A Compressed Sensing Reconstruction Method Based on Non-local Similarity of Images

Abstract

The invention puts forward a compressed sensing reconstruction method based on image nonlocal similarity. According to the method, image nonlocal similarity, a low-rank matrix and minimum total variation (TV) are combined, and two priors, namely, the local similarity and local smoothness of images, are fully utilized. On one hand, the block effect and the loss of global structure information caused by independent processing of single image blocks in the traditional method are eliminated, and on the other hand, real details of images are retained and false details produced by unreliable information are reduced or removed while noise is suppressed. High-quality compressed sensing reconstruction of images is realized. Compared with a general reconstruction method based on transform domain sparse or TV constraints, the method of the invention is of robustness to noise, better reconstruction quality is achieved, and great improvement is achieved both in visual effect and evaluation index.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a compressive sensing reconstruction method based on non-local similarity of images. Background technique [0002] The higher the resolution of the image, the richer the details it contains. High-resolution images are of great significance in practical applications. However, the massive data generated by high-resolution images limits the further improvement of image resolution to some extent. The traditional compression sampling method must obey the Nyquist sampling law, that is, the sampling data can completely retain the information of the original signal when the sampling frequency is greater than twice the highest frequency of the signal, and the compressed sensing theory breaks through the limitation of the Nyquist frequency , it can measure at a sampling rate much lower than Nyquist's theorem, and reconstruct the original signal exactly or approximately through a r...

AI Technical Summary

Problems solved by technology

The sparse prior and image TV prior based on the fixed domain have a good approximation effect for the signal with smooth slices, but the reconstruction effect for the image with rich texture information is not ideal, the texture feature will be smoothed during the reconstruction process, and it is possible generate false information

In 2006, Elad et al. took the lead in proposing a machine learning-based adaptive dictionary (i.e., sparse base) construction method, which uses an adaptive dictionary instead of a fixed sparse base. Although the sparsity of image blocks is fully considered, dictionary training is a big problem. Scale non-convex optimization problem, high computational complexity

Images