Image Tone Conversion Using Multi-Resolution Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems face challenges in efficiently converting bit width of image signals while preserving image quality, particularly due to high computational costs and inability to maintain spatial continuity, especially in regions like face areas.
Innovation Solution
An image processing system that acquires specific image signals from feature areas and performs tone conversion using multi-resolution decomposition and composition methods, along with correction coefficients calculated based on noise information, to optimize tone conversion processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If tone conversion is performed using texture information and area division, then image quality in specific regions is improved, but computational cost increases significantly
Solution Approach 1:
The patent applies local quality by detecting specific feature areas (such as faces) in the image and applying different tone conversion characteristics to these regions compared to the rest of the image. This allows optimization of image quality in important regions without unnecessarily processing the entire image with complex algorithms, thus reducing overall computational cost while maintaining high quality where needed.
2Productivity
If tone conversion is performed using local histogram equalization, then computational efficiency is improved, but adaptability to specific regions like face areas is reduced
Solution Approach 1:
The patent combines the efficiency of histogram equalization with the adaptability of feature-specific processing by first detecting feature areas (faces) and then applying tone conversion with optimized characteristics specifically for these regions. This maintains computational efficiency while achieving region-specific adaptability that standard histogram equalization lacks.
3Speed
If bit width conversion is performed using fixed tone characteristic, then processing speed is maintained, but image quality deterioration occurs due to cancellation of significant digits
Solution Approach 1:
The patent introduces dynamics by adapting the tone conversion characteristics based on the detected feature areas rather than using a fixed characteristic. When feature areas are detected, the system dynamically switches to feature-specific tone conversion parameters that preserve significant digits and maintain image quality, while maintaining processing speed through efficient algorithms.
Data Source
AI summary
A tone conversion unit, when performing tone conversion on an image signal, performs the tone conversion so as to provide a more appropriate tone to a feature area such as a face extracted from the image signal. Specifically, a correction coefficient calculation unit performs multi-resolution decomposition on a specific signal with the extracted feature area, and sets all pixels of low-frequency component to 1, while setting high-frequency components considered to include noise to 0, so as to calculate a correction coefficient. Then, a correction coefficient processing unit performs a processing of multiplying the correction coefficient to the image signal similarly performed with multi-resolution decomposition.