calibration without chessboard single camera

asked 2013-04-03 05:42:19 -0600

surfermac gravatar image

updated 2013-04-03 05:43:34 -0600

I have a set of points on a paper and i want to calibrate my camera using those points. I ve put colors on each point and have done a color thresholding and am currently having all the image points I ve also took object points respectively when i try to calibrate with calibrate camera function its giving me an error. \


OpenCV Error: Assertion failed (ni >= 0) in collectCalibrationData, file /home/mac/OpenCV-2.4.1/modules/calib3d/src/calibration.cpp, line 3172 terminate called after throwing an instance of 'cv::Exception' what(): /home/mac/OpenCV-2.4.1/modules/calib3d/src/calibration.cpp:3172: error: (-215) ni >= 0 in function collectCalibrationData

is there any way to get past this?

edit retag flag offensive close merge delete


What would be the expected result of a single-camera calibration? What do you want it for? Stereo calibration corrects small alignment issues between two cameras mounted on a rig. Calibrating a single camera in this context does not make much sense.

sammy gravatar imagesammy ( 2013-04-03 06:10:26 -0600 )edit

I just want the coordinates of another color dot with respect to my calibrated object points. the z value always remains zero so all i have is a plane. the plane may be inclined so that could be turned back into a straight plane with the calibration values i get. the program i wrote is just a modification of the default calibration program. but it just doesn't calibrate with my points.

surfermac gravatar imagesurfermac ( 2013-04-04 02:45:57 -0600 )edit

i got the object points messed up and that was the cause of the error. I m now able to calibrate but the rms error is a bit high around 3.73 any suggestions to increase the accuracy .

surfermac gravatar imagesurfermac ( 2013-04-15 00:25:52 -0600 )edit