Image Noise Reduction via Local Correlation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image noise reduction algorithms often fail to accurately distinguish between noise and non-noise pixels, leading to erroneous judgments and inadequate protection of image details, especially in high noise ratios.
Innovation Solution
An image noise reduction method based on local correlation, which calculates the number of similar pixels in a motion window and determines if the center pixel is on a 'thin line' to differentiate between noise and non-noise pixels, using either a 5×5 or 3×3 median filter for noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional median filters are used for noise reduction, then impulse noise is reduced, but image details are blurred and edge information is lost
Solution Approach 1:
The patent applies local quality by analyzing the statistical properties of pixels in different regions. It calculates the standard deviation of pixel values within a sliding window and uses this local statistical information to determine whether a pixel is noisy or contains important edge information. This allows the filter to adapt its behavior locally, preserving edges while removing noise in homogeneous regions.
Solution Approach 2:
The patent changes the parameter of filter strength dynamically based on local image characteristics. By computing the standard deviation of pixel values in a sliding window, the algorithm adjusts the filtering intensity adaptively - applying stronger filtering where noise is detected and weaker or no filtering where edges are present, thus resolving the contradiction between noise reduction and detail preservation.
2Reliability
If filtering strength is increased to reduce noise, then noise reduction improves, but image details and edges are lost
Solution Approach 1:
The patent implements feedback by continuously monitoring the standard deviation of pixel values in sliding windows during image processing. This statistical feedback information is used to dynamically adjust the filtering decision for each pixel, allowing the system to increase filtering strength when noise is detected through low standard deviation and reduce or stop filtering when edges are detected through high standard deviation, thus preventing information loss.
3Productivity
If simple threshold-based noise detection is used, then processing speed increases, but detection accuracy decreases due to erroneous judgments
Solution Approach 1:
The patent replaces simple mechanical threshold-based detection with a statistical approach using standard deviation calculation. Instead of relying on fixed thresholds that cause erroneous judgments, the system computes the standard deviation of pixel values in sliding windows, providing a more robust and accurate noise detection mechanism that adapts to local image characteristics while maintaining computational efficiency.
Data Source
AI summary
A method for reducing image noise includes calculating a first pixel amount of pixels that are similar to each other in a first number neighbor of a center pixel in a motion window, determining whether the first pixel amount of pixels that are similar to each other in the first number neighbor is greater than a first predetermined value, and using a mean of those pixels of the first pixel amount of pixels that are similar to each other in the first number neighbor to restore the center pixel of the motion window if the first pixel amount of pixels is greater than the first predetermined value. The method includes determining whether a second pixel amount of pixels that are similar to the center pixel is greater than a second predetermined value if the first pixel amount of pixels is not greater than the first predetermined value.