Hyperspectral image spatial resolution enhancement method

Abstract

The present invention discloses a hyperspectral image spatial resolution enhancement method. The method comprises the steps of: extracting reflectance spectrums of different scenes in an image, and obtaining a spectrum dictionary with high sparsity and low coherence; employing the sparsity, the non-negativity and the space structure self-similarity of signals in the hyperspectral scenes, performing analysis from the extracted scene reflectance spectrums to obtain a low spectral image with a high spatial resolution, performing solving through a synchronous orthogonal matching tracking algorithmto obtain a sparse encoding matrix; and finally, combining the spectrum dictionary and the sparse encoding matrix to obtain a target image. Through joint use of the space and the spectral informationof the image, the simulation data and real data experiment result shows that, the method can effectively reconstruct the hyperspectral image detailed information and the texture structure compared toa traditional method and a matrix decomposition method so as to effectively improve the waveband average peak signal to noise ratio, the waveband average structure similarity and the spectrum angle mapping and better maintain the spectrum information.

Description

technical field [0001] The invention belongs to the field of photoelectric imaging detection and image processing, and relates to a new hyperspectral polarization image super-resolution enhancement method. Background technique [0002] Hyperspectral imaging technology has been an important part of comprehensive earth observation since the 1980s. It combines imaging technology with spectral technology, contains rich spatial, radiative and spectral information, and has the characteristics of continuous spectrum and unified spectrum. Hyperspectral images have been widely used in geological exploration, ocean monitoring, battlefield reconnaissance and other fields due to their good spectral characteristics. Also in medical imaging, spatially resolved spectral imaging obtained by hyperspectral imaging techniques provides diagnostic information on tissue physiology, morphology and composition. The spatial resolution of remote sensing images is an important indicator to measure th...

AI Technical Summary

Problems solved by technology

The above method improves the spatial resolution of the image to a certain extent, but there is no effective method to introduce spectral domain information to improve the reconstruction effect, and there is a problem of spectral information distortion

Images