Cross-Hopping Graph Networks for Hyperspectral–High Spatial Resolution (H2) Image Classification
Remote Sensing · 27.08.2024 · 16 (17) · 3155
DOI: 10.3390/rs16173155 ↗Kurz gefasst
Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 27.08.2024. Verfasst von Tao Chen und 4 weiteren Autor:innen. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Innovation TeamFunds of ChinaWest Normal University, Sichuan Science and Technology Program of China.
Abstract (Originalsprache)
As we take stock of the contemporary issue, remote sensing images are gradually advancing towards hyperspectral–high spatial resolution (H2) double-high images. However, high resolution produces serious spatial heterogeneity and spectral variability while improving image resolution, which increases the difficulty of feature recognition. So as to make the best of spectral and spatial features under an insufficient number of marking samples, we would like to achieve effective recognition and accurate classification of features in H2 images. In this paper, a cross-hop graph network for H2 image classification(H2-CHGN) is proposed. It is a two-branch network for deep feature extraction geared towards H2 images, consisting of a cross-hop graph attention network (CGAT) and a multiscale convolutional neural network (MCNN): the CGAT branch utilizes the superpixel information of H2 images to filter samples with high spatial relevance and designate them as the samples to be classified, then utilizes the cross-hop graph and attention mechanism to broaden the range of graph convolution to obtain more representative global features. As another branch, the MCNN uses dual convolutional kernels to extract features and fuse them at various scales while attaining pixel-level multi-scale local features by parallel cross connecting. Finally, the dual-channel attention mechanism is utilized for fusion to make image elements more prominent. This experiment on the classical dataset (Pavia University) and double-high (H2) datasets (WHU-Hi-LongKou and WHU-Hi-HongHu) shows that the H2-CHGN can be efficiently and competently used in H2 image classification. In detail, experimental results showcase superior performance, outpacing state-of-the-art methods by 0.75–2.16% in overall accuracy.
Förderung
- National Natural Science Foundation of China
- Innovation TeamFunds of ChinaWest Normal University
- Sichuan Science and Technology Program of China
- A Ba Achievements Transformation Program
- Sichuan Province Transfer Payment Application and Development Program
Zitieren
APA
Chen, T., Wang, T., Chen, H., Zheng, B., & Deng, W. (2024). Cross-Hopping Graph Networks for Hyperspectral–High Spatial Resolution (H2) Image Classification. Remote Sensing, 16(17), 3155. https://doi.org/10.3390/rs16173155
Vancouver
Chen T, Wang T, Chen H, Zheng B, Deng W. Cross-Hopping Graph Networks for Hyperspectral–High Spatial Resolution (H2) Image Classification. Remote Sensing. 2024;16(17):3155. doi:10.3390/rs16173155
BibTeX
@article{Chen2024,
author = {Chen, Tao and Wang, Tingting and Chen, Huayue and Zheng, Bochuan and Deng, Wu},
title = {Cross-Hopping Graph Networks for Hyperspectral–High Spatial Resolution (H2) Image Classification},
journal = {Remote Sensing},
year = {2024},
volume = {16},
number = {17},
pages = {3155},
doi = {10.3390/rs16173155}
}
EUVIMED