CPC G06F 3/012 (2013.01) [G06F 3/013 (2013.01); G06F 3/04815 (2013.01); G06T 7/50 (2017.01); G06T 11/00 (2013.01); H04N 7/147 (2013.01); H04N 7/18 (2013.01); H04N 13/117 (2018.05); H04N 13/239 (2018.05); H04N 13/243 (2018.05); H04N 13/254 (2018.05); H04N 13/344 (2018.05); H04N 23/80 (2023.01); G06T 2207/10048 (2013.01); H04N 5/33 (2013.01); H04N 13/271 (2018.05); H04N 2013/0081 (2013.01)] | 16 Claims |
1. A method for generating rich scene information representative of a scene, the method comprising:
in a digital processing resource comprising at least one digital processor:
(1) receiving data from at least two sensors, wherein at least two of the at least two sensors comprise a stereo pair of cameras, the received sensor data being at least in part representative of the scene;
(2) detecting reliability of the received sensor data, thereby generating reliability information, the reliability information being representative of the reliability of the data received from at least one sensor, wherein detecting reliability of the received sensor data comprises:
(a) comparing the output of a sensor to the output from at least one other sensor;
(b) performing stereo disparity analysis between at least two cameras, the stereo disparity analysis generating stereo disparity data; and
(c) wherein the comparing utilizes at least one histogram that pertains to stereo disparity data:
(3) remedying unreliable received sensor data to generate remedied data; and
(4) generating rich scene information representative of the scene and useable to reconstruct a view of the scene as viewed from a virtual viewpoint, wherein the generating of rich scene information utilizes (a) the received sensor data, (b) the remedied data and (c) the reliability information.
|