Fallback mechanism for auto white balancing
Improved fallback mechanisms for auto white balancing are presented. In at least one embodiment, white balance correction factors produced by a first white balance technique are blended with white balance correction factors produced by a second white balance technique based on a confidence level in the white balance correction factors produced by the first white balance technique.
1 . A method comprising:
determining one or more correction factors based at least on an estimated illuminant color of an image;
determining one or more blended correction factors by blending the one or more correction factors with one or more alternative correction factors according to one or more blending weights that are determined based on a difference between a confidence level of the estimated illuminant color and a confidence threshold; and
performing white balancing of the image based at least on applying the one or more blended correction factors to the image.
2 . The method of claim 1 , further comprising:
determining the one or more alternative correction factors based at least on an estimated scene brightness of the image.
3 . The method of claim 1 , further comprising:
determining the one or more alternative correction factors based at least on an alternative estimated illuminant color of the image.
4 . The method of claim 1 , wherein the blending the one or more correction factors with the one or more alternative correction factors comprises:
determining the confidence level in the estimated illuminant color; and
determining one or more blending weights to be applied to the one or more correction factors and the one or more alternative correction factors, the determining the one or more blending weights being based at least on a difference between the confidence level and at least one of a low confidence threshold or a high confidence threshold; and
blending the one or more correction factors with the one or more alternative correction factors according to the one or more blending weights.
5 . The method of claim 1 , wherein the estimated illuminant color is obtained based at least on applying a chromaticity weighting function to the image, and wherein the confidence level is an average weight determined using the chromaticity weighting function.
6 . The method of claim 1 , wherein the estimated illuminant color is determined based at least on normalizing the image, wherein the confidence level is based on a measure of reliability of the normalizing, and wherein the measure of reliability comprises an assessment of a number of valid pixels in the normalized image.
7 . The method of claim 1 , further comprising:
determining one or more additional correction factors; and
based at least on a confidence level associated with the one or more alternative correction factors, determining one or more further blended correction factors by blending the one or more blended correction factors with the one or more additional correction factors, and
wherein the performing white balancing of the image is based at least on applying the one or more further blended correction factors to the image.
8 . A method comprising:
determining a confidence level associated with one or more correction factors;
based at least on comparing the confidence level to a confidence threshold, selecting, from a plurality of alternative correction factors, one or more alternative correction factors corresponding to an estimated scene brightness of a scene depicted by an image; and
performing white balancing of the image based at least on applying the one or more alternative correction factors to the image.
9 . The method of claim 8 , wherein the applying the one or more alternative correction factors to the image comprises:
determining the one or more correction factors based at least on an estimated illuminant color of the image;
blending the one or more alternative correction factors with the one or more correction factors to generate one or more blended correction factors; and
applying the one or more blended correction factors to the image.
10 . The method of claim 9 , wherein the estimated illuminant color is obtained based at least on applying a chromaticity weighting function to the image, and wherein the confidence level is an average weight determined using the chromaticity weighting function.
11 . The method of claim 9 , wherein the estimated illuminant color is determined based at least on normalizing the image, wherein the confidence level is based on a measure of reliability of the normalizing, and wherein the measure of reliability comprises an assessment of a number of valid pixels in the normalized image.
12 . The method of claim 9 , wherein the blending the one or more correction factors with the one or more alternative correction factors comprises:
determining one or more blending weights to be applied to the one or more correction factors and the one or more alternative correction factors based at least on a difference between the confidence level and the confidence threshold; and
blending the one or more correction factors with the one or more alternative correction factors according to the one or more blending weights.
13 . A system comprising:
one or more processing units to perform operations comprising:
determining one or more correction factors based at least on an estimated illuminant color of an image;
determining one or more blended correction factors by blending the one or more correction factors with one or more alternative correction factors according to one or more blending weights that are determined based on a difference between a confidence level of the estimated illuminant color and a confidence threshold; and
performing white balancing of the image based at least on applying the one or more blended correction factors to the image.
14 . The system of claim 13 , wherein the one or more processing units are further to perform operations comprising determining the one or more alternative correction factors based at least on an estimated scene brightness of the image.
15 . The system of claim 13 , wherein the one or more processing units are further to perform operations comprising determining the alternative correction factors based at least on an alternative estimated illuminant color of the image.
16 . The system of claim 13 , wherein the determining the one or more blended correction factors comprises:
determining the confidence level associated with the estimated illuminant color;
determining one or more blending weights to be applied to the one or more correction factors and the one or more alternative correction factors based at least on a difference between the confidence level and at least one of a low confidence threshold or a high confidence threshold; and
blending the one or more correction factors with the one or more alternative correction factors according to the one or more blending weights.
17 . The system of claim 13 , wherein the one or more processing units are further to perform operations comprising obtaining the estimated illuminant color based at least on applying a chromaticity weighting function to the image, and wherein the confidence level is determined based on a number or an amount of image pixels that remain after applying the chromaticity weighting function to the image.
18 . The system of claim 13 , wherein the one or more processing units are further to perform operations comprising obtaining the estimated illuminant color based at least on normalizing the image, wherein the confidence level is determined based on a measure of reliability of the normalizing, and wherein the measure of reliability comprises an assessment of a number of valid pixels in the normalized image.
19 . The system of claim 13 , wherein the one or more processing units are further to perform operations comprising:
determining one or more additional correction factors; and
based at least on a confidence level associated with the one or more alternative correction factors, determining one or more further blended correction factors by blending the one or more blended correction factors with the one or more additional correction factors, and
wherein the performing white balancing of the image is based at least on applying the one or more further blended correction factors to the image.
20 . The system of claim 13 , wherein the system is comprised in at least one of:
a control system for an autonomous or semi-autonomous machine;
a perception system for an autonomous or semi-autonomous machine;
a system for performing simulation operations;
a system for performing digital twin operations;
a system for performing light transport simulation;
a system for performing collaborative content creation for 3D assets;
a system for presenting one or more of virtual reality content, augmented reality content, or mixed reality content;
a system for real-time streaming applications;
a system for performing deep learning operations;
a system implemented using an edge device;
a system implemented using a robot;
a system for performing conversational AI operations;
a system for generating synthetic data;
a system incorporating one or more virtual machines (VMs);
a system implemented at least partially in a data center; or
a system implemented at least partially using cloud computing resources.