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

Tensor repairing method based on non-local self-similarity and low-rank regularization of tensors

A repair method, self-similar technology, applied in the field of image processing, can solve the problem of ignoring non-local self-similar information and so on

Inactive Publication Date: 2019-09-10
李晓彤
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] Although methods that exploit tensor local smoothness priors have achieved good results, they ignore the redundant non-local self-similarity information in tensors

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
  • Tensor repairing method based on non-local self-similarity and low-rank regularization of tensors
  • Tensor repairing method based on non-local self-similarity and low-rank regularization of tensors
  • Tensor repairing method based on non-local self-similarity and low-rank regularization of tensors

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0058] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0059] At first the symbols used in the present invention are described:

[0060] In the present invention, a vector is represented by a lowercase letter (a), a matrix is ​​represented by an uppercase letter (A), and a swash letter is used to represent tensors. The following sections introduce some basic concepts and knowledge of tensors.

[0061] (1) Basic knowledge of tensor

[0062] Define a rank N tensor where its (i 1 , i 2 ,...,i N ) position element is defined as tensor The expansion matrix along the nth direction is defined as where the tensor in (i 1 , i 2 ,...,i N ) po...

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 tensor repairing method based on non-local self-similarity and low-rank regularization of tensors. The method comprises the following steps: S1, establishing a tensor model;S2, optimizing a target function according to the adjacent operator function, and solving the tensor model; and S3, carrying out iterative solution by utilizing a rank growth strategy. A plug-and-playframework is utilized to design a non-explicit non-local self-similarity rule to promote detail recovery of the tensor, a model solving algorithm based on a block continuous upper bound descent method is designed. Numerical experiment shows that the proposed model NLS-LR has obvious advantages in the aspects of structure, contour, details and the like of the recovery target tensor, and experimental results show that the visual effect and evaluation indexes of the model all exceed those of a plurality of existing mainstream methods.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a tensor restoration method based on tensor non-local self-similarity and low-rank regularization. Background technique [0002] In today's information society, information is exploding. In real life, data such as magnetic resonance images (MRI), hyperspectral images / multispectral images (HSI / MSI), color pictures and videos often have high-dimensional structures. As a generalization of vectors and matrices, tensors play a very important role in representing high-dimensional data. Due to the lack of information or the high cost of obtaining information, tensors in real life often show an incomplete structure. The problem of inferring a complete tensor from a missing tensor is called the tensor repair problem (LRTC). The tensor inpainting problem has a wide range of applications in reality, such as image inpainting, nuclear magnetic resonance image restoration, rain rem...

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
CPCG06T2207/10024G06T5/77G06T5/70
Inventor 李晓彤
Owner 李晓彤
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