Hello Community, Hello Peter,
I would be very pleased to get the solution to my problem. The example "Example Stereo Disparity 3D" shows how to calculate a 3D point cloud from two images and the stereo parameters.
I would like to calculate only the 3D coordinates of single 2D image points (marker points of the checkerboard) in real time (as tracking of a probe) with my stereo camera. The 2D checkerboard points on both raw images are easy to find. Also my stereo cameras are calibrated and I have the stereo parameters. At this point I am not getting further.
The "Example Stereo Disparity 3D" rectifies the whole stereo images to one plane. Is it also possible to rectify only my found chess points to this plane? This would save a lot of performance.
If these points are then on the common plane, is there another method to get the 3D coordinate for each stereo pair?
In summary, I need a method that [A] gets a list of 2D chess points on the left ,[B] a list of 2D chess points on the right and [C] stereo parameters and returns me a list of 3D chess board points in the camera coordinate system.
I am very excited about the boofcv library. I've been doing very well with stereo calibration, but I'm a bit lacking with Stereo Disparity 3D. I would really appreciate the support. Thanks a lot in advance!