Hyperspectral image classification method for lightweight depth separable convolution feature fusion network

Abstract

The invention discloses a hyperspectral image classification method for a lightweight depth separable convolution feature fusion network, and the method comprises the steps: processing a hyperspectralimage, carrying out the normalization processing to obtain a sample set, carrying out the classification of the sample set, and completing the data preprocessing; setting a spectral information extraction module, a spatial information extraction module and a multi-layer feature fusion module to complete the construction of a training model; training the preprocessed convolutional neural network by using the constructed training model to obtain a final training result; repeating the operation of the convolutional neural network for N times, carrying out voting through N test results to obtaina final classification result, and carrying out hyperspectral image classification; and outputting a classification image according to the hyperspectral image classification result. According to the method, the spectral information and the spatial information are fused, the number of parameters is reduced, the network depth is increased, the network operation efficiency is improved, and the classification accuracy is improved.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a hyperspectral image classification method of a lightweight depth-separable convolutional feature fusion network. Background technique [0002] In recent years, with the development of hyperspectral remote sensing technology, the dimension of remote sensing data has been increasing, which has brought great challenges to the classification of hyperspectral data. Hyperspectral data has the characteristics of large data volume, correlation, multi-dimensionality, and nonlinearity. Selecting an effective algorithm for the classification of hyperspectral data has become an important issue in the analysis of hyperspectral remote sensing image data. According to the characteristics of deep learning, the theory and model of deep neural network are introduced in the hyperspectral image classification task, so that the rich information obtained through hyperspectral re...

AI Technical Summary

Problems solved by technology

How to make full use of the spatial spectrum information of hyperspectral images and reduce the parameters of hyperspectral image classification network to increase network depth is still a huge challenge

Images