Compressed spectral imaging method based on nonlinear compressed sensing and dictionary learning

Abstract

The invention discloses a compressed spectral imaging method based on nonlinear compressed sensing and dictionary learning, mainly to solve the problem that during a compressed sampling process in the prior art, a learned dictionary and sparse coefficients have negative values. The realization process comprises steps: signals in the original space are projected to feature space, a non-negative condition is introduced, and a non-negative kernel tracking algorithm and a non-negative matrix decomposition method are used for carrying out dictionary learning in the feature space; the learnt dictionary is used in a nonlinear compressed sensing model, and sparse coefficients are solved via the non-negative kernel tracking algorithm; and finally, a pre-image method is used for restoring the original signals. As is shown by an experimental result, under different sampling rates, compared with other existing dictionary learning methods, the reconstruction effects are the best, and the method can be used for capturing a remote sensing image.

Description

technical field [0001] The invention belongs to the technical field of signal processing, in particular to a compressed spectrum imaging method, which can be used for remote sensing image acquisition. Background technique [0002] Compressed sensing is a new sampling theory developed in the field of image processing technology in recent years. By using the sparse characteristics of the signal, it can realize the accurate recovery of information under the condition of much smaller than the traditional Nyquist sampling rate. At present, most of the compressed sensing is done under the linear model, because the sparse representation of the signal under the linear model is simple and intuitive. From the initial orthogonal base dictionary to the current dictionary learning, a large number of related researchers have used various methods to try to find a more suitable description of the transformation space, but they always stay in the linear model, so the development is slow. Ho...

AI Technical Summary

Problems solved by technology

Although this method can achieve a better reconstruction effect with a lower sampling rate and fewer measurement values, due to the negative value of the learned dictionary, there are information errors and Information loss, the original signal cannot be fully expressed, so that the reconstructed signal has a larger visual error, a smaller PSNR, and a poorer recovery effect than the original signal.

Images