Image Denoising Method Based on External Nonlocal Self-similarity and Improved Sparse Representation

Abstract

The invention discloses an image denoising method based on external non-local self-similarity and improved sparse representation, comprising the following steps: (1) dividing the external clean image data set into block groups; (2) dividing the external clean image data sets into block groups External smooth block group and external texture block group; (3) learning the external smooth block group prior; (4) learning the external texture block group prior through Gaussian mixture model; (5) dividing the noisy image into block groups; (6) Using the external non-local self-similarity prior to guide the clustering of noisy image block groups, and calculate the smooth block group ratio of the noisy image; (7) optimize the regularization parameters of the sparse representation model by using the smooth block group ratio, according to the improved sparse representation model respectively Image blocks in each subspace are recovered. By optimizing the regularization parameters of the sparse representation model to increase its adaptability, it solves the technical problems of low peak signal-to-noise ratio and loss of detail information in existing image denoising methods.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image denoising method based on external non-local self-similarity and improved sparse representation. Background technique [0002] After decades of development, digital image processing technology has become more and more mature, and has played a great role in the fields of climate monitoring, medical diagnosis, face recognition and defect detection. Digital images will inevitably be disturbed by noise in the process of generation, transmission and conversion, and operations such as compression and feature extraction of images containing noise will cause degradation of image quality. Therefore, the problem of estimating the original clean image from the noisy image has become one of the important research contents. [0003] Image denoising algorithms are mainly divided into spatial domain methods and transform domain methods. The spatial domain methods use the corr...

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to address the above-mentioned problems in the prior art, to propose an image denoising method based on external non-local self-similarity and improved sparse representation, by optimizing the regularization parameters of the sparse representation model to increase its adaptability, in order to Solve the technical problems of low peak signal-to-noise ratio and loss of detail information in the existing image denoising methods

Images