Image Noise Removal via Uniform Color Space Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods struggle to accurately separate noise from edge structures in images, often resulting in reduced contrast and color loss, especially in high contrast areas, due to the difficulty in distinguishing noise components from brightness variations in non-uniform color spaces.
Innovation Solution
An image processing method that converts images into a work color space, extracts noise components, and then adjusts the noise removal based on the difference in gradation characteristics between the work and output color spaces, using techniques such as uniform noise spaces and contrast ratio functions to optimize noise removal while preserving edge structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If noise removal filtering is executed in the final output color space, then noise can be removed from the image, but edge structures are degraded and color fidelity is lost due to non-uniform noise distribution
Solution Approach 1:
The patent introduces a uniform noise space as an intermediary color space between the input image space and the final output color space. Noise removal filtering is executed in this uniform noise space where noise distribution is standardized, and then the processed image is converted back to the output color space. This intermediary space acts as a mediator that enables effective noise removal while preserving edge structures and color fidelity in the final output.
Solution Approach 2:
The patent transforms the image from the output color space to a uniform noise space by changing the parameter representation of colors. In this transformed space, noise distribution becomes uniform across different brightness levels, allowing consistent noise removal filtering. The parameters are then transformed back to the original color space, achieving noise removal without the degradation problems that occur when filtering is applied directly in the output color space.
2Reliability
If noise removal intensity is increased to remove more noise, then noise removal effectiveness improves, but contrast and saturation are reduced and color overlay occurs
Solution Approach 1:
By changing the parameter space to a uniform noise space, the patent enables noise removal filtering with standardized parameters that do not cause excessive smoothing. The uniform noise distribution in this space allows for optimal noise removal intensity that effectively removes noise while preserving contrast and color information, avoiding the color overlay and saturation loss that occur with traditional noise removal methods.
Data Source
AI summary
An image processing method for removing a noise component contained in an original image includes: extracting a noise component contained in an original image in a work color space; creating a noise-free image in the work color space based upon the extracted noise component and a difference between gradation characteristics in the work color space and gradation characteristics in an output color space; converting the noise-free image in the work color space to an image in the output color space.