Improving Accuracy and Reducing Complexity of Sparse Image Reconstruction Using Structural Prior Constraint

Abstract

A method for improving sparse image reconstruction accuracy and reducing complexity by using structured prior constraints, belonging to the field of remote sensing image compression technology, is characterized in that it is a method of introducing image information on the basis of a multi-observation vector model to improve accuracy and reduce complexity , including the following steps in turn: input observation matrix, matrix composed of multi-observation vectors, number of wavelet transform layers, number of wavelet coefficients in each layer; initialize system model parameters, image wavelet coefficients, maximum number of iterations, precision performance index; use Markov-based The Gibbs sampling of the chain Monte Carlo method reconstructs the wavelet coefficients of the image, and then performs wavelet inverse transform on the reconstructed wavelet coefficient matrix to obtain the reconstructed image. The present invention can apply the required accuracy performance index to the actual reconstructed image, utilize the multi-observation vector model to reduce the complexity, and at the same time adjust the reconstruction accuracy by adjusting the number of wavelet transform layers, so that the quality of the reconstructed image meets the performance index requirements .

Description

technical field [0001] The invention belongs to the field of signal processing, can be applied to sparse reconstruction of image information, and particularly relates to a low-complexity structured reconstruction method based on a multi-observation vector model. Background technique [0002] According to the Compressive Sensing (CS) theory, if the signal satisfies a certain sparsity in a certain transform domain, the signal can be sampled at a rate much lower than the Nyquist sampling rate, and the signal can be accurately reproduced. structure. At present, compressed sensing technology has applications in many engineering fields, such as sub-Nyquist sampling systems, compressed imaging systems, compressed sensing networks, etc. In particular, compressed sensing technology can also be applied to the problem of sparse image reconstruction, and has been widely used in important fields such as medical imaging. [0003] However, most of the existing compressed sensing algorith...

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to solve the problems of high storage complexity and time complexity and limited reconstruction accuracy faced by sparse reconstruction of high-resolution images

Images