Color Balance Correction via Object-Highlight Vector Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for color balance correction in digital images often result in errors due to incorrect detection of gray and skin color regions, leading to adverse effects such as incorrect light source estimation and subsequent color balance issues.
Innovation Solution
An image processing apparatus and method that detects objects and highlights in the image, calculates object and highlight colors, and applies color balance correction by adjusting the positional relationship between these colors on a color space to align with a reference vector, while also considering the reliability of the color information to minimize correction errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional color balance correction methods are used, then the process is simple, but the accuracy of light source estimation deteriorates due to erroneous detection of gray and skin color regions
Solution Approach 1:
The patent segments the color correction process into distinct modules: object detection unit, object color calculation unit, highlight color acquisition unit, correction value calculation unit, and color balance correction unit. This segmentation allows each module to specialize in a specific task, improving overall accuracy while maintaining manageable complexity.
Solution Approach 2:
The patent introduces a correction value calculation unit that acts as an intermediary between color detection and final correction. This unit calculates correction values based on the relative positional relationships between object colors and highlight colors, mediating the transformation to achieve accurate light source estimation without direct error propagation.
2Productivity
If conventional color balance correction is applied, then processing speed is fast, but color balance accuracy deteriorates leading to adverse effects
Solution Approach 1:
The patent performs preliminary object detection and color calculation before applying the final color balance correction. By pre-calculating object colors and highlight colors, and determining their relative positional relationships in advance, the system prepares accurate correction values that improve final color balance reliability without significantly increasing overall processing time.
Solution Approach 2:
The patent uses the relative positional relationship between object colors and highlight colors as feedback to determine correction values. This feedback mechanism allows the system to adjust correction amounts dynamically based on the specific color relationships in each image, improving accuracy while maintaining efficient processing through algorithmic optimization.
Data Source
AI summary
An image processing apparatus for applying a color balance correction to input image data, comprises a detection unit which detects an object included in the image data; an object color calculation unit which calculates an object color as a color of the object; an acquisition unit which acquires a highlight color in the image data; a calculation unit which calculates a correction value for the object color and the highlight color based on a relative positional relationship between the object color and the highlight color on a color space; and a color balance correction unit which applies the color balance correction to the highlight color and the object color using the correction value.