Spatial Multifeature and Dual-Layer Multihop Graph Convolution Networks for Hyperspectral Image Classification
Hyperspectral image (HSI) classification constitutes a crucial research direction within the domain of remote sensing. Convolutional neural networks (CNNs) and graph convolutional network (GCN) have exhibited outstanding classification performance in this field, emerging as current research focuses....
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11062325/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Hyperspectral image (HSI) classification constitutes a crucial research direction within the domain of remote sensing. Convolutional neural networks (CNNs) and graph convolutional network (GCN) have exhibited outstanding classification performance in this field, emerging as current research focuses. Nevertheless, GCN possesses certain limitations in capturing the neighborhood features of images, while traditional 2-D CNNs are incapable of fully extracting the spatial information of HSI. To address these problems, we propose a novel architecture dubbed spatial multifeature and dual-layer multihop graph convolutional network (SMTGCN). This network is capable of concurrently extracting pixel-level spatial features and superpixel-level spectral features. Specifically, a dual-layer multihop graph convolutional network is constructed within the GCN branch, which can take the features of superpixel at different segmentation scales as network nodes to effectively capture and fuse the superpixel features in HSI. In the CNN branch, a multiscale spatial structure is constructed for feature extraction and fusion, and a hybrid attention mechanism model is proposed to enhance the feature capture ability, a multilayer pooling structure is added to retain more detailed information while suppressing excessive redundant data. Finally, the features extracted by the GCN branch and the CNN branch are fused to realize HSI classification. Experimental results conducted on four benchmark HSI datasets indicate that, in comparison with existing classification methods, SMTGCN achieves remarkable improvements in classification performance when using a small number of training samples. |
|---|---|
| ISSN: | 1939-1404 2151-1535 |