CV Register Point Clouds (3D)

Finds the transform that maps one 3D point set onto another given CORRESPONDING points (cv2.estimateAffine3D) - the 3D counterpart of 'OpenCV Find Homography (RANSAC)'. Typical use: two stereo pairs of the same scene. Match features between the two left views ('CV Match Features'), read the 3D position of each match out of the two XYZ maps ('CV Sample Array At Points' on cv2.reprojectImageTo3D's output), and this node returns the 3x4 [R|t] that brings the second cloud into the first one's frame - apply it with 'CV Transform Points 3D' and merge with 'CV Merge Point Clouds'. 'RANSAC + rigid refit' (default) rejects the outliers that noisy stereo depth always produces, then re-fits a clean rigid transform on the inliers only. Failure tolerant: fewer than 3 correspondences, a count mismatch or a degenerate configuration returns the identity transform with found=false instead of raising, so a failed registration merges the clouds unaligned rather than halting the workflow - branch on 'found' with a 'Basic data handling: IfElse' if you would rather skip the second cloud entirely.