White Balance Control Using Light Source Loci to Reduce Color Failure

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional white balance control methods fail to accurately detect light sources in images with chromatic-colored subjects, leading to color failures where the object color is toned down and becomes overly complementary green, especially in environments with dominant green colors like forests or shaded areas.

Innovation Solution

A white balance control method that divides images into areas, corrects color information based on light source loci in a color space, calculates white balance correction values, and adjusts white balance accordingly, using first and second light source loci to minimize the influence of object colors and accurately detect the light source.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If white balance control is performed based on color information from the entire image, then the control process is simple, but the light source cannot be accurately detected when chromatic-colored subjects dominate the image

Engineering Contradiction:

Improvesimplicity of white balance control processVSAvoidaccuracy of light source detection

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The image is divided into multiple divided areas, and color information is obtained for each area separately. This segmentation allows the system to analyze color distribution across different regions and identify areas that represent the actual light source rather than being dominated by chromatic subject colors.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different processing approaches are applied to different regions of the image based on their color characteristics. Areas with color information close to light source loci are identified and given different weight or processing compared to areas dominated by subject colors, allowing localized analysis that improves light source detection accuracy.

Inventive Principle:

Principle #3Local quality

2Productivity

If conventional white balance control is applied in environments with dominant green colors, then the control is performed quickly, but the light source is mistaken and color failure occurs

Engineering Contradiction:

Improvespeed of white balance controlVSAvoidaccuracy of white balance control

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

Light source loci are pre-calculated and stored in the color space before actual image processing. These reference loci represent expected color signatures of various light sources. During white balance control, the system compares actual color information against these pre-established references, enabling fast and reliable light source identification even in challenging green-dominated environments.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses the distribution of color information across multiple divided areas to feedback and determine the actual light source type. By analyzing the aggregate color data and comparing it against light source loci, the system can identify the true light source and adjust white balance accordingly, preventing color failures while maintaining efficient processing.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS7643068B2White balance control method, white balance control apparatus and image-taking apparatus

Publication Date: 2010.01.05 FUJIFILM CORP

AI summary

According to the present invention, it is possible, by using the first and second light source loci, to tone down the influence of the object color of the color information in the image and detect the light source based on the color information having the influence of the object color toned down so as to detect the light source accurately. It is thereby possible to reduce the color failure when controlling the white balance.