Image Restoration Method and System

Abstract

The invention provides a method for restoring an image. The method comprises the following steps that a matrix low-rank sparse decomposition model is built; the matrix low-rank sparse decomposition model is solved to obtain a decomposition result; the image is restored according to the matrix low-rank sparse decomposition model and the decomposition result. According to the method, when the image is damaged by errors seriously, an original image can still be restored accurately. The invention provides a system for restoring the image.

Description

technical field [0001] The present invention relates to the technical fields of mathematical models and algorithms and image / video processing, in particular to an image restoration method and system. Background technique [0002] Matrix low-rank sparse decomposition refers to decomposing an observation matrix into a low-rank matrix and a sparse matrix sum. It has a wide range of applications in signal processing, computer vision, image processing, and other research fields. [0003] In general, low-rank matrices involve data that are linearly related in nature, and sparse matrices are associated with large sparseness errors. Matrix low-rank sparse decomposition is also called low-rank matrix recovery, that is, recovering an error-free low-rank matrix from an observation matrix with large sparse errors. A low-rank sparse decomposition of a matrix can be modeled as follows, i.e., minimizing the weighted sum of the rank function of the data matrix and the (pseudo)zero-norm of...

AI Technical Summary

Problems solved by technology

However, when the structure of the matrix becomes more complex, the existing methods cannot achieve the desired effect

For example, when the rank of the data matrix is ​​relatively high, or the error distribution of the error matrix is ​​not so sparse, the principal component pursuit algorithm may not be successful in recovering the original data matrix from the observation matrix

Images