Color Balancing Method Pruning Saturated Pixels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Imaging devices struggle to accurately maintain color balance under different illumination conditions, as they fail to replicate the human vision system's ability to preserve light source interactions, leading to unnatural color appearances in captured images.
Innovation Solution
A color balancing method that samples pixels in an RGB color space, prunes saturated regions, calculates channel gains based on gray world statistics, and applies these gains to produce a transformed color-balanced image, suitable for display, which can be implemented in hardware, software, or a combination of both.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional color balancing methods are used, then the imaging device can process images quickly, but the color balance becomes inaccurate under different illumination conditions
Solution Approach 1:
The patent changes the parameter of pixel selection by introducing a saturation threshold criterion. Pixels are selected based on their saturation level (excluding highly saturated pixels) in addition to the traditional gray world assumption, allowing the system to adapt to different illumination conditions while maintaining color balance accuracy
Solution Approach 2:
The patent implements a feedback mechanism where the color balancing process iteratively refines the estimation of illumination chromaticity. By using the selected pixels to compute color moments and comparing against the gray world assumption, the system adjusts its color balance parameters to achieve accurate results under varying lighting conditions
2Ease of manufacture
If all pixels are used for color balancing, then processing is simpler, but saturated regions distort the color balance
Solution Approach 1:
The patent extracts and excludes highly saturated pixels from the color balancing calculation by applying a saturation threshold. This removes the distorting influence of saturated regions while maintaining the simplicity of the overall processing approach
Solution Approach 2:
The patent applies different treatment to different regions of the image based on their saturation characteristics. Non-saturated pixels are used for color balance calculation while highly saturated pixels are excluded, creating a locally optimized approach that improves accuracy without significantly increasing complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Pixels from an image are sampled for gray world statistics. To avoid the effect of saturated regions, the pixels are pruned. If a predetermined percentage of the pixels are included in the gray world statistics, color channel gain is calculated and applied to the image. As a result, color balance is achieved in a simple and efficient manner.