Image restoration method and system based on equality structure multiple regularization

Abstract

The invention discloses an image restoration method and system based on equality structure multiple regularization, and the method comprises the steps: introducing an equality structure, and constructing a measurement Gaussian likelihood function; selecting a plurality of sparse transformations based on the constructed measurement Gaussian likelihood function, and giving a plurality of mutually independent probability density functions based on sparse transformation coefficients to calculate the conditional prior probability density of the target image; based on the condition prior probability density of the target image, calculating a mean value estimation model of the target image according to a Bayesian rule; and finally, achieving fast reconstruction of the target image by combination of an expectation maximization method and a conjugate gradient method iteration so that the reversibility limitation of sparse representation introduced by a comprehensive method is overcome, sparse transformation of any number is allowed to be adopted, and the convergence speed of the algorithm can be effectively improved through the joint sparse domain.

Description

technical field [0001] The invention belongs to the field of image restoration, in particular to an image restoration method and system based on multiple regularization of an equation structure. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] As a sparse reconstruction technique, compressed sensing (CS) is widely used in many fields, such as radar and medical imaging, seismic imaging, image processing, etc. According to the CS theory, under the premise that the signal meets the sparsity requirement, the original signal can be reconstructed with high precision by using a sampled data volume much smaller than that required by the Nyquist-Shannon sampling law. But what is even more striking is that CS also has excellent performance in high-precision reconstruction and estimation of signals. In order to balance the reconstruction accuracy of...

AI Technical Summary

Problems solved by technology

Single sparse transformation can be realized by modifying the linear model through the comprehensive method, but for irreversible sparse transformation, or to introduce multiple sparse transformations, the comprehensive method cannot be used

Therefore, in the field of complex image restoration algorithms, the multiple regularization BCS method with a complete theoretical basis and wide applicability has not yet appeared.

Images