stereo calibration

asked 2012-10-05 17:37:57 -0600

I have been using Matlab Calibration Toolbox to calibrate my stereo camera. The result is not satisfying. I extract salient features from both images and verify the calibration result based on rectification result. After rectification, the features in the same pair are misaligned by roughly 10 pixels in y axis. I would expect the rectification error in y axis to be around 1 pixel.

Following the Opencv calibration example, for some images, the software fails to detect all the corner. Can someone comment on this? Also I have attached sample left and right images taken for calibration.

C:\fakepath\left0.jpg C:\fakepath\right0.jpg

edit retag flag offensive close merge delete

add a comment

1

answered 2012-10-05 20:07:05 -0600

Martin Peris
1742 ●10 ●23 http://blog.martinperi...

Hi there!

After watching the images here are my comments:

I see that you posted JPG files, I don't know if those are the original files you used, but whenever possible use raw images (not compressed, as compression generates artifacts that affect the ability of detecting the corners of the chessboard)
The size of the squares might be too small in relation to the resolution of your camera. How about using bigger squares?
This is not related to corner detection, but I see that you are using a 16x16 chessboard. I would recommend you to use a non-square chessboard (16x15, for example), this way the orientation of the chessboard can be uniquely determined. Otherwise you can get strange effects on the calibration.

I hope this helps