Image Processing Apparatus Noise Reduction Weight Derivation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The non-local means method for image noise reduction faces challenges in achieving effective noise reduction while retaining edge and texture, as it assigns excessively high weights to the target pixel when the reference region is the same as the target region, leading to insufficient noise reduction and image blurring.
Innovation Solution
An image processing apparatus that determines pixel values using a weighted mean of reference pixels, with a derivation unit calculating weights based on similarity between target and reference regions, and a weighted mean processing unit performing calculations using a pixel other than the target pixel as the reference, to address the issue of excessive weight assignment to the target pixel.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the non-local means method uses block matching to calculate similarity between target pixel and reference pixel, then noise reduction accuracy is improved, but the method produces insufficient noise reduction effect when the same region is set as both target and reference pixel
Solution Approach 1:
The patent extracts and removes the target pixel from the set of reference pixels used for weighted mean calculation. By separating the target pixel from potential reference pixels, the method prevents the target pixel from being used as its own reference, thereby avoiding the complete block match problem that causes insufficient noise reduction while maintaining the use of other similar regions for accurate noise reduction.
2Object-affected harmful factors
If simple noise reduction processing is applied to reduce noise, then noise level is reduced, but resolution deteriorates and edges/textures are lost
Solution Approach 1:
The patent applies local quality by using non-local mean filtering that selectively processes different regions with appropriate weights. The method calculates similarity between target pixel regions and reference pixel regions, assigning higher weights to more similar regions while preserving edges and textures through the similarity-based weighting mechanism, thus reducing noise while maintaining local image quality.
Data Source
AI summary
A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.