Object-based automatic white balance based on neural network models

An apparatus comprising:

• an interface configured to receive pixel data;
• and
• a processor configured to (i) process said pixel data arranged as video frames, (ii) perform computer vision operations on said video frames to detect objects in said video frames, (iii) analyze characteristics of said objects detected in a current video frame of said video frames, (iv) determine adaptive auto-white balance weightings for each block region of said current video frame in response to said characteristics of said objects detected, (v) generate an auto-white balance weighting table based on said adaptive auto-white balance weightings specific to said current video frame and (vi) generate an output video frame in response to (A) using said auto-white balance weighting table to adjust white balance gain values in said current video frame when one of said objects is detected in said current video frame, said adaptive auto-white balance weightings are used to determine corresponding white balance gain values for respective block regions based on said characteristics of said objects and (B) using a fixed weighting table to adjust said white balance gain values in said current video frame when none of said objects are detected in said current video frame, wherein
• (a) said output video frame is generated in response to using said auto-white balance weighting table generated specific to said current video frame to adjust white balance gain values in said current video frame,
• (b) said auto-white balance weighting table is individually generated for only specific use with said current video frame to dynamically adjust to said characteristics specific to said objects detected in said current video frame being analyzed, and
• (c) said fixed weighting table comprises pre-defined white balance gain values selected for a particular type of scene captured in said video frames from a plurality of predefined scene types.