You need two images - stereo - in order to reconstruct depth information easily. For anything else, you need the "magic" of AI.