Hyperspectral Image Classification Method Based on Spatial-Spectral Joint of Deep Convolutional Neural Networks

Abstract

The present invention relates to a hyperspectral image classification method based on a space-spectrum combination of a deep convolutional neural network, which introduces the traditional deep convolutional neural network applied to two-dimensional images into the three-dimensional hyperspectral image classification problem. First, use a small amount of labeled data to train the convolutional neural network, and use the network to autonomously extract the spatial spectral features of the hyperspectral image without any compression and dimensionality reduction; then, use the extracted spatial spectral features to train the support vector machine ( SVM) classifier to classify the image; finally, in combination with the trained neural network and the trained classifier, the neural network extracts the spatial spectral features of the target to be classified, and the classifier determines the specific category of the extracted spatial spectral features, A structure (DCNN‑SVM) that can autonomously extract the spatial spectral features of hyperspectral images and classify them is obtained, thus forming a set of hyperspectral image classification methods.

Description

technical field [0001] The invention belongs to the technical field of remote sensing information processing, and relates to a hyperspectral image classification method, in particular to a hyperspectral image classification method based on a space-spectrum combination of a deep convolutional neural network. Background technique [0002] Hyperspectral remote sensing images have high spectral resolution, multiple imaging bands, and large amounts of information, and are widely used in remote sensing applications. Hyperspectral image classification technology plays an important role in these applications. The features used for classification are extracted from the original hyperspectral image. This step has a great impact on the classification accuracy of hyperspectral images. Improve the classification accuracy; on the contrary, the classification features with poor robustness will significantly reduce the classification effect. [0003] In recent years, deep learning has made...

AI Technical Summary

Problems solved by technology

[0004] However, the existing methods for extracting spatial spectral features of hyperspectral images using deep models are very complicated. First, it is often necessary to reduce the dimensionality of the original hyperspectral image in spectral space, and then The information after dimensionality reduction is combined with the spectral information to obtain the spatial spectral feature

Dimensionality reduction is computationally intensive, and certain spectral information is lost, affecting accuracy

Images