EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Hyperspectral Image Classification Based on Multi-Scale Convolutional Features and Multi-Attention Mechanisms

Qian Sun, Guangrui Zhao, Xinyuan Xia, Yu Xie, Chenrong Fang, Le Sun, Zebin Wu, Chengsheng Pan

Remote Sensing · 16.06.2024 · 16 (12) · 2185

DOI: 10.3390/rs16122185 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 16.06.2024. Verfasst von Qian Sun und 7 weiteren Autor:innen. Die Arbeit wurde bisher 23-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, Natural Science Foundation of Jiangsu Province, Startup Foundation for Introducing Talent of NUIST.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Convolutional neural network (CNN)-based and Transformer-based methods for hyperspectral image (HSI) classification have rapidly advanced due to their unique characterization capabilities. However, the fixed kernel sizes in convolutional layers limit the comprehensive utilization of multi-scale features in HSI land cover analysis, while the Transformer’s multi-head self-attention (MHSA) mechanism faces challenges in effectively encoding feature information across various dimensions. To tackle this issue, this article introduces an HSI classification method, based on multi-scale convolutional features and multi-attention mechanisms (i.e., MSCF-MAM). Firstly, the model employs a multi-scale convolutional module to capture features across different scales in HSIs. Secondly, to enhance the integration of local and global channel features and establish long-range dependencies, a feature enhancement module based on pyramid squeeze attention (PSA) is employed. Lastly, the model leverages a classical Transformer Encoder (TE) and linear layers to encode and classify the transformed spatial–spectral features. The proposed method is evaluated on three publicly available datasets—Salina Valley (SV), WHU-Hi-HanChuan (HC), and WHU-Hi-HongHu (HH). Extensive experimental results have demonstrated that the MSCF-MAM method outperforms several representative methods in terms of classification performance.

Förderung

  • National Natural Science Foundation of China
  • Natural Science Foundation of Jiangsu Province
  • Startup Foundation for Introducing Talent of NUIST
  • Jiangsu Innovation & Entrepreneurship Group Talents Plan

Zitieren

APA

Sun, Q., Zhao, G., Xia, X., Xie, Y., Fang, C., Sun, L., Wu, Z., & Pan, C. (2024). Hyperspectral Image Classification Based on Multi-Scale Convolutional Features and Multi-Attention Mechanisms. Remote Sensing, 16(12), 2185. https://doi.org/10.3390/rs16122185

Vancouver

Sun Q, Zhao G, Xia X, Xie Y, Fang C, Sun L, et al. Hyperspectral Image Classification Based on Multi-Scale Convolutional Features and Multi-Attention Mechanisms. Remote Sensing. 2024;16(12):2185. doi:10.3390/rs16122185

BibTeX
@article{Sun2024,
  author = {Sun, Qian and Zhao, Guangrui and Xia, Xinyuan and Xie, Yu and Fang, Chenrong and Sun, Le and Wu, Zebin and Pan, Chengsheng},
  title = {Hyperspectral Image Classification Based on Multi-Scale Convolutional Features and Multi-Attention Mechanisms},
  journal = {Remote Sensing},
  year = {2024},
  volume = {16},
  number = {12},
  pages = {2185},
  doi = {10.3390/rs16122185}
}
Kostenlos starten →