I was wondering whether it is possible to obtain the camera intrinsics from the rectified left and right stereo pair, as I am only able to obtain the original intrinsics for each of the monocams. I need these to compute some Q matrices from the rectified image to rgb camera outside of the pipelines.
Reading out rectified left/right camera intrinsics.
Our method to try to align the depth, projected on (rectified) left image, to rgb camera (outside of pipeline) is not working, and we are trying to figure out why. Why would they stay the same? are there then any other computations going on to do this alignment internally?