Hybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification
Remote Sensing · 16.01.2023 · 15 (2) · 526
DOI: 10.3390/rs15020526 ↗Kurz gefasst
Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 16.01.2023. Verfasst von Zheng Fang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 18-mal zitiert (Crossref) und nennt 54 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key Research and Development Program of China, National Natural Science Foundation of China, Postgraduate Research and Practice Innovation Program of Jiangsu Province.
Abstract (Originalsprache)
Recently, methods based on convolutional neural networks (CNNs) achieve superior performance in polarimetric synthetic aperture radar (PolSAR) image classification. However, the current CNN-based classifiers follow patch-based frameworks, which need input images to be divided into overlapping patches. Consequently, these classification approaches have the drawback of requiring repeated calculations and only relying on local information. In addition, the receptive field size in conventional CNN-based methods is fixed, which limits the potential to extract features. In this paper, a hybrid attention-based encoder–decoder fully convolutional network (HA-EDNet) is presented for PolSAR classification. Unlike traditional CNN-based approaches, the encoder–decoder fully convolutional network (EDNet) can use an arbitrary-size image as input without dividing. Then, the output is the whole image classification result. Meanwhile, the self-attention module is used to establish global spatial dependence and extract context characteristics, which can improve the performance of classification. Moreover, an attention-based selective kernel module (SK module) is included in the network. In the module, softmax attention is employed to fuse several branches with different receptive field sizes. Consequently, the module can capture features with different scales and further boost classification accuracy. The experiment results demonstrate that the HA-EDNet achieves superior performance compared to CNN-based and traditional fully convolutional network methods.
Förderung
- National Key Research and Development Program of China
- National Natural Science Foundation of China
- Postgraduate Research and Practice Innovation Program of Jiangsu Province
- Aeronautical Science Foundation of China
- Science and Technology Program of Shenzhen, China
- Open Research Program of Shanghai Key Lab of Intelligent Information Processing
- Fujian Provincial Key Lab of Coastal Basin Environment
- Key Laboratory of Southeast Coast Marine Information Intelligent Perception and Application, MNR
Zitieren
APA
Fang, Z., Zhang, G., Dai, Q., Xue, B., & Wang, P. (2023). Hybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification. Remote Sensing, 15(2), 526. https://doi.org/10.3390/rs15020526
Vancouver
Fang Z, Zhang G, Dai Q, Xue B, Wang P. Hybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification. Remote Sensing. 2023;15(2):526. doi:10.3390/rs15020526
BibTeX
@article{Fang2023,
author = {Fang, Zheng and Zhang, Gong and Dai, Qijun and Xue, Biao and Wang, Peng},
title = {Hybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification},
journal = {Remote Sensing},
year = {2023},
volume = {15},
number = {2},
pages = {526},
doi = {10.3390/rs15020526}
}
EUVIMED