Revision history [back]

if care to read tutorial and paper , you'll see it needs 20 stereo images (+2) for graycode and 3 for sinusodial

apart from that, there is block-matching(from stereo cams), photometric stereo, structure from motion, and also dnns trained to get disparity from a single image are on the way.

if care to read tutorial and paper , you'll see it needs 20 stereo images (+2) for graycode and 3 (mono cam) for sinusodial

apart from that, there is block-matching(from stereo cams), photometric stereo, structure from motion, and also dnns trained to get disparity from a single image are on the way.