Row and column loss image filling method based on low-rank matrix reconstruction and sparse representation

Abstract

The invention belongs to the field of computer vision, and aims at achieving the accurate filling of a pixel row and column loss image. The technical scheme employed in the invention is that a row and column loss image filling method based on low-rank matrix reconstruction and sparse representation comprises the steps: introducing low-rank priori to carry out the constraining of a potential image based on the low-rank matrix reconstruction theory; introducing a separable two-dimensional sparse priori based on the sparse representation theory with the consideration to a condition that the sparse representation of each column of a row loss image can be achieved through a column dictionary and the sparse representation of each row of a column loss image can be achieved through a row dictionary; specifically describing a filling problem of an image with the row and column loss into a solving problem of a constraint optimization equation based on the above combined low-rank and separable two-dimensional sparse priori, thereby achieving the filling of the row and column loss image. The method is mainly used for a computer vision processing occasion.

Description

technical field [0001] The invention belongs to the field of computer vision. In particular, it concerns row and column missing image filling methods based on low-rank matrix reconstruction and sparse representation. Background technique [0002] The problem of recovering an unknown complete matrix from a part of the known pixels of the matrix has attracted great attention in recent years. Such problems are frequently encountered in many application domains of computer vision and machine learning, such as image inpainting, recommender systems, and background modeling. [0003] There have been many research results on methods to solve the image filling problem. Due to the ill-conditioned nature of the matrix filling problem, current matrix filling methods generally consider the latent matrix to be low-rank or approximately low-rank, and then fill in the missing pixel values ​​through low-rank matrix reconstruction. Such as singular value threshold method (SVT), augmented L...

AI Technical Summary

Problems solved by technology

However, due to the lack of prior conditions, the problem of missing both rows and columns of the matrix has not yet been solved.

Images