stereo camera calibration error

Hi all,

I am implementing a simple program to calibrate two cameras and compute a disparity map. I was given a set of 425 pictures of a chessboard, taken using the two cameras. Using the "Camera Calibration" tool I calibrated each camera individually, and I selected the images with reprojection errors lower than 0.7 (I kept about 70 pictures for each camera). Then, I calibrated the stereo pair using the same set of images.

The issue is that when I load the fixed intrinsics in the "Stereo Camera Calibration" tool, I get the following error message:

The value of 'principalPoint' is invalid. Expected principalPoint to be positive.

I am new to computer vision and I am pretty sure that I am doing something wrong.

I tried to compute the disparity maps without single-camera calibration. I got very poor results, and I think that it is due to bad camera calibration.

Thank you in advance for your help.

Have a nice day!

Answer 1

[Vinai Datta Thatiparthi]

Hi,

I noticed a similar discussion on the forum that could be relevant to solving the error that you're reporting:

LINK to MATLAB Answer: https://www.mathworks.com/matlabcentral/answers/370485-camera-calibration-failed

Hope this helps!

[DavideC]

Hi,

Unfortunately, my issue concerns the "Principal Point" and not image size.

Thank you for your answer.

Have a nice day!

Answer 2

[Kevin Mallery]

I'm getting the same error. So far, all I've been able to figure out as a "solution" is to just disable the positivity check in the checkPrincipalPoint function in cameraIntrinsics.m. As far as I can tell, I don't see any reason why the principal point should be constrained to be positive.

Could someone at Mathworks comment on this? I would prefer a solution that is less hacky.