A Hyperspectral Image Unmixing Method Based on Weighted Joint Sparse Regression

Abstract

The invention provides a high-spectrum image unmixing method based on weighted joint sparse regression. According to the method, a problem of weighted joint sparse regression is solved so that all the mixed image elements of a high-spectrum image can be unmixed simultaneously. In the unmixing process, the weight value used by next iteration is obtained by current solving calculation. Compared with non-weighted norm, sparsity of an abundance matrix can be further enhanced by the weighted norm, and thus unmixing precision can be enhanced. The experiment results indicate that the high-spectrum image unmixing method has characteristics of high speed and high precision.

Description

technical field [0001] The invention provides a hyperspectral image unmixing method based on weighted joint sparse regression, which belongs to the field of hyperspectral image analysis. Background technique [0002] Hyperspectral remote sensing plays an important role in the observation of the earth's land, ocean, and atmosphere. However, due to the low spatial resolution of hyperspectral imaging spectrometers and the complex distribution of ground features, there are a large number of mixed pixels in hyperspectral images. In order to make full use of the information in hyperspectral image data, it is necessary to decompose each mixed pixel into a set of several components (also called end members) and their relative proportions (also called abundance). [0003] In the past two decades, many hyperspectral image unmixing algorithms have been proposed, most of them are based on the linear mixture model, which assumes that the spectral signal of the hyperspectral image collect...

AI Technical Summary

Problems solved by technology

However, due to the relatively low spatial resolution of current imaging spectrometers and the existence of mixing phenomena at various scales, the assumption of the existence of pure pixels is usually not valid (M. Iordache,, J.M.Bioucas-Dias, and A. Plaza, Collaborative Sparse Regression for Hyperspectral Unmixing, IEEE Trans. Geosci. Remote Sens., vol.52, no.1, pp.341-354, Jan.2014.)

Images