Matching pursuit reconstruction method based on structured observation and direction guidance

Abstract

The invention discloses a matching pursuit reconstruction method based on structured observation and direction guidance for mainly solving the problems of inaccurate compressed sensing image reconstruction and low robustness at low sampling rate. The implementation process is as follows: defining available sketching blocks and unavailable sketching blocks according to a sketching graph of an image; clustering the available sketching blocks adopting direction guidance based on sketching direction, an performing grey clustering on the unavailable sketching blocks; during observation, adopting structured observation expressed by sketching on smooth image blocks in the unavailable sketching blocks, and adopting observation of multiple observation vectors on unsmooth image blocks; and during reconstruction, adopting an OMP algorithm based on sketching expression and direction guidance to obtain a reconstructed image. Compared with the traditional observation OMP method, the matching pursuit reconstruction method has the advantages of high reconstructed image quality and good robustness, and the matching pursuit reconstruction method can be applied to the reconstruction of natural images.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a matching tracking reconstruction method based on structured observation and direction guidance, which can be used to reconstruct natural images. Background technique [0002] In recent years, a new data theory compressive sensing CS has emerged in the field of signal processing. This theory realizes compression while collecting data, breaks through the limitations of the traditional Nyquist sampling theorem, and brings new advantages to data collection technology. The revolutionary changes make the theory have broad application prospects in compressed imaging systems, military cryptography, wireless sensing and other fields. Compressed sensing theory mainly includes three aspects: sparse representation of signal, observation of signal and reconstruction of signal. [0003] There are three main categories of current CS reconstruction algorithms, including c...

AI Technical Summary

Problems solved by technology

Gaussian observation is a random observation method, which treats all image blocks equally and does not capture the inherent a priori characteristics of image blocks. At the same time, the corresponding OMP reconstruction method directly reconstructs the image according to the measurement, and does not have the image block The judgment of the structure, both of which lead to the lack of accuracy and robustness of the reconstruction at low sampling rates

Images