Hyperspectral image compression method based on non local and low-rank decomposition

Abstract

The invention discloses a hyperspectral image compression method based on non local and low-rank decomposition. Non local similarity is used for partitioning the hyperspectral image, hyperspectral small blocks with the same size are acquired; and then, according to the similarity, similar blocks are gathered in one class, four-order data formed by blocks in the same class can be subjected to low-rank decomposition. As the size of the decomposed data is far smaller than that of the original data, compression of the hyperspectral data is thus realized, and the spatial structure characteristics and the spectral structure characteristics of the hyperspectral data are made full use of.

Description

technical field [0001] The invention belongs to the technical field of computer image processing, and relates to an image compression and reconstruction method, in particular to a hyperspectral image compression method based on non-local and low-rank decomposition. Background technique [0002] Hyperspectral images contain multiple bands. Compared with grayscale images and RGB images, hyperspectral images contain more spectral information, so the accuracy of image processing can be greatly improved. Although hyperspectral images are used more and more with the maturity of hyperspectral imaging technology and the reduction of cost, there are still some limitations in hyperspectral images. [0003] 1) Multiple spectra of hyperspectral images provide more useful information, but also provide more redundant information, which greatly increases the time complexity and space complexity of image processing. [0004] 2) Hyperspectral images are often treated as multiple grayscale i...

AI Technical Summary

Problems solved by technology

[0004] 2) Hyperspectral images are often treated as multiple grayscale images in previous applications, and each spectrum can be used as a grayscale image. This processing process will lose spectral structure information

[0005] 3) When some algorithms process hyperspectral images, they not only need to process a single spectrum separately, but also pull the 2D image represented by each spectrum into a 1D vector. This processing method not only loses the spectral structure information of the image, but also destroys spatial structure information of a single grayscale image

Images