Method for restoring degraded image based on L0 convex approximation

A technology for degrading images and images, applied in the fields of image processing and computer vision, which can solve the problems of slow convergence, poor accuracy, and artifacts.

Active Publication Date: 2016-11-23
杭州视熵科技有限公司
View PDF2 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the disadvantages of poor accuracy of existing degraded image restoration methods, avoid artifacts in the results due to the sensitivity of the solution method to noise, and slow convergence speed, the present inventio

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
  • Method for restoring degraded image based on L0 convex approximation
  • Method for restoring degraded image based on L0 convex approximation
  • Method for restoring degraded image based on L0 convex approximation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0056] The present invention is further described below.

[0057] A degraded image restoration method based on L0 convex approximation, comprising the following steps:

[0058] Given the original input blurred image y, the purpose of our algorithm is to estimate its blur kernel k and the corresponding clear image x on the premise that only y is known.

[0059] Construct an image pyramid according to the blur kernel size: According to the blur kernel size ks=41, the original blur map is down-sampled and the image is deconvolved at different scales. The first layer number is: That is, the downsampling is 7 scales, and the size of the blur kernel in each scale is: kslist=kslist+(kslist%2==0)=[7 9 11 15 21 29 41] (the size is generally required to be an odd number). The corresponding blurred image is also down-sampled to 7 scales, and the size of the i-th scale is

[0060] 1). Construct fitting items

[0061] Define the fitting term as: Where x and k represent the desir...

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 method for restoring a degraded image based on an L0 convex approximation. A new regular term is provided based on the sparse priori knowledge of an image; thus, the method can be used for quite well approximating to the L0 norm of an image gradient; moreover, the constraint on the L1 norm of an original image is increased; a half quadratic splitting method is adopted; meanwhile, the interactive iteration and the solving are carried out on formulas in a frequency domain at the same time; in a solving process, the validity of an approximately equivalent function is discussed and proved. By using the method for restoring the degraded image based on the L0 convex approximation, the extra wave filtration and plentiful iteration times are not needed; moreover, the convergence speed is quite quick; the method for restoring the degraded image based on the L0 convex approximation is applied to the restoration of the image; the processing time is short; the running speed of an algorithm is quick; moreover, a processing result is also better than that of an existing method.

Description

technical field [0001] The invention relates to the fields of computer vision, image processing and the like, more precisely, it relates to a blurred image restoration method based on sparse prior. Background technique [0002] In the process of image acquisition and transmission, affected by objective conditions (such as care, dust, etc.), it is easy to cause quality degradation and even affect the use. For example, the image is blurred, which is mainly caused by the shaking of the shooting equipment or the change of the scene. With the help of image processing methods, we can restore the image as much as possible to restore its original appearance to a certain extent, which is the main task of image restoration. [0003] Blurred image restoration is also called image deblurring. In most cases, it can also be called image deconvolution. It inverts the blurring process of clear images under certain prior knowledge, and at the same time transforms the inverse operation into ...

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
CPCG06T5/003
Inventor 刘盛宋洪章张少波陈宏峰陈胜勇
Owner 杭州视熵科技有限公司
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