MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection

Remote sensing (RS) image change detection (CD) aims to identify and localize the differences in the same geographical area observed across bitemporal RS images. Existing CD methods typically focus on characterizing and distinguishing change regions, while often overlooking the nonchange areas (back...

Full description

Saved in:
Bibliographic Details
Main Authors: Andong Huang, Chuan Xu, Liye Mei, Zhiwei Ye, Wei Yang, Ying Wang, Xinghua Li
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/10972294/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850255734108848128
author Andong Huang
Chuan Xu
Liye Mei
Zhiwei Ye
Wei Yang
Ying Wang
Xinghua Li
author_facet Andong Huang
Chuan Xu
Liye Mei
Zhiwei Ye
Wei Yang
Ying Wang
Xinghua Li
author_sort Andong Huang
collection DOAJ
description Remote sensing (RS) image change detection (CD) aims to identify and localize the differences in the same geographical area observed across bitemporal RS images. Existing CD methods typically focus on characterizing and distinguishing change regions, while often overlooking the nonchange areas (background information) in RS images. Therefore, the influence of subtle changes in nonchange regions may be amplified, leading to false negatives or false positives in the change regions. To address this issue, we propose a multiview complementary learning network (MVCL) for RS image CD, which considers CD from both background view and change view perspectives, while focusing on the extraction and aggregation of both change and nonchange (background) features. Specifically, we introduce a background branch during the encoder stage dedicated to extracting background information. This background information is then injected into the change regions through the background-dependent feature fusion module, serving as prior knowledge. By leveraging the stable background information, irrelevant and noisy changes are suppressed, guiding the generation of change regions and reducing the interference of background noise. Furthermore, to mitigate the noise introduced during multiscale fusion in the nonchange (background) areas, we propose a prior-guided spatial background enhancement module, which consists of a prior semantic guidance strategy and a spatial background enhancement module. According to the results of experiments conducted on three public datasets, the proposed method outperforms several state-of-the-art (SOTA) methods in both quantitative metrics and visual effectiveness.
format Article
id doaj-art-ea9d041a094e405cb15e29925febf7db
institution OA Journals
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-ea9d041a094e405cb15e29925febf7db2025-08-20T01:56:48ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118128311284410.1109/JSTARS.2025.356165610972294MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change DetectionAndong Huang0https://orcid.org/0009-0007-0644-7080Chuan Xu1https://orcid.org/0000-0002-7099-7833Liye Mei2https://orcid.org/0000-0002-2555-9199Zhiwei Ye3https://orcid.org/0000-0002-1218-0681Wei Yang4https://orcid.org/0000-0002-2014-8120Ying Wang5https://orcid.org/0009-0007-3655-9702Xinghua Li6https://orcid.org/0000-0002-2094-6480School of Information Science and Engineering, Wuchang Shouyi University, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Computer Science, Hubei University of Technology, Wuhan, ChinaSchool of Information Science and Engineering, Wuchang Shouyi University, Wuhan, ChinaSchool of Information Science and Engineering, Wuchang Shouyi University, Wuhan, ChinaSchool of Remote Sensing Information Engineering, Wuhan University, Wuhan, ChinaRemote sensing (RS) image change detection (CD) aims to identify and localize the differences in the same geographical area observed across bitemporal RS images. Existing CD methods typically focus on characterizing and distinguishing change regions, while often overlooking the nonchange areas (background information) in RS images. Therefore, the influence of subtle changes in nonchange regions may be amplified, leading to false negatives or false positives in the change regions. To address this issue, we propose a multiview complementary learning network (MVCL) for RS image CD, which considers CD from both background view and change view perspectives, while focusing on the extraction and aggregation of both change and nonchange (background) features. Specifically, we introduce a background branch during the encoder stage dedicated to extracting background information. This background information is then injected into the change regions through the background-dependent feature fusion module, serving as prior knowledge. By leveraging the stable background information, irrelevant and noisy changes are suppressed, guiding the generation of change regions and reducing the interference of background noise. Furthermore, to mitigate the noise introduced during multiscale fusion in the nonchange (background) areas, we propose a prior-guided spatial background enhancement module, which consists of a prior semantic guidance strategy and a spatial background enhancement module. According to the results of experiments conducted on three public datasets, the proposed method outperforms several state-of-the-art (SOTA) methods in both quantitative metrics and visual effectiveness.https://ieeexplore.ieee.org/document/10972294/Additional branchchange detection (CD)prior-guidedremote sensing (RS)spatial background enhancement
spellingShingle Andong Huang
Chuan Xu
Liye Mei
Zhiwei Ye
Wei Yang
Ying Wang
Xinghua Li
MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Additional branch
change detection (CD)
prior-guided
remote sensing (RS)
spatial background enhancement
title MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
title_full MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
title_fullStr MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
title_full_unstemmed MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
title_short MVCL: Multiview Complementary Learning Network for Remote Sensing Image Change Detection
title_sort mvcl multiview complementary learning network for remote sensing image change detection
topic Additional branch
change detection (CD)
prior-guided
remote sensing (RS)
spatial background enhancement
url https://ieeexplore.ieee.org/document/10972294/
work_keys_str_mv AT andonghuang mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT chuanxu mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT liyemei mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT zhiweiye mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT weiyang mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT yingwang mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection
AT xinghuali mvclmultiviewcomplementarylearningnetworkforremotesensingimagechangedetection