Image Denoising via Empirical Mode Decomposition and Anisotropic Diffusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image denoising methods using anisotropic diffusion equations are ineffective due to the sensitivity of gradient operators to noise, leading to inaccurate edge detection and noise generation of non-existent edges.
Innovation Solution
The proposed method involves performing empirical mode decomposition to blur images, followed by edge detection and the calculation of a diffusion threshold for an improved anisotropic diffusion equation, which guides the diffusion process to remove noise and enhance denoising effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If gradient operator is used for edge detection in anisotropic diffusion, then edge detection capability is provided, but noise sensitivity increases and spurious edges are generated
Solution Approach 1:
The patent applies preliminary blurring processing to the image before performing edge detection. This pre-processing step removes high-frequency noise components that would otherwise interfere with accurate edge detection, allowing the gradient operator to work on a cleaner image and avoid generating spurious edges from noise.
Solution Approach 2:
The patent segments the image processing into distinct frequency components through blurring, separating noise (high-frequency) from actual edge information (low-frequency). This segmentation allows edge detection to focus on meaningful structures while filtering out noise interference.
2Measurement precision
If edge detection is performed on original image, then real edges are detected, but noise generates false edges reducing detection accuracy
Solution Approach 1:
The patent applies preliminary blurring processing to the image before performing edge detection. This pre-processing step removes high-frequency noise components that would otherwise interfere with accurate edge detection, allowing the gradient operator to work on a cleaner image and avoid generating spurious edges from noise.
3Productivity
If diffusion processing is performed without improved diffusion threshold, then computational simplicity is maintained, but denoising effect is reduced due to noise-affected edge detection
Solution Approach 1:
The patent introduces a feedback mechanism where the diffusion threshold is dynamically adjusted based on edge detection results from the blurred image. The edge map obtained from the pre-processed image provides feedback to control the diffusion process, allowing the algorithm to preserve real edges while removing noise without requiring complex real-time adjustments during diffusion.
Data Source
AI summary
An image denoising method and system, and storage medium are disclosed, which relate to the technical field of image processing. The method including: performing blur processing on an image to be processed by means of empirical mode decomposition to obtain a blurred feature image; performing edge detection processing on the feature image to obtain an edge detection operator of the feature image; calculating a diffusion threshold of a preset anisotropic diffusion equation according to the edge detection operator, and determining an improved anisotropic diffusion equation according to the calculated diffusion threshold; and performing, by using the improved anisotropic diffusion equation, diffusion processing on the image to be processed to obtain denoised image information. With the above method, the impact of the image noise on edge detection results can be weakened and the image denoising effect can be improved.