Hyperspectral Image Migration Classification Method Based on Deep Joint Distribution Adaptation Network

Abstract

A hyperspectral image migration classification method based on a deep joint distribution adaptation network, the steps are: input the hyperspectral images of the source domain and the target domain, perform feature normalization and dimension unification; combine the hyperspectral images of the source domain and the target domain Features; construct a marginal probability distribution adaptation network to adapt the marginal probability distribution of source domain and target domain hyperspectral images; according to the principle of one-to-many classification, select training samples of source domain and a small number of target domain hyperspectral images; construct The conditional probability distribution adaptation network is used to adapt the conditional probability distribution of the hyperspectral images in the source domain and the target domain; and perform one-to-many classification on the hyperspectral images in the target domain. The present invention proposes an adaptation network based on a deep joint distribution, which realizes the feature adaptation of the hyperspectral image in the source domain and the target domain, and reduces the joint probability distribution difference between the two; at the same time, it adopts a one-to-many classification model to improve the classification Intra-class and inter-class discrimination, thereby improving the accuracy of hyperspectral image migration classification.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, in particular to a hyperspectral image migration classification based on a deep joint distribution adaptation network. Background technique [0002] Hyperspectral images have high spectral resolution, wide band coverage, and rich spectral details of ground objects, which are conducive to fine ground object analysis. Hyperspectral image classification is an important part of hyperspectral image interpretation, and it is widely used in mineral exploration, vegetation survey, agricultural monitoring and other fields. Due to the large amount of hyperspectral image data and redundancy, the same target has spectral differences in different data, which affects the classification effect. [0003] Hyperspectral image classification is the process of analyzing and classifying the acquired target spectral signal and other features, and mainly adopts the method of supervised classifi...

AI Technical Summary

Problems solved by technology

The above method does not analyze the joint probability distribution of the hyperspectral image in the source domain and the target domain, and does not effectively use the joint probability distribution of the data to further improve the migration classification effect

Images