White Balance Adjustment Using Surface Inclination Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques require manual segmentation of images into regions for white balance adjustment, which is time-consuming and may not accurately account for varying light sources, especially when the direction and intensity of flash light differ from assumed values.
Innovation Solution
An image processing apparatus determines a representative point in an image, obtains information about the inclination of surfaces including this point and pixels of interest, and adjusts white balance based on this information to automatically interpolate and set color temperature parameters for each pixel.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual region segmentation is performed for white balance adjustment, then white balance can be adjusted for different regions, but user time and effort increase significantly
Solution Approach 1:
The system performs automatic region segmentation and white balance adjustment without requiring manual user input. The determination unit automatically identifies regions with different light sources and calculates appropriate white balance coefficients, making the system self-sufficient and eliminating the need for manual segmentation operations.
Solution Approach 2:
The patent replaces manual mechanical segmentation operations with automated computational processing. Instead of users manually drawing region boundaries, the system uses image processing algorithms to automatically detect and segment regions based on lighting characteristics, substituting human operation with computational methods.
2Adaptability or versatility
If region segmentation is performed based on distance information, then white balance can be adjusted according to flash reach, but segmentation accuracy decreases when flash direction and intensity differ from assumptions
Solution Approach 1:
The system changes from using distance information alone to using multiple parameters including image luminance values and normal vector information. By incorporating luminance ratios and surface orientation data, the system adapts to varying flash directions and intensities, improving segmentation accuracy under different lighting conditions.
Solution Approach 2:
The determination unit uses feedback from image luminance values to refine region segmentation. By comparing actual luminance ratios with expected values and adjusting segmentation accordingly, the system adapts to real lighting conditions rather than relying on predetermined distance-based assumptions.
3Ease of manufacture
If uniform white balance adjustment is applied to the entire image, then processing is simple, but accuracy decreases when multiple light sources with different color temperatures are present
Solution Approach 1:
The patent divides the image into multiple regions based on lighting characteristics, with each region receiving appropriate white balance adjustment. The determination unit identifies boundaries between regions illuminated by different light sources, allowing differentiated processing that maintains accuracy while keeping individual region processing simple.
Solution Approach 2:
Instead of applying a single uniform white balance coefficient to the entire image, the system applies different white balance coefficients to different regions. Each region's coefficient is determined based on its specific lighting conditions, providing locally optimized adjustment that improves overall accuracy.
Data Source
AI summary
In order to provide a more preferable white balance adjustment, an image processing apparatus obtains an input image, determines a representative point in the input image and a white balance coefficient of the representative point, obtains first information representing the inclination of a surface that includes the representative point and second information representing the inclination of a surface that includes a pixel of interest in the input image, and adjusts white balance of the pixel of interest based on the first information, the second information, and the white balance coefficient.