I'm using a RGB-D camera (Intel Realsense D345) to implement a table top projected augmented reality system. Using chessboard calibration I obtain a transformation matrix which I use to transform each incoming frame using warpPerspective from openCV. It works really well for the color frames. The problem is, am I allowed to do this for depth images as well ? Considering depth images are 3D geometrical data. What's the right way to apply a transformation matrix to depth images?
This forum is disabled, please visit https://forum.opencv.org