基于EmguCV的摄像机标定及矫正

标签： EmguCV摄像头标定C#
2015-05-03 14:55 501人阅读 评论(6) 收藏 举报
分类：
C#版权声明：本文为博主原创文章，未经博主允许不得转载。目录(?)[+]

前言

之前用OpenCV做过摄像机标定及矫正，现在平台换了，改用C#，就用EmguCV做一下，其实就是OpenCV的C#版。在EmguCV中有两类摄像机标定的类，一个是CvInvoke类，一个是CameraCalibration类，两种标定效果差不多，只不过CvInvoke涉及的函数大多都是指针类型的，而C#对于指针的操作比较麻烦。本文是在CameraCalibration类的基础上完成摄像机标定，用CvInvoke类完成矫正图像。

函数说明

1、角点检测

[csharp] view plain copy print ?public static PointF[] FindChessboardCorners(Image<Gray, byte> image,Size patternSize,CALIB_CB_TYPE flags)

ParametersimageType: Emgu.CV.Image<Gray, Byte>Source chessboard viewpatternSizeType: System.Drawing.SizeThe number of inner corners per chessboard row and columnflagsType: Emgu.CV.CvEnum.CALIB_CB_TYPEVarious operation flags
Return ValueType: PointF[]The corners detected if the chess board pattern is found, otherwise null is returned注：输入图像需是灰度图，检测之前角点需要开辟空间，如：[csharp] view plain copy print ?cornersDetected = new PointF[nPoints]; //nPoints表示单幅图像角点总数cornersDetected = CameraCalibration.FindChessboardCorners(chessboardImage, patternSize,CALIB_CB_TYPE.ADAPTIVE_THRESH | CALIB_CB_TYPE.NORMALIZE_IMAGE);

2、标定函数

[csharp] view plain copy print ?public static double CalibrateCamera(MCvPoint3D32f[][] objectPoints,PointF[][] imagePoints,Size imageSize,IntrinsicCameraParameters intrinsicParam,CALIB_TYPE calibrationType,MCvTermCriteria termCriteria,out ExtrinsicCameraParameters[] extrinsicParams)

ParametersobjectPointsType: Emgu.CV.Structure.MCvPoint3D32f[][]The 3D location of the object points. The first index is the index of image, second index is the index of the pointimagePointsType: System.Drawing..::..PointF[][]The 2D image location of the points. The first index is the index of the image, second index is the index of the pointimageSizeType: System.Drawing.SizeThe size of the image, used only to initialize intrinsic camera matrixintrinsicParamType: Emgu.CV.IntrinsicCameraParametersThe intrisinc parameters, might contains some initial values. The values will be modified by this function.calibrationTypeType: Emgu.CV.CvEnum.CALIB_TYPEcCalibration typetermCriteriaType: Emgu.CV.Structure.MCvTermCriteriaThe termination criteriaextrinsicParamsType: Emgu.CV.ExtrinsicCameraParameters[]The output array of extrinsic parameters.
Return ValueType: DoubleThe final reprojection error注：objectPoints表示棋盘角点在世界坐标系下的坐标，有多少幅棋盘图像就应有多少角点坐标集，以物理尺寸为单位。imagePoints表示角点在图像中的坐标，以像素为单位。返回值是重投影误差。

3、映射矩阵求取

[csharp] view plain copy print ?public static void cvInitUndistortMap(IntPtr intrinsicMatrix,IntPtr distortionCoeffs,IntPtr mapx,IntPtr mapy)

ParametersintrinsicMatrixType: System.IntPtrThe camera matrix (A) [fx 0 cx; 0 fy cy; 0 0 1]distortionCoeffsType: System.ntPtrThe vector of distortion coefficients, 4x1 or 1x4 [k1, k2, p1, p2].mapxType: System.IntPtrThe output array of x-coordinates of the mapmapyType: System.ntPtrThe output array of y-coordinates of the map注：定义Matrix类的映射矩阵变量，其属性中包含Ptr，可以直接当指针用，如：[csharp] view plain copy print ?private Matrix<float> mapx = new Matrix<float>(height, width);private Matrix<float> mapy = new Matrix<float>(height, width);4、几何变换[csharp]
view plain
copy
print
?public static void cvRemap(IntPtr src,IntPtr dst,IntPtr mapx,IntPtr mapy,int flags,MCvScalar fillval)ParameterssrcType: System.ntPtr
Source imagedstType: System.IntPtr
Destination imagemapxType: System.IntPtr
The map of x-coordinates (32fC1 image)mapyType: System.ntPtr
The map of y-coordinates (32fC1 image)flagsType: System.Int32
A combination of interpolation method and the optional flag CV_WARP_FILL_OUTLIERSfillvalType: Emgu.CV.Structure.MCvScalar
A value used to fill outliers注：flags定义在 CvEnum下的 WARP枚举类型，调用： (int)WARP.CV_WARP_INVERSE_MAP

程序说明

基于EmguCV摄像机标定及矫正的软件界面如下：如图所示，界面包含棋盘格信息设置，标定及矫正事件的实现等等。矫正前后图像对比:代码实现了从摄像头读取棋盘格图像或者从本地读取图像，图像个数有Imges指定，棋盘格大小有Square Size指定；然后成功检测到角点之后进行摄像头标定，保存角点值和摄像头内参数，通过Rectify按钮实现畸变矫正功能。为避免每次标定时都要检测角点，设置Read Corners按钮，读取角点(包含objectPoints和imagePoints)，然后Start Calibrate实现标定。所有的数据都是保存到xml文件中，方便查看和提取。