Auto White Balance Fallback Using Chromaticity Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional auto white balancing methods struggle to accurately estimate illuminant color in images with large colored regions or objects, leading to unsatisfactory results, especially in scenarios where underlying statistics are unreliable.
Innovation Solution
Employ a chromaticity-based weighting function to exclude or minimize the impact of large colored regions or objects within an image when estimating illuminant color, using a weighting distribution defined around the Planckian Locus, and apply white balance correction factors based on this estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional auto white balancing methods are used, then the process is simple and fast, but color accuracy deteriorates in images with large colored regions or objects
Solution Approach 1:
The patent segments the image into multiple local regions and computes separate statistics for each region. By dividing the image into zones and evaluating illuminant color independently in each zone, the method avoids being dominated by large colored regions while maintaining processing efficiency through localized analysis rather than global processing.
Solution Approach 2:
The patent applies different processing strategies to different local regions based on their characteristics. By computing local statistics and identifying regions with reliable illuminant information versus those dominated by colored objects, the method selectively applies white balance corrections tailored to each region's properties, improving overall color accuracy without uniform complex processing.
2Reliability
If statistics from the entire image are used for white balance, then processing is efficient, but reliability deteriorates when large colored regions dominate the image
Solution Approach 1:
The patent divides the image into local regions and computes statistics separately for each region. This segmentation allows the system to identify and exclude regions dominated by colored objects from the global statistics calculation, improving reliability by ensuring that only regions with reliable illuminant information contribute to the overall white balance determination.
Solution Approach 2:
The patent computes statistics for multiple local regions beyond what a simple global average would require. By evaluating more regions than strictly necessary and then selecting or combining results based on reliability criteria, the method ensures robust white balance even when some regions are dominated by colored objects, while still maintaining reasonable processing efficiency.
Data Source
AI summary
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.