Hyperspectral image classification method based on double-branch convolution auto-encoder

Abstract

The invention provides a hyperspectral image classification method based on a double-branch convolution auto-encoder. The method mainly solves the problem of low classification precision caused by insufficient feature extraction of the hyperspectral data in the prior art, and comprises the following steps: obtaining a supervised training sample set, a supervised training sample label vector set, an unsupervised training sample set and a test sample set of the hyperspectral data; constructing and constructing a double-branch convolution auto-encoder; iterative training is carried out on the double-branch convolution auto-encoder; and obtaining a hyperspectral image classification result. The invention discloses a classification method based on a double-branch convolution auto-encoder. Two modes of unsupervised learning and supervised learning are comprehensively utilized to perform feature extraction on the hyperspectral data, data information of the hyperspectral image is fully considered, the classification precision of the hyperspectral image is effectively improved, and the method can be used for distinguishing and distinguishing ground features in the fields of agricultural monitoring, geological exploration, disaster environment evaluation and the like.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to a hyperspectral image classification method, in particular to a hyperspectral image classification method based on a double-branch convolutional autoencoder, which can be used to classify hyperspectral image features. Background technique [0002] A spectral image with a spectral resolution within the range of 101 is called a hyperspectral image, and its most notable feature is its rich spectral information. Compared with ordinary two-dimensional images, hyperspectral data can be represented as a hyperspectral data cube, which is a three-dimensional data structure. While ensuring spatial resolution, hyperspectral data also contains a large number of spectral dimension features, so it can provide spatial domain information and spectral domain information, and has the characteristics of "integration of graph and spectrum". Based on these characteristics, hyperspectral image...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the shortcomings of the above-mentioned prior art, and propose a hyperspectral image classification method based on a dual-branch convolutional autoencoder, which is used to solve the technical problem of low classification accuracy in the prior art

Images