Paul Doliotis's profile - activity

I just found out that calling findChessboardCorners with these flags, seems to be working for all images:

found = findChessboardCorners( view, s.boardSize, pointBuf, CV_CALIB_CB_ADAPTIVE_THRESH | CV_CALIB_CB_FILTER_QUADS);

Actually the answer is:

from calibration grid coordinate system (aka World coordinate system) to the camera calibration system

Hi all,

I am using the OpenCV calibration code from this tutorial: http://docs.opencv.org/doc/tutorials/calib3d/camera_calibration/camera_calibration.html#cameracalibrationopencv

Here is the arguments list for the calibration function: runCalibration(s,imageSize, cameraMatrix, distCoeffs, imagePoints, rvecs, tvecs, reprojErrs, totalAvgErr);

rvecs and tvecs are essentially the rotational and transnational components of the Homogeneous transformation matrix.

My question is the following:

does the transformation matrix( using rvecs, tvecs) take us from:

calibration grid coordinate system to the camera coordinate system

or

from camera coordinate system to the calibration grid coordinate system.

Thanks,

Paul

Hey all,

I am giving some images as input to findChessboardCorners but it fails to detect the corners. I believe the calibration images are not very challenging for automatic corner extractions. Here are two sample images: http://vlm1.uta.edu/~doliotis/temp/opencv_calibration/9.bmp http://vlm1.uta.edu/~doliotis/temp/opencv_calibration/old_0.bmp

Any ideas why findChessboardCorners is failing? Are the squares too small? Thanks, Paul

p.s. I have verified that the function works with other images. I am using opencv 243