String>
</ComboBox>
<TextBlock Text="边缘阈值" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,6">
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding EdgeThreshold, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="1" Maximum="200" Value="{Binding EdgeThreshold}" VerticalAlignment="Center" SmallChange="1" IsSnapToTickEnabled="True" TickFrequency="1" />
</DockPanel>
<TextBlock Text="平滑 Sigma" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,6">
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding Sigma, StringFormat=F1, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="0.1" Maximum="5.0" Value="{Binding Sigma}" SmallChange="0.1" LargeChange="0.5" VerticalAlignment="Center" />
</DockPanel>
</StackPanel>
</Border>
<!-- 拟合参数 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="拟合参数" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,8" />
<TextBlock Text="RANSAC 阈值 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="50" Text="{Binding RansacThreshold, StringFormat=F1, UpdateSourceTrigger=PropertyChanged}" VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="0.5" Maximum="20" Value="{Binding RansacThreshold}" SmallChange="0.5" LargeChange="2" VerticalAlignment="Center" />
</DockPanel>
</StackPanel>
</Border>
<!-- 结果 -->
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="结果" FontWeight="SemiBold" FontSize="12" Margin="0,0,0,6" />
<TextBlock Text="{Binding ResultText}" FontSize="11.5" Foreground="#333" TextWrapping="Wrap" />
</StackPanel>
</Border>
</StackPanel>
</ScrollViewer>
</Grid>
</Window>
@@ -0,0 +1,53 @@
using System.Windows;
using Prism.Ioc;
using XP.ImageProcessing.RoiControl.Controls;
using XplorePlane.Services.MainViewport;
using XplorePlane.ViewModels.ImageProcessing;
namespace XplorePlane.Views.ImageProcessing
{
public partial class EdgeLineFitPanel : Window
{
public EdgeLineFitPanel()
{
InitializeComponent();
var viewportService = ContainerLocator.Current?.Resolve<IMainViewportService>();
DataContext = new EdgeLineFitViewModel(viewportService);
Loaded += (s, e) =>
{
var mainWin = Owner as MainWindow;
if (mainWin != null)
{
var canvas = FindChild<PolygonRoiCanvas>(mainWin);
if (DataContext is EdgeLineFitViewModel vm)
{
vm.SetCanvas(canvas);
// 自动进入绘制模式
vm.DrawCaliperCommand.Execute();
}
}
};
Closed += (s, e) =>
{
if (DataContext is EdgeLineFitViewModel vm)
vm.OnPanelClosed();
};
}
private static T FindChild<T>(DependencyObject parent) where T : DependencyObject
{
int count = System.Windows.Media.VisualTreeHelper.GetChildrenCount(parent);
for (int i = 0; i < count; i++)
{
var child = System.Windows.Media.VisualTreeHelper.GetChild(parent, i);
if (child is T t) return t;
var result = FindChild<T>(child);
if (result != null) return result;
}
return null;
}
}
}
@@ -0,0 +1,345 @@
<Window x:Class="XplorePlane.Views.ImageProcessing.TemplateMatchAssistantWindow"
xmlns="http://schemas.microsoft.com/winfx/2006/xaml/presentation"
xmlns:x="http://schemas.microsoft.com/winfx/2006/xaml"
Title="模板匹配助手"
Height="560"
Width="560"
FontFamily="Microsoft YaHei UI"
WindowStartupLocation="CenterOwner"
ResizeMode="CanResizeWithGrip">
<Window.Resources>
<Style x:Key="TmMdl2Glyph" TargetType="TextBlock">
<Setter Property="FontFamily" Value="Segoe MDL2 Assets" />
<Setter Property="FontSize" Value="15" />
<Setter Property="Margin" Value="0,0,6,0" />
<Setter Property="VerticalAlignment" Value="Center" />
</Style>
</Window.Resources>
<Grid Margin="12">
<Grid.RowDefinitions>
<RowDefinition Height="Auto" />
<RowDefinition Height="*" />
</Grid.RowDefinitions>
<TextBlock
Grid.Row="0"
Text="{Binding StatusMessage}"
TextWrapping="Wrap"
Margin="0,0,0,10"
FontSize="12" />
<TabControl Grid.Row="1">
<TabItem Header="单张与参数">
<Grid Margin="0,8,0,0">
<Grid.RowDefinitions>
<RowDefinition Height="Auto" />
<RowDefinition Height="*" />
<RowDefinition Height="Auto" />
</Grid.RowDefinitions>
<StackPanel
Grid.Row="0"
Orientation="Horizontal"
Margin="0,0,0,8">
<Button
Margin="0,0,6,0"
Padding="8,4"
Command="{Binding SelectTemplateRoiCommand}"
ToolTip="在主视图上拖拽框选模板区域">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE7B8;" />
<TextBlock VerticalAlignment="Center" Text="框选模板 ROI" />
</StackPanel>
</Button>
<Button
Margin="0,0,6,0"
Padding="8,4"
Command="{Binding LearnFromRoiCommand}"
ToolTip="根据已框选的 ROI 学习模板">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE8AF;" />
<TextBlock VerticalAlignment="Center" Text="从 ROI 训练模板" />
</StackPanel>
</Button>
<Button
Margin="0,0,6,0"
Padding="8,4"
Command="{Binding LoadModelCommand}"
ToolTip="从磁盘加载已保存的模板模型">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE8E5;" />
<TextBlock VerticalAlignment="Center" Text="加载模型" />
</StackPanel>
</Button>
<Button
Padding="8,4"
Command="{Binding SaveModelCommand}"
ToolTip="将当前模板保存为模型文件">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE74E;" />
<TextBlock VerticalAlignment="Center" Text="保存模型" />
</StackPanel>
</Button>
</StackPanel>
<ScrollViewer Grid.Row="1" VerticalScrollBarVisibility="Auto">
<StackPanel>
<Grid Margin="0,4">
<Grid.ColumnDefinitions>
<ColumnDefinition Width="132" />
<ColumnDefinition Width="*" />
<ColumnDefinition Width="52" />
</Grid.ColumnDefinitions>
<Grid.RowDefinitions>
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
</Grid.RowDefinitions>
<TextBlock Grid.Row="0" Grid.Column="0" VerticalAlignment="Center" Text="匹配阈值" />
<Slider
Grid.Row="0"
Grid.Column="1"
Margin="6,6,6,6"
Minimum="0"
Maximum="1"
SmallChange="0.01"
LargeChange="0.05"
TickFrequency="0.05"
IsSnapToTickEnabled="False"
AutoToolTipPlacement="TopLeft"
VerticalAlignment="Center"
Value="{Binding MatchThreshold, Mode=TwoWay, UpdateSourceTrigger=PropertyChanged}" />
<TextBlock
Grid.Row="0"
Grid.Column="2"
VerticalAlignment="Center"
TextAlignment="Right"
Text="{Binding MatchThreshold, StringFormat={}{0:F2}}" />
<TextBlock Grid.Row="1" Grid.Column="0" VerticalAlignment="Center" Text="角度容差 (°)" />
<Slider
Grid.Row="1"
Grid.Column="1"
Margin="6,6,6,6"
Minimum="0"
Maximum="180"
SmallChange="1"
LargeChange="5"
TickFrequency="5"
IsSnapToTickEnabled="True"
AutoToolTipPlacement="TopLeft"
VerticalAlignment="Center"
Value="{Binding ToleranceAngle, Mode=TwoWay, UpdateSourceTrigger=PropertyChanged}" />
<TextBlock
Grid.Row="1"
Grid.Column="2"
VerticalAlignment="Center"
TextAlignment="Right"
Text="{Binding ToleranceAngle, StringFormat={}{0:F0}}" />
<TextBlock Grid.Row="2" Grid.Column="0" VerticalAlignment="Center" Text="最大匹配数" />
<Slider
Grid.Row="2"
Grid.Column="1"
Margin="6,6,6,6"
Minimum="1"
Maximum="100"
SmallChange="1"
LargeChange="5"
TickFrequency="1"
IsSnapToTickEnabled="True"
AutoToolTipPlacement="TopLeft"
VerticalAlignment="Center"
Value="{Binding MaxMatchCount, Mode=TwoWay, UpdateSourceTrigger=PropertyChanged}" />
<TextBlock
Grid.Row="2"
Grid.Column="2"
VerticalAlignment="Center"
TextAlignment="Right"
Text="{Binding MaxMatchCount, StringFormat={}{0:F0}}" />
<TextBlock Grid.Row="3" Grid.Column="0" VerticalAlignment="Center" Text="最大重叠" />
<Slider
Grid.Row="3"
Grid.Column="1"
Margin="6,6,6,6"
Minimum="0"
Maximum="1"
SmallChange="0.05"
LargeChange="0.1"
TickFrequency="0.05"
IsSnapToTickEnabled="True"
AutoToolTipPlacement="TopLeft"
VerticalAlignment="Center"
Value="{Binding MaxOverlap, Mode=TwoWay, UpdateSourceTrigger=PropertyChanged}" />
<TextBlock
Grid.Row="3"
Grid.Column="2"
VerticalAlignment="Center"
TextAlignment="Right"
Text="{Binding MaxOverlap, StringFormat={}{0:F2}}" />
<TextBlock Grid.Row="4" Grid.Column="0" VerticalAlignment="Center" TextWrapping="Wrap" Text="金字塔最小面积" />
<Slider
Grid.Row="4"
Grid.Column="1"
Margin="6,6,6,6"
Minimum="64"
Maximum="4096"
SmallChange="32"
LargeChange="128"
TickFrequency="32"
IsSnapToTickEnabled="True"
AutoToolTipPlacement="TopLeft"
VerticalAlignment="Center"
Value="{Binding MinReduceArea, Mode=TwoWay, UpdateSourceTrigger=PropertyChanged}" />
<TextBlock
Grid.Row="4"
Grid.Column="2"
VerticalAlignment="Center"
TextAlignment="Right"
Text="{Binding MinReduceArea, StringFormat={}{0:F0}}" />
<CheckBox
Grid.Row="5"
Grid.Column="0"
Grid.ColumnSpan="3"
Margin="0,8,0,0"
Content="使用 SIMD"
IsChecked="{Binding UseSimd}" />
<CheckBox
Grid.Row="6"
Grid.Column="0"
Grid.ColumnSpan="3"
Margin="0,4,0,0"
Content="亚像素"
IsChecked="{Binding UseSubPixel}" />
</Grid>
</StackPanel>
</ScrollViewer>
<StackPanel
Grid.Row="2"
Orientation="Horizontal"
HorizontalAlignment="Right"
Margin="0,10,0,0">
<Button
Padding="10,4"
Command="{Binding RunMatchCommand}"
ToolTip="对主视图当前图像执行模板匹配">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE768;" />
<TextBlock VerticalAlignment="Center" Text="运行匹配" />
</StackPanel>
</Button>
</StackPanel>
</Grid>
</TabItem>
<TabItem Header="批量测试">
<Grid Margin="0,8,0,0" DataContext="{Binding Batch}">
<Grid.RowDefinitions>
<RowDefinition Height="Auto" />
<RowDefinition Height="Auto" />
<RowDefinition Height="*" />
<RowDefinition Height="Auto" />
</Grid.RowDefinitions>
<TextBlock
Grid.Row="0"
Text="{Binding BatchStatusText}"
TextWrapping="Wrap"
Margin="0,0,0,8"
FontSize="12" />
<Grid Grid.Row="1" Margin="0,0,0,8">
<Grid.ColumnDefinitions>
<ColumnDefinition Width="*" />
<ColumnDefinition Width="Auto" />
</Grid.ColumnDefinitions>
<StackPanel Grid.Column="0" Orientation="Horizontal">
<Button
Padding="10,4"
Margin="0,0,8,0"
Command="{Binding PickFolderCommand}"
ToolTip="选择包含待测图像的文件夹(仅当前层级)">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE838;" />
<TextBlock VerticalAlignment="Center" Text="选择文件夹…" />
</StackPanel>
</Button>
<Button
Padding="10,4"
Margin="0,0,8,0"
Command="{Binding StartBatchCommand}"
ToolTip="对文件夹内图像依次执行模板匹配">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE768;" />
<TextBlock VerticalAlignment="Center" Text="开始批量匹配" />
</StackPanel>
</Button>
<Button
Padding="10,4"
Margin="0,0,8,0"
Command="{Binding StopBatchCommand}"
ToolTip="停止当前批量任务">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE71A;" />
<TextBlock VerticalAlignment="Center" Text="停止" />
</StackPanel>
</Button>
<Button
Padding="10,4"
Command="{Binding OpenSelectedInMainViewportCommand}"
ToolTip="将表格当前选中行对应图像在主视图中打开并显示匹配结果">
<StackPanel Orientation="Horizontal">
<TextBlock Style="{StaticResource TmMdl2Glyph}" Text="&#xE944;" />
<TextBlock VerticalAlignment="Center" Text="在主视图打开所选" />
</StackPanel>
</Button>
</StackPanel>
<StackPanel Grid.Column="1" VerticalAlignment="Center" Margin="8,0,0,0">
<TextBlock Foreground="Gray" FontSize="11" Text="{Binding ImageFileCount, StringFormat={}{0} 个图像文件}" />
<TextBlock
Foreground="Gray"
FontSize="11"
Text="{Binding FolderPath}"
TextTrimming="CharacterEllipsis"
MaxWidth="220" />
</StackPanel>
</Grid>
<DataGrid
x:Name="BatchResultGrid"
Grid.Row="2"
AutoGenerateColumns="False"
IsReadOnly="True"
CanUserAddRows="False"
CanUserDeleteRows="False"
SelectionMode="Single"
ItemsSource="{Binding Rows}"
SelectedItem="{Binding SelectedRow, Mode=TwoWay}"
MouseDoubleClick="BatchResultGrid_OnMouseDoubleClick">
<DataGrid.Columns>
<DataGridTextColumn Header="文件名" Binding="{Binding FileName}" Width="*" MinWidth="120" />
<DataGridTextColumn Header="结果" Binding="{Binding Result}" Width="72" />
<DataGridTextColumn Header="匹配数" Binding="{Binding MatchCount}" Width="64" />
<DataGridTextColumn Header="最佳分数" Binding="{Binding BestScore, StringFormat=F3}" Width="80" />
<DataGridTextColumn Header="耗时(ms)" Binding="{Binding TimeMs, StringFormat=F1}" Width="80" />
<DataGridTextColumn Header="备注" Binding="{Binding ErrorDetail}" Width="2*" MinWidth="100" />
</DataGrid.Columns>
</DataGrid>
<TextBlock
Grid.Row="3"
Margin="0,8,0,0"
FontSize="11"
Foreground="Gray"
TextWrapping="Wrap"
Text="参数与「单张与参数」选项卡一致;仅扫描所选文件夹当前层级。双击一行可在主视图打开该图并显示匹配框。" />
</Grid>
</TabItem>
</TabControl>
</Grid>
</Window>
@@ -0,0 +1,28 @@
using System;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Input;
using XplorePlane.ViewModels.ImageProcessing;
namespace XplorePlane.Views.ImageProcessing;
public partial class TemplateMatchAssistantWindow : Window
{
public TemplateMatchAssistantWindow()
{
InitializeComponent();
}
private void BatchResultGrid_OnMouseDoubleClick(object sender, MouseButtonEventArgs e)
{
if (sender is DataGrid g && g.DataContext is TemplateMatchBatchViewModel vm)
vm.OpenSelectedFromDoubleClick();
}
protected override void OnClosed(EventArgs e)
{
if (DataContext is IDisposable d)
d.Dispose();
base.OnClosed(e);
}
}
+45 -1
View File
@@ -166,7 +166,7 @@
Size="Medium"
SmallImage="/Assets/Icons/crosshair.png"
Text="辅助线" />
<telerik:RadRibbonToggleButton
<telerik:RadRibbonToggleButton
telerik:ScreenTip.Description="显示/隐藏图像比例尺"
telerik:ScreenTip.Title="比例尺"
IsChecked="{Binding IsScaleBarVisible, Mode=TwoWay}"
@@ -537,9 +537,53 @@
Text="坐标标定" />
</StackPanel>
</telerik:RadRibbonGroup>
<<<<<<< HEAD
=======
<telerik:RadRibbonGroup
telerik:ScreenTip.Description="Show the Alignment tab of the Format Cells dialog box."
telerik:ScreenTip.Title="Format Cells: Alignment"
DialogLauncherCommand="{Binding Path=ShowFormatCellsDialog.Command}"
DialogLauncherCommandParameter="Alignment"
DialogLauncherVisibility="{Binding Path=ShowFormatCellsDialog.IsEnabled, Converter={StaticResource BoolToVisibilityValueConverter}}"
Header="识别定位"
IsEnabled="{Binding Path=AlignmentGroup.IsEnabled}">
<telerik:RadRibbonGroup.Variants>
<telerik:GroupVariant Priority="0" Variant="Large" />
</telerik:RadRibbonGroup.Variants>
<telerik:RadRibbonGroup.Resources>
<spreadsheetControls:RadHorizontalAlignmentToBooleanConverter x:Key="horizontalAlignmentToBooleanConverter" />
<spreadsheetControls:RadVerticalAlignmentToBooleanConverter x:Key="verticalAlignmentToBooleanConverter" />
</telerik:RadRibbonGroup.Resources>
<StackPanel />
<StackPanel>
<telerik:RadRibbonButton
telerik:ScreenTip.Description="框选模板、调参并在当前图像上试跑旋转模板匹配"
telerik:ScreenTip.Title="模板匹配助手"
Command="{Binding OpenTemplateMatchAssistantCommand}"
Size="Medium"
SmallImage="/Assets/Icons/Matching.png"
Text="模板助手" />
<telerik:RadRibbonButton
telerik:ScreenTip.Description="沿搜索线放置卡尺检测边缘点并拟合直线"
telerik:ScreenTip.Title="拟合直线"
Command="{Binding EdgeLineFitCommand}"
Size="Medium"
SmallImage="/Assets/Icons/FittedLine.png"
Text="拟合直线" />
<telerik:RadRibbonButton
telerik:ScreenTip.Description="沿搜索线放置卡尺检测边缘点并拟合圆"
telerik:ScreenTip.Title="拟合圆"
Command="{Binding EdgeCircleFitCommand}"
Size="Medium"
SmallImage="/Assets/Icons/FittedCircle.png"
Text="拟合圆" />
</StackPanel>
</telerik:RadRibbonGroup>
>>>>>>> TURBO-615-RecognitionAndPositioning
<telerik:RadRibbonGroup Header="多语言">
<telerik:RadRibbonGroup.Variants>
<telerik:GroupVariant Priority="0" Variant="Large" />
@@ -1,6 +1,7 @@
using System;
using System.ComponentModel;
using System.IO;
using System.Linq;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Media;
@@ -75,6 +76,7 @@ namespace XplorePlane.Views
"FillRate" => "填锡率",
"BgaVoid" => "BGA空隙",
"BubbleVoid" => "气泡空隙",
"EdgeLineFit" => "直线拟合",
_ => "点点距"
};
string valueText = args.MeasureType switch
@@ -83,9 +85,16 @@ namespace XplorePlane.Views
"FillRate" => $"{args.Distance:F1}%",
"BgaVoid" => $"{args.Distance:F1}%",
"BubbleVoid" => $"{args.Distance:F1}%",
"EdgeLineFit" => "处理中...",
_ => $"{args.Distance:F2} px"
};
SetStatus($"{typeLabel}: {valueText} | 共 {args.TotalCount} 条测量");
// 边缘查找拟合直线:获取搜索线后执行算子
if (args.MeasureType == "EdgeLineFit")
{
ExecuteEdgeLineFitProcessor(args.P1, args.P2);
}
}
};
RoiCanvas.MeasureStatusChanged += (s, e) =>
@@ -114,6 +123,7 @@ namespace XplorePlane.Views
MeasurementToolMode.ThroughHoleFillRate => XP.ImageProcessing.RoiControl.Models.MeasureMode.FillRate,
MeasurementToolMode.BgaVoid => XP.ImageProcessing.RoiControl.Models.MeasureMode.BgaVoid,
MeasurementToolMode.BubbleMeasure => XP.ImageProcessing.RoiControl.Models.MeasureMode.BubbleMeasure,
MeasurementToolMode.EdgeLineFit => XP.ImageProcessing.RoiControl.Models.MeasureMode.EdgeLineFit,
_ => XP.ImageProcessing.RoiControl.Models.MeasureMode.None
};
}, Prism.Events.ThreadOption.UIThread);
@@ -134,8 +144,224 @@ namespace XplorePlane.Views
var vm = GetMainVm();
if (vm != null) vm.CursorInfoText = RoiCanvas.CursorInfo;
});
// 行灰度分布
try
{
var ea2 = ContainerLocator.Current?.Resolve<Prism.Events.IEventAggregator>();
ea2?.GetEvent<ToggleLineProfileEvent>().Subscribe(() =>
{
ToggleLineProfile();
}, Prism.Events.ThreadOption.UIThread);
// 白底检测:进入ROI绘制模式
ea2?.GetEvent<WhiteBackgroundDetectionEvent>().Subscribe(() =>
{
_bgDefectDrawing = false;
_bgDefectRoiMode = BackgroundDefectRoiMode.WhiteBackground;
RegisterBackgroundDefectRoiMouseHandlers();
}, Prism.Events.ThreadOption.UIThread);
// 黑底检测:进入ROI绘制模式
ea2?.GetEvent<BlackBackgroundDetectionEvent>().Subscribe(() =>
{
_bgDefectDrawing = false;
_bgDefectRoiMode = BackgroundDefectRoiMode.BlackBackground;
RegisterBackgroundDefectRoiMouseHandlers();
}, Prism.Events.ThreadOption.UIThread);
// 白底检测:渲染结果(红色标识)
ea2?.GetEvent<WhiteBackgroundResultEvent>().Subscribe(payload =>
{
RenderBackgroundDefectResult(payload.RoiRect, payload.Detections, isBlackBackground: false);
}, Prism.Events.ThreadOption.UIThread);
// 黑底检测:渲染结果(绿色标识)
ea2?.GetEvent<BlackBackgroundResultEvent>().Subscribe(payload =>
{
RenderBackgroundDefectResult(payload.RoiRect, payload.Detections, isBlackBackground: true);
}, Prism.Events.ThreadOption.UIThread);
ea2?.GetEvent<TemplateMatchEnterRoiModeEvent>().Subscribe(() =>
{
_bgDefectDrawing = false;
_bgDefectRoiMode = BackgroundDefectRoiMode.TemplateAssistant;
RegisterBackgroundDefectRoiMouseHandlers();
SetStatus("模板助手:请在图像上拖拽框选模板区域");
}, Prism.Events.ThreadOption.UIThread);
ea2?.GetEvent<TemplateMatchPreviewResultEvent>().Subscribe(payload =>
{
RenderTemplateMatchPreview(payload);
}, Prism.Events.ThreadOption.UIThread);
ea2?.GetEvent<TemplateMatchClearRoiOverlayEvent>().Subscribe(() =>
{
RemoveTemplateAssistantPersistRoi();
}, Prism.Events.ThreadOption.UIThread);
}
catch { }
}
#region
private bool _lineProfileEnabled;
private System.Windows.Shapes.Line _profileRefLine; // 透明命中区域
private System.Windows.Shapes.Line _profileRefLineVisible; // 1px红线显示
private System.Windows.Shapes.Polyline _profileCurve;
private double _profileLineY;
private bool _profileDragging;
private void ToggleLineProfile()
{
_lineProfileEnabled = !_lineProfileEnabled;
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
if (_lineProfileEnabled)
{
// 参考线默认在图像中间
_profileLineY = RoiCanvas.CanvasHeight / 2;
// 创建参考线(红色水平线,可拖动)
// 用透明粗线作为命中区域,叠加1px红线显示
_profileRefLine = new System.Windows.Shapes.Line
{
X1 = 0,
Y1 = _profileLineY,
X2 = RoiCanvas.CanvasWidth,
Y2 = _profileLineY,
Stroke = System.Windows.Media.Brushes.Transparent,
StrokeThickness = 7, // 上下3px命中区域
IsHitTestVisible = true,
Cursor = System.Windows.Input.Cursors.SizeNS
};
_profileRefLineVisible = new System.Windows.Shapes.Line
{
X1 = 0,
Y1 = _profileLineY,
X2 = RoiCanvas.CanvasWidth,
Y2 = _profileLineY,
Stroke = System.Windows.Media.Brushes.Red,
StrokeThickness = 1,
IsHitTestVisible = false
};
_profileRefLine.MouseLeftButtonDown += ProfileLine_MouseDown;
_profileRefLine.MouseMove += ProfileLine_MouseMove;
_profileRefLine.MouseLeftButtonUp += ProfileLine_MouseUp;
canvas.Children.Add(_profileRefLineVisible);
canvas.Children.Add(_profileRefLine);
// 创建灰度折线(固定显示在图像中间位置)
_profileCurve = new System.Windows.Shapes.Polyline
{
Stroke = System.Windows.Media.Brushes.Red,
StrokeThickness = 1,
IsHitTestVisible = false
};
canvas.Children.Add(_profileCurve);
UpdateLineProfile();
SetStatus("行灰度分布:拖动红线改变采样行,再次点击按钮关闭");
}
else
{
if (_profileRefLine != null)
{
canvas.Children.Remove(_profileRefLine);
_profileRefLine = null;
}
if (_profileRefLineVisible != null)
{
canvas.Children.Remove(_profileRefLineVisible);
_profileRefLineVisible = null;
}
if (_profileCurve != null)
{
canvas.Children.Remove(_profileCurve);
_profileCurve = null;
}
SetStatus("行灰度分布已关闭");
}
}
private void ProfileLine_MouseDown(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
_profileDragging = true;
_profileRefLine?.CaptureMouse();
e.Handled = true;
}
private void ProfileLine_MouseMove(object sender, System.Windows.Input.MouseEventArgs e)
{
if (!_profileDragging || _profileRefLine == null) return;
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
var pos = e.GetPosition(canvas);
_profileLineY = Math.Clamp(pos.Y, 0, RoiCanvas.CanvasHeight - 1);
_profileRefLine.Y1 = _profileLineY;
_profileRefLine.Y2 = _profileLineY;
_profileRefLineVisible.Y1 = _profileLineY;
_profileRefLineVisible.Y2 = _profileLineY;
UpdateLineProfile();
}
private void ProfileLine_MouseUp(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
_profileDragging = false;
_profileRefLine?.ReleaseMouseCapture();
e.Handled = true;
}
private void UpdateLineProfile()
{
if (_profileCurve == null) return;
// 从当前显示图像获取像素数据
var viewportVm = DataContext as ViewportPanelViewModel;
var imageSource = viewportVm?.ImageSource as System.Windows.Media.Imaging.BitmapSource;
if (imageSource == null) return;
int imgWidth = imageSource.PixelWidth;
int imgHeight = imageSource.PixelHeight;
int row = (int)Math.Clamp(_profileLineY, 0, imgHeight - 1);
// 转为 Gray8 获取行像素
System.Windows.Media.Imaging.BitmapSource gray8;
if (imageSource.Format != System.Windows.Media.PixelFormats.Gray8)
gray8 = new System.Windows.Media.Imaging.FormatConvertedBitmap(
imageSource, System.Windows.Media.PixelFormats.Gray8, null, 0);
else
gray8 = imageSource;
byte[] rowPixels = new byte[imgWidth];
int stride = imgWidth;
gray8.CopyPixels(new System.Windows.Int32Rect(0, row, imgWidth, 1), rowPixels, stride, 0);
// 构建折线点集:折线固定显示在图像垂直中间位置
// 参考线位置决定采样哪一行,折线位置固定在画布中间
double canvasH = RoiCanvas.CanvasHeight;
double curveCenter = canvasH / 2.0; // 折线基线固定在图像中间
double displayHeight = canvasH * 0.25; // 折线振幅为画布高度的25%
var points = new System.Windows.Media.PointCollection(imgWidth);
for (int x = 0; x < imgWidth; x++)
{
double normalizedGray = rowPixels[x] / 255.0;
double y = curveCenter - normalizedGray * displayHeight;
points.Add(new System.Windows.Point(x, y));
}
_profileCurve.Points = points;
SetStatus($"行灰度分布 | Y={row} | 均值={rowPixels.Select(b => (double)b).Average():F1} | 最大={rowPixels.Max()} | 最小={rowPixels.Min()}");
}
#endregion
private void OnDataContextChanged(object sender, DependencyPropertyChangedEventArgs e)
{
if (e.OldValue is INotifyPropertyChanged oldVm)
@@ -162,7 +388,31 @@ namespace XplorePlane.Views
RoiCanvas.SelectedROI = null;
if (DataContext is ViewportPanelViewModel vm)
vm.ResetMeasurementState();
SetStatus("已清除所有测量");
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas != null)
{
if (_bgDefectPreview != null)
{
canvas.Children.Remove(_bgDefectPreview);
_bgDefectPreview = null;
}
ClearBackgroundDefectOverlays(canvas);
ClearTemplateMatchOverlays(canvas);
RemoveTemplateAssistantPersistRoi();
}
else
{
_bgDefectOverlays.Clear();
_tmMatchOverlays.Clear();
RemoveTemplateAssistantPersistRoi();
}
_bgDefectDrawing = false;
_bgDefectRoiMode = BackgroundDefectRoiMode.None;
try { RoiCanvas.ReleaseMouseCapture(); } catch { /* 未捕获则无影响 */ }
SetStatus("已清除所有测量、白底/黑底检测、模板匹配试跑叠加及模板助手 ROI");
}
private void SaveOriginalImage_Click(object sender, RoutedEventArgs e)
@@ -215,6 +465,486 @@ namespace XplorePlane.Views
#endregion
#region /
private enum BackgroundDefectRoiMode
{
None,
WhiteBackground,
BlackBackground,
TemplateAssistant
}
private BackgroundDefectRoiMode _bgDefectRoiMode;
private bool _bgDefectDrawing;
private System.Windows.Point _bgDefectStart;
private System.Windows.Shapes.Rectangle _bgDefectPreview;
private readonly System.Collections.Generic.List<System.Windows.UIElement> _bgDefectOverlays = new();
private readonly System.Collections.Generic.List<System.Windows.UIElement> _tmMatchOverlays = new();
private System.Windows.Shapes.Rectangle _templateAssistantRoiPersist;
private bool _bgDefectMouseHandlersRegistered;
private void RegisterBackgroundDefectRoiMouseHandlers()
{
if (_bgDefectMouseHandlersRegistered) return;
RoiCanvas.PreviewMouseLeftButtonDown += OnMainCanvasPreviewMouseDown;
RoiCanvas.PreviewMouseMove += OnMainCanvasPreviewMouseMove;
RoiCanvas.PreviewMouseLeftButtonUp += OnMainCanvasPreviewMouseUp;
_bgDefectMouseHandlersRegistered = true;
}
// 需要在 mainCanvas 的 MouseDown/Move/Up 中处理
// 由于 PolygonRoiCanvas 内部已经处理了鼠标事件,我们通过 PreviewMouse 事件来拦截
private void OnMainCanvasPreviewMouseDown(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (_bgDefectRoiMode == BackgroundDefectRoiMode.None || e.LeftButton != System.Windows.Input.MouseButtonState.Pressed) return;
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
_bgDefectStart = e.GetPosition(canvas);
_bgDefectDrawing = true;
// 创建预览矩形(不清除之前的检测结果)
_bgDefectPreview = new System.Windows.Shapes.Rectangle
{
Stroke = System.Windows.Media.Brushes.Blue,
StrokeThickness = 1,
StrokeDashArray = new System.Windows.Media.DoubleCollection { 4, 2 }
};
System.Windows.Controls.Canvas.SetLeft(_bgDefectPreview, _bgDefectStart.X);
System.Windows.Controls.Canvas.SetTop(_bgDefectPreview, _bgDefectStart.Y);
canvas.Children.Add(_bgDefectPreview);
RoiCanvas.CaptureMouse();
e.Handled = true;
}
private void OnMainCanvasPreviewMouseMove(object sender, System.Windows.Input.MouseEventArgs e)
{
if (!_bgDefectDrawing || _bgDefectPreview == null) return;
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
var current = e.GetPosition(canvas);
double x = Math.Min(_bgDefectStart.X, current.X);
double y = Math.Min(_bgDefectStart.Y, current.Y);
double w = Math.Abs(current.X - _bgDefectStart.X);
double h = Math.Abs(current.Y - _bgDefectStart.Y);
System.Windows.Controls.Canvas.SetLeft(_bgDefectPreview, x);
System.Windows.Controls.Canvas.SetTop(_bgDefectPreview, y);
_bgDefectPreview.Width = Math.Max(1, w);
_bgDefectPreview.Height = Math.Max(1, h);
}
private void OnMainCanvasPreviewMouseUp(object sender, System.Windows.Input.MouseButtonEventArgs e)
{
if (!_bgDefectDrawing) return;
_bgDefectDrawing = false;
var completedMode = _bgDefectRoiMode;
_bgDefectRoiMode = BackgroundDefectRoiMode.None;
RoiCanvas.ReleaseMouseCapture();
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
var end = e.GetPosition(canvas);
int x = (int)Math.Min(_bgDefectStart.X, end.X);
int y = (int)Math.Min(_bgDefectStart.Y, end.Y);
int w = (int)Math.Abs(end.X - _bgDefectStart.X);
int h = (int)Math.Abs(end.Y - _bgDefectStart.Y);
// 移除预览矩形
if (_bgDefectPreview != null)
{
canvas.Children.Remove(_bgDefectPreview);
_bgDefectPreview = null;
}
if (w < 10 || h < 10) return; // 太小忽略
// 模板助手:在画布上保留 ROI 矩形(与试跑匹配叠加分开管理)
if (completedMode == BackgroundDefectRoiMode.TemplateAssistant)
{
RemoveTemplateAssistantPersistRoi();
_templateAssistantRoiPersist = new System.Windows.Shapes.Rectangle
{
Stroke = System.Windows.Media.Brushes.DeepSkyBlue,
StrokeThickness = 1.5,
StrokeDashArray = new System.Windows.Media.DoubleCollection { 4, 2 },
Fill = System.Windows.Media.Brushes.Transparent,
Width = Math.Max(1, w),
Height = Math.Max(1, h),
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(_templateAssistantRoiPersist, x);
System.Windows.Controls.Canvas.SetTop(_templateAssistantRoiPersist, y);
canvas.Children.Add(_templateAssistantRoiPersist);
}
// 发布ROI绘制完成事件
try
{
var ea = ContainerLocator.Current?.Resolve<Prism.Events.IEventAggregator>();
var rect = new System.Windows.Int32Rect(x, y, w, h);
if (completedMode == BackgroundDefectRoiMode.WhiteBackground)
ea?.GetEvent<WhiteBackgroundRoiDrawnEvent>().Publish(rect);
else if (completedMode == BackgroundDefectRoiMode.BlackBackground)
ea?.GetEvent<BlackBackgroundRoiDrawnEvent>().Publish(rect);
else if (completedMode == BackgroundDefectRoiMode.TemplateAssistant)
ea?.GetEvent<TemplateMatchRoiDrawnEvent>().Publish(rect);
}
catch { }
e.Handled = true;
}
private void ClearTemplateMatchOverlays(System.Windows.Controls.Canvas canvas)
{
if (canvas != null)
{
foreach (var el in _tmMatchOverlays)
canvas.Children.Remove(el);
}
_tmMatchOverlays.Clear();
}
private void RemoveTemplateAssistantPersistRoi()
{
if (_templateAssistantRoiPersist == null) return;
var rect = _templateAssistantRoiPersist;
_templateAssistantRoiPersist = null;
if (VisualTreeHelper.GetParent(rect) is Panel p)
p.Children.Remove(rect);
}
private void RenderTemplateMatchPreview(TemplateMatchPreviewPayload payload)
{
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
ClearTemplateMatchOverlays(canvas);
if (payload?.Hits == null || payload.Hits.Count == 0)
return;
var stroke = new SolidColorBrush(Color.FromRgb(255, 140, 0));
stroke.Freeze();
const int crossHalf = 8;
foreach (var h in payload.Hits)
{
var poly = new System.Windows.Shapes.Polygon
{
Stroke = stroke,
StrokeThickness = 2,
Fill = Brushes.Transparent,
IsHitTestVisible = false,
Points = new PointCollection
{
new System.Windows.Point(h.LtX, h.LtY),
new System.Windows.Point(h.RtX, h.RtY),
new System.Windows.Point(h.RbX, h.RbY),
new System.Windows.Point(h.LbX, h.LbY)
}
};
canvas.Children.Add(poly);
_tmMatchOverlays.Add(poly);
var cx = h.CenterX;
var cy = h.CenterY;
var hLine = new System.Windows.Shapes.Line
{
X1 = cx - crossHalf,
Y1 = cy,
X2 = cx + crossHalf,
Y2 = cy,
Stroke = stroke,
StrokeThickness = 1.5,
IsHitTestVisible = false
};
var vLine = new System.Windows.Shapes.Line
{
X1 = cx,
Y1 = cy - crossHalf,
X2 = cx,
Y2 = cy + crossHalf,
Stroke = stroke,
StrokeThickness = 1.5,
IsHitTestVisible = false
};
canvas.Children.Add(hLine);
canvas.Children.Add(vLine);
_tmMatchOverlays.Add(hLine);
_tmMatchOverlays.Add(vLine);
var tb = new System.Windows.Controls.TextBlock
{
Text = $"{h.Score:F2}",
Foreground = stroke,
FontSize = 10,
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(tb, cx + crossHalf + 2);
System.Windows.Controls.Canvas.SetTop(tb, cy - 8);
canvas.Children.Add(tb);
_tmMatchOverlays.Add(tb);
}
}
private void RenderBackgroundDefectResult(
System.Drawing.Rectangle roiRect,
System.Collections.Generic.IReadOnlyList<BackgroundDefectDetectionItem> detections,
bool isBlackBackground)
{
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null || detections == null) return;
// 绘制ROI矩形(蓝色实线,两种模式一致)
var roiShape = new System.Windows.Shapes.Rectangle
{
Stroke = System.Windows.Media.Brushes.Blue,
StrokeThickness = 1,
Width = roiRect.Width,
Height = roiRect.Height,
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(roiShape, roiRect.X);
System.Windows.Controls.Canvas.SetTop(roiShape, roiRect.Y);
canvas.Children.Add(roiShape);
_bgDefectOverlays.Add(roiShape);
var defectBrush = isBlackBackground
? System.Windows.Media.Brushes.LimeGreen
: System.Windows.Media.Brushes.Red;
const int labelPadRightOfRoi = 4;
const double labelLineHeight = 15;
int validCount = detections.Count(d => d.Contour != null && d.Contour.Count >= 2);
double roiMidY = roiRect.Y + roiRect.Height * 0.5;
double labelLeft = roiRect.X + roiRect.Width + labelPadRightOfRoi;
double labelStartY = roiMidY - validCount * labelLineHeight * 0.5;
int labelRow = 0;
foreach (var d in detections)
{
if (d.Contour == null || d.Contour.Count < 2) continue;
var fig = new PathFigure
{
StartPoint = new System.Windows.Point(d.Contour[0].X, d.Contour[0].Y),
IsClosed = true
};
if (d.Contour.Count > 1)
{
fig.Segments.Add(new PolyLineSegment(
d.Contour.Skip(1).Select(p => new System.Windows.Point(p.X, p.Y)), true));
}
var geom = new PathGeometry();
geom.Figures.Add(fig);
var contourPath = new System.Windows.Shapes.Path
{
Data = geom,
Stroke = defectBrush,
StrokeThickness = 1,
Fill = System.Windows.Media.Brushes.Transparent,
IsHitTestVisible = false
};
canvas.Children.Add(contourPath);
_bgDefectOverlays.Add(contourPath);
var chordLine = new System.Windows.Shapes.Line
{
X1 = d.ChordP1.X,
Y1 = d.ChordP1.Y,
X2 = d.ChordP2.X,
Y2 = d.ChordP2.Y,
Stroke = defectBrush,
StrokeThickness = 1.5,
IsHitTestVisible = false
};
canvas.Children.Add(chordLine);
_bgDefectOverlays.Add(chordLine);
double um = d.SizeMicrometers;
string label = um >= 1000 ? $"{um / 1000:F2} mm" : $"{um:F0} μm";
var text = new System.Windows.Controls.TextBlock
{
Text = label,
Foreground = defectBrush,
FontSize = 11,
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(text, labelLeft);
System.Windows.Controls.Canvas.SetTop(text, labelStartY + labelRow * labelLineHeight);
canvas.Children.Add(text);
_bgDefectOverlays.Add(text);
labelRow++;
}
}
private void ClearBackgroundDefectOverlays(System.Windows.Controls.Canvas canvas)
{
foreach (var el in _bgDefectOverlays)
canvas.Children.Remove(el);
_bgDefectOverlays.Clear();
}
#endregion
#region 线
private void ExecuteEdgeLineFitProcessor(Point startPoint, Point endPoint)
{
try
{
var vm = GetMainVm();
if (vm == null) return;
// 获取当前图像
var viewportVm = ContainerLocator.Current?.Resolve<ViewportPanelViewModel>();
var imageSource = viewportVm?.ImageSource as BitmapSource;
if (imageSource == null)
{
SetStatus("直线拟合失败:无可用图像");
return;
}
// 转换为 Emgu.CV Image
BitmapSource source = imageSource;
if (imageSource.Format != System.Windows.Media.PixelFormats.Gray8)
source = new FormatConvertedBitmap(imageSource, System.Windows.Media.PixelFormats.Gray8, null, 0);
int width = source.PixelWidth;
int height = source.PixelHeight;
int stride = width;
byte[] pixels = new byte[height * stride];
source.CopyPixels(pixels, stride, 0);
using var inputImage = new Emgu.CV.Image<Emgu.CV.Structure.Gray, byte>(width, height);
for (int y = 0; y < height; y++)
for (int x = 0; x < width; x++)
inputImage.Data[y, x, 0] = pixels[y * stride + x];
// 创建并配置算子
var processor = new XP.ImageProcessing.Processors.EdgeLineFitProcessor();
processor.SetParameter("StartX", (int)startPoint.X);
processor.SetParameter("StartY", (int)startPoint.Y);
processor.SetParameter("EndX", (int)endPoint.X);
processor.SetParameter("EndY", (int)endPoint.Y);
// 执行处理
var result = processor.Process(inputImage);
// 获取输出数据并在画布上绘制结果
var outputData = processor.OutputData;
if (outputData.ContainsKey("LineFitResult"))
{
var fitResult = outputData["LineFitResult"] as XP.ImageProcessing.Processors.LineFitResult;
if (fitResult != null && fitResult.Success)
{
DrawEdgeLineFitResult(fitResult, outputData);
SetStatus($"直线拟合完成: 角度={fitResult.AngleDegrees:F2}°, 内点={fitResult.Inliers.Count}/{fitResult.EdgePointCount}, 误差={fitResult.FitError:F3}px");
}
else
{
SetStatus("直线拟合失败:未找到足够的边缘点");
}
}
result.Dispose();
}
catch (Exception ex)
{
SetStatus($"直线拟合异常: {ex.Message}");
}
}
private readonly System.Collections.Generic.List<System.Windows.UIElement> _elfOverlays = new();
private void DrawEdgeLineFitResult(
XP.ImageProcessing.Processors.LineFitResult fitResult,
System.Collections.Generic.Dictionary<string, object> outputData)
{
var canvas = FindChildByName<System.Windows.Controls.Canvas>(RoiCanvas, "mainCanvas");
if (canvas == null) return;
// 清除之前的拟合结果
foreach (var el in _elfOverlays)
canvas.Children.Remove(el);
_elfOverlays.Clear();
// 绘制拟合直线(绿色)
var fitLine = new System.Windows.Shapes.Line
{
X1 = fitResult.Endpoint1.X,
Y1 = fitResult.Endpoint1.Y,
X2 = fitResult.Endpoint2.X,
Y2 = fitResult.Endpoint2.Y,
Stroke = System.Windows.Media.Brushes.Lime,
StrokeThickness = 2,
IsHitTestVisible = false
};
canvas.Children.Add(fitLine);
_elfOverlays.Add(fitLine);
// 绘制内点(绿色小圆点)
foreach (var pt in fitResult.Inliers)
{
var dot = new System.Windows.Shapes.Ellipse
{
Width = 6,
Height = 6,
Fill = System.Windows.Media.Brushes.Lime,
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(dot, pt.X - 3);
System.Windows.Controls.Canvas.SetTop(dot, pt.Y - 3);
canvas.Children.Add(dot);
_elfOverlays.Add(dot);
}
// 绘制外点(红色小圆点)
foreach (var pt in fitResult.Outliers)
{
var dot = new System.Windows.Shapes.Ellipse
{
Width = 6,
Height = 6,
Fill = System.Windows.Media.Brushes.Red,
IsHitTestVisible = false
};
System.Windows.Controls.Canvas.SetLeft(dot, pt.X - 3);
System.Windows.Controls.Canvas.SetTop(dot, pt.Y - 3);
canvas.Children.Add(dot);
_elfOverlays.Add(dot);
}
// 绘制角度标注
var labelText = $"∠{fitResult.AngleDegrees:F2}° | Err:{fitResult.FitError:F2}px";
var label = new System.Windows.Controls.TextBlock
{
Text = labelText,
Foreground = System.Windows.Media.Brushes.Yellow,
FontSize = 12,
FontWeight = FontWeights.Bold,
IsHitTestVisible = false
};
double labelX = (fitResult.Endpoint1.X + fitResult.Endpoint2.X) / 2 + 5;
double labelY = (fitResult.Endpoint1.Y + fitResult.Endpoint2.Y) / 2 - 20;
System.Windows.Controls.Canvas.SetLeft(label, labelX);
System.Windows.Controls.Canvas.SetTop(label, labelY);
canvas.Children.Add(label);
_elfOverlays.Add(label);
}
#endregion
private static T FindChildByName<T>(DependencyObject parent, string name) where T : FrameworkElement
{
int count = VisualTreeHelper.GetChildrenCount(parent);
@@ -59,21 +59,11 @@
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="图像宽度 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel Margin="0,0,0,8">
<TextBox DockPanel.Dock="Right" Width="65"
Text="{Binding ImageWidth, UpdateSourceTrigger=PropertyChanged}"
VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="64" Maximum="8192" Value="{Binding ImageWidth}"
SmallChange="1" LargeChange="100" VerticalAlignment="Center" />
</DockPanel>
<TextBox Text="{Binding ImageWidth, Mode=OneWay}" IsReadOnly="True" Background="#F0F0F0"
Height="28" FontSize="11.5" VerticalContentAlignment="Center" Margin="0,0,0,8" />
<TextBlock Text="图像高度 (px)" Style="{StaticResource ParamLabel}" />
<DockPanel>
<TextBox DockPanel.Dock="Right" Width="65"
Text="{Binding ImageHeight, UpdateSourceTrigger=PropertyChanged}"
VerticalContentAlignment="Center" Margin="6,0,0,0" />
<Slider Minimum="64" Maximum="8192" Value="{Binding ImageHeight}"
SmallChange="1" LargeChange="100" VerticalAlignment="Center" />
</DockPanel>
<TextBox Text="{Binding ImageHeight, Mode=OneWay}" IsReadOnly="True" Background="#F0F0F0"
Height="28" FontSize="11.5" VerticalContentAlignment="Center" />
</StackPanel>
</Border>
@@ -81,9 +71,9 @@
<Border Style="{StaticResource CardStyle}">
<StackPanel>
<TextBlock Text="像素格式" Style="{StaticResource ParamLabel}" />
<ComboBox SelectedItem="{Binding SelectedPixelFormat}"
ItemsSource="{Binding PixelFormatOptions}"
Height="28" FontSize="11.5" VerticalContentAlignment="Center" />
<TextBox Text="{Binding SelectedPixelFormat, Mode=OneWay}"
IsReadOnly="True" Background="#F0F0F0"
Height="28" FontSize="11.5" VerticalContentAlignment="Center" />
</StackPanel>
</Border>
@@ -18,9 +18,6 @@ namespace XplorePlane.Views
var type = dc.GetType();
ExecuteCommand(type, dc, "ApplyExposureCommand");
ExecuteCommand(type, dc, "ApplyGainCommand");
ExecuteCommand(type, dc, "ApplyWidthCommand");
ExecuteCommand(type, dc, "ApplyHeightCommand");
ExecuteCommand(type, dc, "ApplyPixelFormatCommand");
}
private static void ExecuteCommand(System.Type type, object dc, string cmdName)
+6
View File
@@ -55,6 +55,12 @@
<HintPath>Libs\Native\BR.AN.PviServices.dll</HintPath>
<Private>True</Private>
</Reference>
<!-- 海康威视相机 SDK (.NET Framework 4.0) -->
<Reference Include="MvCameraControl.Net">
<HintPath>..\ExternalLibraries\MvCameraControl.Net.dll</HintPath>
<Private>true</Private>
</Reference>
</ItemGroup>