camera_calibration分析（Camera_calibration analysis）.docVIP

camera_calibration分析（Camera_calibration analysis） 设置; / /读取设置 FileStorage fs(inputSettingsFile,FileStorage::READ);/ /读取设置 fs(“设置”) ; fs.release(); 矢量 Point2f imagePoints;/ /每幅图像的角点向量集 垫cameraMatrix distCoeffs; 垫;/ /图像缓存 / /采集图像 视图= s.nextImage(); 垫的结果; 如果(inputCapture.isOpened()) { 垫view0; inputCapture view0; view0。复制(结果); } 返回结果; imageSize = view . size();/ /格式化输入图像。 / /如果采集完毕,进行校正 if(mode = =捕获 imagePoints.size() =(unsigned)。nrFrames)/ / 25 { if(run校准和save(s,imageSize,cameraMatrix,distCoeffs,imagePoints)) 向量垫 rvecs tvecs; 飘 reprojErrs向量; 双totalAvgErr = 0; bool ok = run校准(s、imageSize、cameraMatrix、distCoeffs、imagePoints、rvecs、tvecs、reprojErrs、totalAvgErr); cameraMatrix =垫:眼睛(3 3 CV_64F);/ /对角矩阵 distCoeffs =垫::0(8 1 CV_64F);/ /全零矩阵 矢量 Point3f objectPoints(1);/ / 3 d点物理坐标向量 calcBoardCornerPositions(s。boardSize,s。squareSize,objectPoints[0],s.calibrationPattern); corners.clear(); for(int i = 0;i boardSize)。高度;+ + i) for(int j = 0;j boardSize)。宽度;+ + j) / /棋盘角点的3 d点物理坐标 角落。push_back(Point3f(浮点数(j *平方esize)，float(i *平方esize)，0); objectPoints.resize(imagePoints.size(),objectPoints[0]); 双rms =校准照相机(objectPoints,imagePoints,imageSize,cameraMatrix,distCoeffs,rvecs,tvecs,s.flag | CV_CALIB_FIX_K4 | CV_CALIB_FIX_K5); The rotation vector of the parameters of the parameters of the parameters of the coordinates of the coordinates of the coordinates of the coordinates of the coordinates of the coordinates of the coordinates of the coordinates of the angular coordinates Size_t nimages = _objectPoints. Total (); Mat objPt, imgPt, npoints, rvecM ((int) nimages, 3, CV_64FC1), tvecM ((int) nimages, 3, CV_64FC1); CollectCalibrationData (_objectPoints _imagePoints, noArray (), objPt, imgPt, 0, npoints); Int nimages = (int) objectPoints. Total (); / / how much memory does the objectPoints occupy? For (I = 0; I nimages; I + +) { Ni = objectPoints. GetMat (I). CheckVector (3, CV_32F); CV_Assert (ni = 0); Total + = ni; } / / allocate cache space Npoints. Create (1, (int) nimages,