Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Nonlocal Mean Denoising Optimization Method Based on Structural Similarity

A non-local mean and structural similarity technology, which is applied in image enhancement, image analysis, instruments, etc., can solve the problems of large difference in denoising performance with different noise intensities, poor image detail processing ability, strong noise denoising ability decline, etc.

Active Publication Date: 2020-05-12
WUHAN UNIV OF SCI & TECH
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The core problem of the non-local mean filtering algorithm is to determine the weighted kernel function. The original non-local mean denoising algorithm uses an exponential kernel function for weighting, which leads to excessive smoothing and blurring of image details, and the pure exponential or cosine kernel function It cannot adapt to the change of noise, and the denoising ability of strong noise is obviously reduced; the improved quadratic exponential kernel function makes the weighting value decrease rapidly with the increase of distance, resulting in poor processing ability of image details and different noise intensity The denoising performance varies greatly, etc.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Nonlocal Mean Denoising Optimization Method Based on Structural Similarity
  • A Nonlocal Mean Denoising Optimization Method Based on Structural Similarity
  • A Nonlocal Mean Denoising Optimization Method Based on Structural Similarity

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] A non-local mean denoising optimization method based on structural similarity. The specific steps of the method are:

[0065] Step 1. Select an image such as figure 1 The noise-contaminated image X shown, figure 1 It is a noise-contaminated image to be denoised in this embodiment. The size of the noise-contaminated image X is 256×256, and the noise standard deviation is 25. Let any pixel i,j∈X,(m,n) be the coordinates of any pixel, select a pixel i in the noise-polluted image X, and take the pixel i as the center to establish a 7×7 noise image search box.

[0066] Step 2. Take a 3×3 noise image similarity box X in the noise image search box t , with the noise image similarity box X t Swipe in the noise image search box to find all noise image similar boxes X in the noise image search box t A combination of similar boxes X recorded for each noisy image t The pixel point j in the center is in the noise image search box centered on the pixel point i, that is, the w...

Embodiment 2

[0119] A non-local mean denoising optimization method based on structural similarity. The specific steps of the method are:

[0120] Step 1. Select an image such as image 3 The noise-contaminated image X shown, image 3 It is a noise-contaminated image to be denoised in this embodiment. The size of the noise-contaminated image X is 256×256, and the noise standard deviation is 25. Let any pixel i,j∈X,(m,n) be the coordinates of any pixel, select a pixel i in the noise-polluted image X, and take the pixel i as the center to establish a 7×7 noise image search box.

[0121] Step 2. Take a 3×3 noise image similarity box X in the noise image search box t , with the noise image similarity box X t Swipe in the noise image search box to find all noise image similar boxes X in the noise image search box t A combination of similar boxes X recorded for each noisy image t The pixel point j in the center is in the noise image search box centered on the pixel point i, that is, the w...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a non-local mean denoising optimization method based on structural similarity. The technical scheme is characterized by obtaining a noise pollution image X, and calculating a weighted value w(i,j) of all pixel points j with respect to a pixel point i; carrying out primary denoising on the noise pollution image X through a weight kernel function improved non-local means method to obtain a primarily-denoised image X<^>; in the primarily-denoised image X<^>, calculating weighted value wSSIM(i,j) of all primarily-denoised image similar boxes X<^>j with respect to d*d primarily-denoised image similar boxes X<^>i with the pixel point i being as the center; and adjusting all of noise image similar boxes in the noise pollution image X, and carrying out secondary denoising on the noise pollution image X to obtain a secondarily-denoised image X<->. The method has the advantages of being capable of keeping image details, suitable for different noise intensity and good in visual quality.

Description

technical field [0001] The invention belongs to the technical field of non-local mean denoising optimization. Specifically, it relates to a non-local mean denoising optimization method based on structural similarity. Background technique [0002] Image denoising is the most basic and widely studied hot issue in the field of image processing, and its purpose is to remove all kinds of noise pollution introduced in the process of image acquisition and transmission. Among many image denoising algorithms, the original non-local mean algorithm proposed by Buades et al. has been shown to outperform other classical denoising methods. The basic idea of ​​the algorithm is to use a large amount of redundant information in the image to perform a global search for the image block where each pixel in the image is located to find blocks that are similar to it, and to reduce random noise through the weighted average of similar structural blocks. The effect of denoising. [0003] The core...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06T5/00
CPCG06T5/002G06T2207/10004G06T2207/20021G06T2207/20024
Inventor 柴利张璐盛玉霞
Owner WUHAN UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com