Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network
Extracting difference features is a key technique for polarimetric synthetic aperture radar (PolSAR) image change detection. Although the current PolSAR change detection algorithms based on convolutional neural networks (CNNs) can capture the local information of difference features well, the global...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2023-11-01
|
Series: | International Journal of Applied Earth Observations and Geoinformation |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S1569843223003357 |
_version_ | 1797633288797421568 |
---|---|
author | Zhifei Yang Yan Wu Ming Li Xin Hu Zhikang Li |
author_facet | Zhifei Yang Yan Wu Ming Li Xin Hu Zhikang Li |
author_sort | Zhifei Yang |
collection | DOAJ |
description | Extracting difference features is a key technique for polarimetric synthetic aperture radar (PolSAR) image change detection. Although the current PolSAR change detection algorithms based on convolutional neural networks (CNNs) can capture the local information of difference features well, the global structure information cannot be extracted effectively, resulting in low detection accuracy. In this paper, we propose a graph-context attention-based siamese encoder–decoder network (GCA-SEDN) for unsupervised change detection in PolSAR images. The GCA-SEDN can mine local and global polarization features simultaneously. Firstly, based on the local features extracted by CNN, the feature optimization graph attention (FOGA) module is constructed to capture global features of PolSAR images. At the same time, the FOGA module greatly refines the image structure feature representation and extracts more discriminative features. Secondly, the designed context-aware dilated pyramid (CADP) module uses multiple dilated group convolutional layers to further extract deep data features with different receptive fields. The obtained multi-scale context data features can adapt well to change targets of different sizes. Finally, by considering both the reconstruction error of the dual-branch encoder–decoder network and the pixel-level classification error, a new hybrid loss function is constructed so that the GCA-SEDN can fully learn change features, thus effectively improving the accuracy of label prediction. Experiments on five real Gaofen-3 PolSAR datasets prove that the proposed GCA-SEDN is more competitive than other existing representative algorithms. |
first_indexed | 2024-03-11T11:51:59Z |
format | Article |
id | doaj.art-8474e1748e1c40c8990ce3e41aaf6068 |
institution | Directory Open Access Journal |
issn | 1569-8432 |
language | English |
last_indexed | 2024-03-11T11:51:59Z |
publishDate | 2023-11-01 |
publisher | Elsevier |
record_format | Article |
series | International Journal of Applied Earth Observations and Geoinformation |
spelling | doaj.art-8474e1748e1c40c8990ce3e41aaf60682023-11-09T04:11:37ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322023-11-01124103511Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention networkZhifei Yang0Yan Wu1Ming Li2Xin Hu3Zhikang Li4Remote Sensing Image Processing and Fusion Group, School of Electronic Engineering, Xidian University, Xi’an 710071, ChinaRemote Sensing Image Processing and Fusion Group, School of Electronic Engineering, Xidian University, Xi’an 710071, China; Corresponding author.National Key Laboratory of Radar Signal Processing, Xidian University, Xi’an 710071, ChinaRemote Sensing Image Processing and Fusion Group, School of Electronic Engineering, Xidian University, Xi’an 710071, ChinaRemote Sensing Image Processing and Fusion Group, School of Electronic Engineering, Xidian University, Xi’an 710071, ChinaExtracting difference features is a key technique for polarimetric synthetic aperture radar (PolSAR) image change detection. Although the current PolSAR change detection algorithms based on convolutional neural networks (CNNs) can capture the local information of difference features well, the global structure information cannot be extracted effectively, resulting in low detection accuracy. In this paper, we propose a graph-context attention-based siamese encoder–decoder network (GCA-SEDN) for unsupervised change detection in PolSAR images. The GCA-SEDN can mine local and global polarization features simultaneously. Firstly, based on the local features extracted by CNN, the feature optimization graph attention (FOGA) module is constructed to capture global features of PolSAR images. At the same time, the FOGA module greatly refines the image structure feature representation and extracts more discriminative features. Secondly, the designed context-aware dilated pyramid (CADP) module uses multiple dilated group convolutional layers to further extract deep data features with different receptive fields. The obtained multi-scale context data features can adapt well to change targets of different sizes. Finally, by considering both the reconstruction error of the dual-branch encoder–decoder network and the pixel-level classification error, a new hybrid loss function is constructed so that the GCA-SEDN can fully learn change features, thus effectively improving the accuracy of label prediction. Experiments on five real Gaofen-3 PolSAR datasets prove that the proposed GCA-SEDN is more competitive than other existing representative algorithms.http://www.sciencedirect.com/science/article/pii/S1569843223003357PolSAR imageUnsupervised change detectionSiamese encoder–decoder frameworkGraph attention moduleMulti-scale context information |
spellingShingle | Zhifei Yang Yan Wu Ming Li Xin Hu Zhikang Li Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network International Journal of Applied Earth Observations and Geoinformation PolSAR image Unsupervised change detection Siamese encoder–decoder framework Graph attention module Multi-scale context information |
title | Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network |
title_full | Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network |
title_fullStr | Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network |
title_full_unstemmed | Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network |
title_short | Unsupervised change detection in PolSAR images using siamese encoder–decoder framework based on graph-context attention network |
title_sort | unsupervised change detection in polsar images using siamese encoder decoder framework based on graph context attention network |
topic | PolSAR image Unsupervised change detection Siamese encoder–decoder framework Graph attention module Multi-scale context information |
url | http://www.sciencedirect.com/science/article/pii/S1569843223003357 |
work_keys_str_mv | AT zhifeiyang unsupervisedchangedetectioninpolsarimagesusingsiameseencoderdecoderframeworkbasedongraphcontextattentionnetwork AT yanwu unsupervisedchangedetectioninpolsarimagesusingsiameseencoderdecoderframeworkbasedongraphcontextattentionnetwork AT mingli unsupervisedchangedetectioninpolsarimagesusingsiameseencoderdecoderframeworkbasedongraphcontextattentionnetwork AT xinhu unsupervisedchangedetectioninpolsarimagesusingsiameseencoderdecoderframeworkbasedongraphcontextattentionnetwork AT zhikangli unsupervisedchangedetectioninpolsarimagesusingsiameseencoderdecoderframeworkbasedongraphcontextattentionnetwork |