I have a question, when I run findChessboardCorners
in the Qt-MinGW platform, I can get the correct board_feature_locations_xy results, but when I run the same code on the VS2015 platform, I get the wrong results, but if I change the definition std::vector<cv::Point2f> board_feature_locations_xy;
to
cv::Mat board_feature_locations_xy;
I can get the correct results.
So anyone can solve this problem?
I use opencv-3.1.0.
std::vector<cv::Point2f> board_feature_locations_xy;
cv::Size board_feature_size(8, 6);
cv::findChessboardCorners(calibration_image_cv, board_feature_size, board_feature_locations_xy, CV_CALIB_CB_ADAPTIVE_THRESH | CV_CALIB_CB_FILTER_QUADS);