Hello,
I did eye-in-hand calibration with camera calibration before that and now have both transformation matrices. Where I am confused is how to use them to get robot picking point from camera pixel point?
I multiply matrices like: H_gripper2base @ H_cam2gripper @ point_homogeneous.T but I cannot get the good result. I think I am wrong how to get this point_homogeneous point, how to convert pixel coordinates to camera frame world point if I have camera matrix (from cv.calibrateCamera()), point of interest in pixels and robot gripper position.