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

Image denoising method and apparatus based on global and local prior cooperative constraints

An image and global technology, applied in the field of image denoising and device based on global and local prior joint constraints, can solve the problems of incomplete restoration of image fine structure and low image signal-to-noise ratio, etc., to remove image noise and improve image quality. The effect of the signal-to-noise ratio

Active Publication Date: 2018-11-06
NORTHWEST UNIV(CN)
View PDF9 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide an image denoising method and device based on global and local prior joint constraints, to solve the problem of incomplete recovery of image fine structure and image information in the prior art when denoising a single image. low noise ratio

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
  • Image denoising method and apparatus based on global and local prior cooperative constraints
  • Image denoising method and apparatus based on global and local prior cooperative constraints
  • Image denoising method and apparatus based on global and local prior cooperative constraints

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0068] An image denoising method based on joint global and local prior constraints, which is used to denoise a single image Y to be processed, such as figure 1 shown, including the following steps:

[0069] Step 1, using formula I to establish a denoising model for the single image to be processed Y:

[0070]

[0071] Among them, X is the estimated pure image of the image Y to be processed, ||·|| w,* is the weighted kernel norm, Z represents the kernel Wiener filter image of image Y, λ WNN is the parameter of the set weighted kernel norm regularization term, λ KWF is the parameter of the kernel Wiener filter regularization term, λ WNN >0,λ KWF >0;

[0072] The observation model of the traditional image to be processed is usually expressed as:

[0073] Y=X+η Formula II

[0074] Among them, Y is the image to be processed, X is the clean image to be estimated, and η represents the noise.

[0075] In order to restore the pure image X to be estimated from the image Y to ...

Embodiment 2

[0134] This embodiment provides an image denoising device based on global and local prior joint constraints, which is used to denoise a single image Y to be processed, such as figure 2 As shown, the device includes:

[0135] The denoising model building module adopts the denoising model of formula I to remove noise from the image Y to be processed to obtain a denoising image

[0136]

[0137] Among them, X is the estimated pure image of the image Y to be processed, ||·|| w,* is the weighted kernel norm, Z represents the kernel Wiener filter image of image Y, λ WNN is the parameter of the set weighted kernel norm regularization term, λ KWF is the parameter of the kernel Wiener filter regularization term, λ WNN >0,λ KWF >0;

[0138] Decomposition module for rewriting Equation I into Equation III by using the scaled dual variable using the method of alternating direction multipliers:

[0139]

[0140] Among them, M is an auxiliary variable, ρ>0, and A is a Lagrangi...

Embodiment 3

[0159] In this embodiment, the commonly used pictures in the standard image test set are selected to test the method proposed in the present invention. Three-dimensional block matching (BM3D) algorithm and weighted kernel norm minimization denoising (WNNM) algorithm are the existing classic traditional denoising methods respectively. The deep convolutional neural network residual learning denoising (DnCNN) algorithm and the convolutional neural network-based fast and flexible image denoising (FFDNet) algorithm are currently the most advanced convolutional neural network-based denoising algorithms. The experimental results are shown in the table 1.

[0160] image 3 Pure image for Barbara, Figure 4-6 For the denoising method provided by the present invention when the noise is 25, 50, 75 image 3 Effect picture of denoising; Figure 7 For C-man pure image, Figure 8-10 For the denoising method provided by the present invention when the noise is 25, 50, 75 Figure 7 Image ...

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 discloses an image denoising method and apparatus based on global and local prior cooperative constraints. The image denoising method based on global and local prior cooperative constraints is provided, through comprehensive utilization of self-similarity prior of the global structure and the nonlinear mapping based on a local kernel function, being able to effectively recover the fine structure of the image while removing noise, and further improving the signal-to-noise ratio and the subjective objective quality of the image.

Description

technical field [0001] The invention relates to an image denoising method and device, in particular to an image denoising method and device based on global and local prior joint constraints. Background technique [0002] Image denoising is a method to recover the underlying clean image from the noisy observed image, and it has been an important problem in the field of image processing and computer vision. In the prior art, the denoising method based on the partial differential equation is to model the image denoising problem as a partial differential equation with an anisotropic diffusion process. The non-local mean method calculates the weight of neighboring pixels according to the self-similarity of the natural image, and uses the weighted average method to estimate the pure image. The method based on wavelet transform is to use forward wavelet transform to decompose the input image at multiple scales, then attenuate the wavelet coefficients at each scale to suppress nois...

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 Applications(China)
IPC IPC(8): G06T5/00
CPCG06T2207/20021G06T2207/20024G06T5/70
Inventor 康睿文章勇勤彭进业祝轩李展王珺许鹏飞郑霞
Owner NORTHWEST UNIV(CN)
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