Hyperspectral image band selection method based on low-rank expression

Abstract

The invention discloses a hyperspectral image band selection method based on low-rank expression to mainly solve the problem that complexity of hyperspectral data processing is high and classification accuracy of hyperspectral images is low. The processing procedure comprises: (1) obtaining hyperspectral data and carrying out normalization processing on the data; (2) carrying out low-rank expression on the processed hyperspectral data; (3) solving low-rank expression coefficient by the use of augmented Lagrangian multiplier (ALM); (4) clustering bands according to the low-rank expression coefficient; (6) selecting a representative band from each cluster to be used as a final selected band; and (6) classifying the selected band. According to the invention, redundant information between bands is eliminated, and the band containing large amounts of information is selected. The selected band is more beneficial to classification. Classification accuracy of hyperspectral images is raised, and complexity of hyperspectral data processing is also reduced. The method provided by the invention can be used in dimensionality reduction of hyperspectral data.

Description

technical field [0001] The invention belongs to the technical field of image processing and relates to the selection of hyperspectral image bands, which can be applied to dimensionality reduction of hyperspectral data and reduce the computational complexity of data processing. Background technique [0002] Hyperspectral remote sensing technology is an important remote sensing technology developed in the 1920s. The hyperspectral imager can obtain almost continuous surface object spectrum images under the condition of multi-band and narrow spacing, which makes the hyperspectral image have higher spatial resolution and spectral resolution than the traditional remote sensing image, and is used in agriculture, Geology, coastal and inland water environment, atmospheric research, global environmental research and other fields have been widely used. However, hundreds or even thousands of bands also bring problems such as large data volume, "dimension disaster", information redundan...

AI Technical Summary

Problems solved by technology

However, it still results in the loss of important features

[0011] The above methods only consider the overall structure of the features, but do not fully consider the correlation between the features, so the optimal feature subset cannot be selected, resulting in low classification accuracy.

Images