应用介绍

此项目是打地鼠游戏的视觉识别实现 。

#include "calibration.h"

const int imageWidth = 640;                             //摄像头的分辨率 

const int imageHeight = 480;

const int boardWidth = 7;                               //横向的角点数目 

const int boardHeight = 5;                              //纵向的角点数据 

const int boardCorner = boardWidth * boardHeight;       //总的角点数据 

const int frameNumber = 13;                             //相机标定时需要采用的图像帧数 

const int squareSize = 30;                              //标定板黑白格子的大小 单位mm 

const Size boardSize = Size(boardWidth, boardHeight);   // 

Mat intrinsic;                                          //相机内参数 

Mat distortion_coeff;                                   //相机畸变参数 

vector<Mat> rvecs;                                        //旋转向量 

vector<Mat> tvecs;                                        //平移向量 

vector<vector<Point2f>> corners;                        //各个图像找到的角点的集合 和objRealPoint 一一对应 

vector<vector<Point3f>> objRealPoint;                   //各副图像的角点的实际物理坐标集合 

vector<Point2f> corner;                                   //某一副图像找到的角点 

Mat rgbImage, grayImage;

/*计算标定板上模块的实际物理坐标*/

void calRealPoint(vector<vector<Point3f>>& obj, int boardwidth, int boardheight, int imgNumber, int squaresize)

{

 //  Mat imgpoint(boardheight, boardwidth, CV_32FC3,Scalar(0,0,0)); 

 vector<Point3f> imgpoint;

 for (int rowIndex = 0; rowIndex < boardheight; rowIndex++)

 {

  for (int colIndex = 0; colIndex < boardwidth; colIndex++)

  {

   //  imgpoint.at<Vec3f>(rowIndex, colIndex) = Vec3f(rowIndex * squaresize, colIndex*squaresize, 0); 

   imgpoint.push_back(Point3f(rowIndex * squaresize, colIndex * squaresize, 0));

  }

 }

 for (int imgIndex = 0; imgIndex < imgNumber; imgIndex++)

 {

  obj.push_back(imgpoint);

 }

}

/*设置相机的初始参数 也可以不估计*/

void guessCameraParam(void)

{

 /*分配内存*/

 intrinsic.create(3, 3, CV_64FC1);

 distortion_coeff.create(5, 1, CV_64FC1);

 /*

 fx 0 cx

 0 fy cy

 0 0  1

 */

 intrinsic.at<double>(0, 0) = 256.8093262;   //fx        

 intrinsic.at<double>(0, 2) = 160.2826538;   //cx 

 intrinsic.at<double>(1, 1) = 254.7511139;   //fy 

 intrinsic.at<double>(1, 2) = 127.6264572;   //cy 

 intrinsic.at<double>(0, 1) = 0;

 intrinsic.at<double>(1, 0) = 0;

 intrinsic.at<double>(2, 0) = 0;

 intrinsic.at<double>(2, 1) = 0;

 intrinsic.at<double>(2, 2) = 1;

 /*

 k1 k2 p1 p2 p3

 */

 distortion_coeff.at<double>(0, 0) = -0.193740;  //k1 

 distortion_coeff.at<double>(1, 0) = -0.378588;  //k2 

 distortion_coeff.at<double>(2, 0) = 0.028980;   //p1 

 distortion_coeff.at<double>(3, 0) = 0.008136;   //p2 

 distortion_coeff.at<double>(4, 0) = 0;          //p3 

}

void outputCameraParam(void)

{

 /*保存数据*/

 CvMat cv_mat_intrinsic = intrinsic;

 CvMat cv_mat_distortion_coeff = distortion_coeff;

 cvSave("cameraMatrix.xml", &cv_mat_intrinsic);

 cvSave("cameraDistoration.xml", &cv_mat_distortion_coeff);

 //cvSave("rotatoVector.xml", &rvecs); 

 //cvSave("translationVector.xml", &tvecs); 

 /*输出数据*/

 cout << "fx :" << intrinsic.at<double>(0, 0) << endl << "fy :" << intrinsic.at<double>(1, 1) << endl;

 cout << "cx :" << intrinsic.at<double>(0, 2) << endl << "cy :" << intrinsic.at<double>(1, 2) << endl;

 cout << "k1 :" << distortion_coeff.at<double>(0, 0) << endl;

 cout << "k2 :" << distortion_coeff.at<double>(1, 0) << endl;

 cout << "p1 :" << distortion_coeff.at<double>(2, 0) << endl;

 cout << "p2 :" << distortion_coeff.at<double>(3, 0) << endl;

 cout << "p3 :" << distortion_coeff.at<double>(4, 0) << endl;

}

void calibrate()

{

 Mat img;

 int goodFrameCount = 0;

 namedWindow("chessboard");

 cout << "按Q退出 ..." << endl;

 while (goodFrameCount < frameNumber)

 {

  char filename[100];

  sprintf_s(filename, "chessboard\\chess%d.jpg", goodFrameCount++);

  //  cout << filename << endl; 

  rgbImage = imread(filename, CV_LOAD_IMAGE_COLOR);

  cvtColor(rgbImage, grayImage, CV_BGR2GRAY);

  imshow("Camera", grayImage);

  bool isFind = findChessboardCorners(rgbImage, boardSize, corner, 0);

  if (isFind == true) //所有角点都被找到 说明这幅图像是可行的 

  {

   /*

   Size(5,5) 搜索窗口的一半大小

   Size(-1,-1) 死区的一半尺寸

   TermCriteria(CV_TERMCRIT_EPS | CV_TERMCRIT_ITER, 20, 0.1)迭代终止条件

   */

   cornerSubPix(grayImage, corner, Size(5, 5), Size(-1, -1), TermCriteria(CV_TERMCRIT_EPS | CV_TERMCRIT_ITER, 20, 0.1));

   drawChessboardCorners(rgbImage, boardSize, corner, isFind);

   imshow("chessboard", rgbImage);

   corners.push_back(corner);

   //string filename = "res\\image\\calibration"; 

   //filename += goodFrameCount + ".jpg"; 

   //cvSaveImage(filename.c_str(), &IplImage(rgbImage));       //把合格的图片保存起来 

   cout << "The image is good" << endl;

  }

  else

  {

   cout << "The image is bad please try again" << endl;

  }

  //  cout << "Press any key to continue..." << endl; 

  //  waitKey(0); 

  if (waitKey(10) == 'q')

  {

   break;

  }

    imshow("chessboard", rgbImage); 

    //waitKey(0);

 }

 /*

 图像采集完毕 接下来开始摄像头的校正

 calibrateCamera()

 输入参数 objectPoints  角点的实际物理坐标

 imagePoints   角点的图像坐标

 imageSize     图像的大小

 输出参数

 cameraMatrix  相机的内参矩阵

 distCoeffs    相机的畸变参数

 rvecs         旋转矢量(外参数)

 tvecs         平移矢量(外参数）

 */

 /*设置实际初始参数 根据calibrateCamera来 如果flag = 0 也可以不进行设置*/

 guessCameraParam();

 cout << "guess successful" << endl;

 /*计算实际的校正点的三维坐标*/

 calRealPoint(objRealPoint, boardWidth, boardHeight, frameNumber, squareSize);

 cout << "cal real successful" << endl;

 /*标定摄像头*/

 calibrateCamera(objRealPoint, corners, Size(imageWidth, imageHeight), intrinsic, distortion_coeff, rvecs, tvecs, 0);

 cout << "calibration successful" << endl;

 /*保存并输出参数*/

 outputCameraParam();

 cout << "out successful" << endl;

 /*显示畸变校正效果*/

 //Mat cImage;

 //undistort(rgbImage, cImage, intrinsic, distortion_coeff);

 //imshow("Corret Image", cImage);

 //cout << "Correct Image" << endl;

 //cout << "Wait for Key" << endl;

 //waitKey(0);

 //system("pause");

}

。。。。。。。。想了解详情请下载附件