校准，上标，stereoRectify的结果错误

如何解决校准，上标，stereoRectify的结果错误

/请理解英语是尴尬的，因为我是韩国人。/

Ubuntu：20.04，opencv：4.0

double rms = cv::calibrateCamera(objectPoints,imagePoints,imageSize,cameraMatrix,distCoeffs,rvecs,tvecs,cv::CALIB_FIX_K5);

校准

cv::BFMatcher matcher;
std::vector<cv::DMatch> matches;
matcher.match(descriptors1,descriptors2,matches);
//-- Filter matches using the Lowe's ratio test
std::vector<cv::Point2f> pt1;
std::vector<cv::Point2f> pt2;
std::vector<cv::DMatch> good_matches;


for (size_t i = 0; i < matches.size(); i++)
{
    if (matches[i].distance < 0.6 * matches[i+1].distance)
    {
        good_matches.push_back(matches[i]);
        pt1.push_back(keypoints1[matches[i].queryIdx].pt);
        pt2.push_back(keypoints2[matches[i].trainIdx].pt);

    }
}

存储好点（我使用过SURF）

std::cout<< pt1 <<std::endl;
std::cout<< pt2 <<std::endl;

std::vector<cv::Point2f> upt1;
std::vector<cv::Point2f> upt2;
cv::undistortPoints(pt1,upt1,distCoeffs);
cv::undistortPoints(pt2,upt2,distCoeffs);    
cv::Mat f=cv::findFundamentalMat(cv::Mat(upt1),cv::Mat(upt2),cv::FM_RANSAC,3,0.99);

找到基本面

std::vector<cv::Vec3f> lines1;
cv::computeCorrespondEpilines(cv::Mat(upt1),1,f,lines1);
for (auto iter = lines1.begin(); iter != lines1.end(); ++iter){
    cv::line(img2,cv::Point(0,-(*iter)[2] / (*iter)[1]),cv::Point(img2.cols,-((*iter)[2] + (*iter)[0] * img2.cols) / (*iter)[1]),cv::Scalar(255,255,255));
}

// Draw the RIGHT points corresponding epipolar lines in left image
std::vector<cv::Vec3f> lines2;
cv::computeCorrespondEpilines(cv::Mat(upt2),2,lines2);
for (auto iter = lines2.begin(); iter != lines2.end(); ++iter){

    cv::line(img1,cv::Point(img1.cols,-((*iter)[2] + (*iter)[0] * img1.cols) / (*iter)[1]),255));

绘制apilines

cv::Mat rr = cv::Mat::zeros(3,CV_64F);

for(int i=0; i<(int)rvecs.size();i++)
{
cv::Rodrigues(rvecs[i],rr,cv::noArray());
}

cv::Mat tt(3,CV_64F);
for( const auto& t: tvecs)
{
  tt += t;
}
// avoid division by zero
if (!tvecs.empty())
  tt /= tvecs.size();


cv::stereoRectify( cameraMatrix,imgsize,tt,R1,R2,P1,P2,Q,1.0,&RL,&RR );


initUndistortRectifyMap(cameraMatrix,CV_16SC2,map11,map12);
initUndistortRectifyMap(cameraMatrix,map21,map22);

remap(img1,view1,map12,cv::INTER_LINEAR,cv::BORDER_CONSTANT,cv::Scalar());
remap(img2,view2,map22,cv::Scalar());

cv::imwrite("View1.jpg",view1);
cv::imwrite("View2.jpg",view2);

我开病了，现在我想改正。 但是，结果图像是 enter image description here

和 我是否对cv :: computeCorrespondEpilines做错了，所以Epilines也很奇怪。enter image description here

我只有一台摄像机，所以我只有一台cameraMatrix和一台Disrcoeffs。

rms = 1.16223

我想知道的是我应该对cv :: stereoRectify（）使用cv :: stereoCalibrate（）而不是cv :: calibrateCamera（）吗？ 为何左/右image.jpg的极线和结果怪异？ 我引用了此网站。enter link description here

和enter link description here