Image compressed sensing reconstruction method based on approximate message passing and double-low-rank constraint

Abstract

The invention relates to the technical field of image processing, in particular to an image compressed sensing reconstruction method based on approximate message passing and double-low rank constraint, which comprises the following steps of: estimating an Onsager correction item in residual updating of an approximate message passing algorithm by adopting a Monte Carlo method, calculating a noisy image, and performing similar image block clustering on the noisy image, and performing low-rank correction on the similar image blocks, performing low-rank correction on the residual error, and averaging the overlapped image blocks to obtain a reconstructed image. The approximate message passing algorithm is adopted to solve the reconstruction problem, and compared with a common iterative threshold algorithm, the compressed sensing reconstruction effect is more ideal. According to the non-local self-similarity of the natural image, similar image block low-rank constraint is adopted, and the problem that the natural image does not have wavelet domain or gradient domain sparse constraint is solved. In addition, residual low-rank constraint is used on the basis of similar image block low-rankconstraint, so that the reconstruction quality is improved, and the problem of partial region blurring when the similarity of the image similar blocks is not high is effectively solved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image compression sensing reconstruction method based on approximate message passing and double low-rank constraints. Background technique [0002] Compressed sensing is a new signal processing method that has emerged in recent years and is now being used more and more widely. Compressed sensing can break through the constraints of the Shannon-Nyquist sampling theorem, sample at a rate much smaller than twice the Nyquist bandwidth, and realize signal sampling and compression at the same time, obtain observation values ​​through dimensionality reduction sampling, and use reconstruction Algorithms to accurately restore the original signal have received high attention and applications in medical imaging, wireless communication, radar detection and other fields. Compressed sensing mainly includes three parts: sparse representation, nonlinear measurement, and image recons...

AI Technical Summary

Problems solved by technology

In the current image reconstruction algorithms, the approximate message passing algorithm based on sparse signals often uses sparse constraints in the wavelet domain or gradient domain, while natural images do not have such sparse properties but have non-local self-similarity, and the general iteration used The threshold algorithm compressive sensing reconstruction effect is not ideal, there is a problem of blurring in some areas in the process of image reconstruction, and its calculation load is also large

Images