Video denoising method based on scale mixing model and low-rank approximation

A hybrid model and video technology, applied in image data processing, instrumentation, computing, etc., can solve problems such as lack of consideration of the correlation between before and after frames, video blurring, and affecting video processing.

Inactive Publication Date: 2016-11-23
西安电子科技大学昆山创新研究院 +1
View PDF4 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] These two types of methods lack the consideration of the correlation between the front and back frames, and do not fully utilize the existing information in the video, so the useful texture information in the video will be lost, resulting in local blurring of the video, which will affect the subsequent video processing.

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
  • Video denoising method based on scale mixing model and low-rank approximation
  • Video denoising method based on scale mixing model and low-rank approximation
  • Video denoising method based on scale mixing model and low-rank approximation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] Referring to the accompanying drawings, the technical solutions and effects of the present invention are described in detail below:

[0041] refer to figure 1 , the implementation steps of the present invention are as follows:

[0042] Step 1. Perform median filter preprocessing on the input noisy video sequence y to obtain the initial estimated video y 0 .

[0043] (1a) For noisy video y={y 1 ,...,y t ,...,y T}∈R M×N×T Noisy image y in the tth frame t ∈ R M×N The i-th pixel in Perform a median filter, that is, to Take an image block x of size n as the center i,t , take image block x i,t The median value of internal pixels is Me, if The absolute value of is greater than or equal to the preset threshold τ, it is considered that the pixel Impulse noise, its pixel value after median filtering Otherwise, consider the pixel Not affected by impulse noise, the pixel value remains unchanged after median filtering, where t∈{1,2,...,T}, T represents the numbe...

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 a video denoising method based on a scale mixing model and low-rank approximation, and mainly aims to solve the problem that Gaussian pulse mixed noise is hard to be eliminated accurately in the prior art. The video denoising method comprises the following steps: 1, acquiring an initial estimation of a video by adopting a median value filtering method, and searching a similar image block matrix in front and back frames of a test image; 2, modeling an abnormal point set by using the Laplace scale mixing model to convert the abnormal point estimation problem into a problem of joint solution for abnormal points and hidden factors so as to eliminate the abnormal points caused by the mixed noise; 3, performing low-rank approximation on the similar image block matrix, and calculating a denoised image by using a nonlocal low-rank model; 4, performing iterative calculating through the Laplace scale mixing model and the nonlocal low-rank model to obtain a recovered single-frame image; 5, repeating the steps 1 to 4 to obtain a denoised video. The video denoising method disclosed by the invention can eliminate the mixed noise and retain detailed information of the image, is good in visual effect, and can be used for denoising a video media, a remote-sensing image and a medical image.

Description

technical field [0001] The invention relates to the field of video denoising, in particular to a video denoising method based on a scale mixture model and low-rank approximation, which can be applied to the fields of video multimedia, remote sensing images, medical images, and the like. Background technique [0002] Video sequences will inevitably receive noise interference during storage and transmission, which will be directly related to subsequent video processing applications, such as target tracking, target recognition, video compression, etc. Therefore, in video processing, video denoising plays an important role. very important role. The video acquisition process will introduce Gaussian noise, while dead pixels or transmission errors in the camera equipment will introduce impulse noise. Impulse noise can be divided into salt and pepper noise and random value noise. The pixel value affected by salt and pepper noise is 0 or 255, while the pixel value affected by random...

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
IPC IPC(8): G06T5/00
CPCG06T2207/20032G06T2207/10016G06T5/73G06T5/70
Inventor 董伟生石光明黄韬
Owner 西安电子科技大学昆山创新研究院
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap