Luminance Correction Using Illumination Component Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods for luminance correction, such as the Retinex process, face challenges in performing good luminance correction without cumbersome processes, particularly in maintaining gradation of high-luminance parts and avoiding halo phenomena and dark-part noise, especially under biased illumination conditions like backlighting.
Innovation Solution
An image processing apparatus and method that includes an illumination light component estimation unit, an estimated illumination light component correction unit, a Retinex processing unit, and a gain adjustment unit, which corrects the estimated illumination light component to converge the luminance distribution into a uniform range, using a property where the increasing ratio of the corrected component decreases with higher values, and sets a clip luminance value based on histogram analysis to maintain gradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the luminance component is divided by the estimated illumination light component obtained by blurring, then luminance correction is performed, but halo phenomena and dark-part noise occur and gradation of high-luminance parts is lost
Solution Approach 1:
The patent applies different processing strategies to different regions of the image based on local characteristics. The clip luminance value is determined by analyzing the luminance distribution histogram, and pixels with luminance values above this clip level are processed differently to preserve gradation while preventing halo effects in high-luminance regions.
Solution Approach 2:
The patent modifies the parameter distribution of the corrected luminance component by introducing a clip luminance value threshold. This parameter change allows the system to maintain appropriate luminance correction in most regions while preserving the subtle gradations in high-luminance areas that would otherwise be lost or cause halo artifacts.
2Productivity
If gain correction is performed to associate the distribution with luminance signal range, then luminance range is adjusted, but optimal gain settings require cumbersome processes for each image
Solution Approach 1:
The system performs automatic luminance distribution analysis and clip luminance value determination without requiring manual intervention. The histogram analysis and automated threshold selection enable the processing apparatus to adapt to different images independently, eliminating the need for cumbersome manual gain settings while maintaining high processing efficiency.
3Stability of the object's composition
If the luminance component is corrected by dividing by the estimated illumination light component, then luminance distribution is uniformized, but gradation of high-luminance parts is lost
Solution Approach 1:
The patent preserves local quality characteristics in high-luminance regions by applying the clip luminance value threshold. Pixels with luminance values above the clip level maintain their original gradation relationships, preventing the loss of subtle tonal variations that would otherwise occur during uniform luminance correction.
Solution Approach 2:
Instead of applying uniform correction to all luminance values, the patent applies partial correction by selectively processing only those pixels below the clip luminance threshold. This partial action approach maintains sufficient luminance uniformity while preserving the essential gradation information in high-luminance regions.
Data Source
AI summary
An image processing apparatus capable of performing good luminance correction without performing a cumbersome process is provided. The image processing apparatus includes an illumination light component estimation unit which calculates an estimated illumination light component of an original image, an estimated illumination light component correction unit which outputs a corrected estimated illumination light component obtained by correcting the estimated illumination light component, and a Retinex processing unit which corrects a luminance component of the original image by dividing the luminance component by the corrected estimated illumination light component.