Image Processing Apparatus Shine Correction via Chrominance Signal Ratio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques, such as those described in Japanese Patent No. 5867390, often erroneously correct areas with steep brightness changes, like the eyes, due to relying solely on the difference between the input image and low frequency components for reflection component extraction, leading to unnecessary corrections.
Innovation Solution
An image processing apparatus with a detection unit for identifying specific object areas, a correction processing unit that reduces luminance and adjusts chrominance signal ratios, and an extraction unit for peripheral color information, allowing for controlled correction intensity based on chrominance signal ratio comparisons.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reflection component extraction is performed based only on the difference between the input image and low frequency component, then the shine area can be detected and corrected, but areas with steep brightness changes (such as the eyes) are erroneously detected and corrected
Solution Approach 1:
The patent segments the image processing into multiple components: low frequency component extraction, reflection component calculation, and peripheral area color extraction. By dividing the correction process and applying different processing to different areas (using the peripheral area chrominance as a reference), it distinguishes true shine areas from areas with steep brightness changes, thereby improving detection accuracy while preventing erroneous corrections.
2Productivity
If correction intensity is uniformly applied to all high-luminance pixels, then processing is simple and fast, but the quality of correction varies in different areas
Solution Approach 1:
The patent applies local quality by extracting chrominance signal ratios from peripheral areas and using them to determine area-specific correction parameters. Different regions of the image receive customized correction intensities based on their local chrominance characteristics, ensuring consistent correction quality across different areas while maintaining processing efficiency through automated parameter determination.
Data Source
AI summary
A reflection correction processing unit extracts at least color information from a peripheral area of a processing target pixel having a high luminance in an input image, and performs control, to the processing target pixel, to reduce the luminance of the processing target pixel and to bring a chrominance signal of the processing target pixel close to a predetermined chrominance signal ratio. In a case where a chrominance signal ratio based on the color information is close to the predetermined chrominance signal ratio, the reflection correction processing unit increases an intensity of the correction to the processing target pixel.