Dear all,
I am working with OpenCV since some weeks. Camera calibration and generation of point clouds (PLY) works perfectly. What I now experience is problems putting everything together into a coordinate system. My setup looks like this:
The top of the box needs to be my Z-layer. I played with extrinsic calibration but my coordinate system always stays at the camera viewpoint. Maybe I am on the wrong way but I need to get managed that the box comes into the world coordinate system.
For the beginning it would be enough to do this for one camera as I can rotate the other ones with a simple transformation as soon as I found a solution for one camera.
May basis was the stereo_calib.cpp and stereo_match.cpp from the “samples/CPP”. I am using an asymmetric chessboard calibration plate.
Do you have some hits what I need to do? Maybe also existing samples in Python or CPP?
Thank you and best regards
Markus