A Transformer-Based Multiscale Difference Enhancement Network for Change Detection

Change detection (CD) is an important research field in remote sensing, aimed at identifying differences in multitemporal images. Despite the progress made by convolutional neural networks and Transformer architectures in visual analysis, challenges remain in achieving robust feature representation...

Full description

Saved in:
Bibliographic Details
Main Authors: Mengyang Pan, Hang Yang, Chengkang Yu, Mingqing Li, Anping Deng
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/10876581/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850215400724234240
author Mengyang Pan
Hang Yang
Chengkang Yu
Mingqing Li
Anping Deng
author_facet Mengyang Pan
Hang Yang
Chengkang Yu
Mingqing Li
Anping Deng
author_sort Mengyang Pan
collection DOAJ
description Change detection (CD) is an important research field in remote sensing, aimed at identifying differences in multitemporal images. Despite the progress made by convolutional neural networks and Transformer architectures in visual analysis, challenges remain in achieving robust feature representation and global contextual understanding. To address these issues, we propose a transformer-based multiscale difference enhancement network (TMDENet). Our approach utilizes a multiscale feature extraction module to capture diverse feature representations, while incorporating a channel-spatial cooperation mechanism for refined detail enhancement. The extracted features are encoded into semantic tokens, which are processed by a Transformer to capture long-range dependencies. This is further complemented by a cross-hierarchical linear fusion module for multiscale feature integration. Additionally, a difference enhancement module isolates and emphasizes change-related features. Extensive evaluations on benchmark datasets (LEVIR, CDD, DSIFN, WHUCD) show that TMDENet achieves state-of-the-art results in boundary delineation and change localization. This study establishes TMDENet as a robust framework for high-resolution remote sensing CD, offering significant improvements in both precision and reliability.
format Article
id doaj-art-e08b65477c024ec2b608b476226f3de9
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-e08b65477c024ec2b608b476226f3de92025-08-20T02:08:38ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118137781379310.1109/JSTARS.2025.353941610876581A Transformer-Based Multiscale Difference Enhancement Network for Change DetectionMengyang Pan0https://orcid.org/0009-0009-9143-9759Hang Yang1https://orcid.org/0000-0001-6027-1337Chengkang Yu2https://orcid.org/0009-0006-2874-8039Mingqing Li3Anping Deng4https://orcid.org/0000-0002-0419-7176Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, ChinaChangchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun, ChinaChange detection (CD) is an important research field in remote sensing, aimed at identifying differences in multitemporal images. Despite the progress made by convolutional neural networks and Transformer architectures in visual analysis, challenges remain in achieving robust feature representation and global contextual understanding. To address these issues, we propose a transformer-based multiscale difference enhancement network (TMDENet). Our approach utilizes a multiscale feature extraction module to capture diverse feature representations, while incorporating a channel-spatial cooperation mechanism for refined detail enhancement. The extracted features are encoded into semantic tokens, which are processed by a Transformer to capture long-range dependencies. This is further complemented by a cross-hierarchical linear fusion module for multiscale feature integration. Additionally, a difference enhancement module isolates and emphasizes change-related features. Extensive evaluations on benchmark datasets (LEVIR, CDD, DSIFN, WHUCD) show that TMDENet achieves state-of-the-art results in boundary delineation and change localization. This study establishes TMDENet as a robust framework for high-resolution remote sensing CD, offering significant improvements in both precision and reliability.https://ieeexplore.ieee.org/document/10876581/Change detection (CD)channel-spatial feature cooperation (CSFC)convolution neural network (CNN)remote sensing imageTransformer
spellingShingle Mengyang Pan
Hang Yang
Chengkang Yu
Mingqing Li
Anping Deng
A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Change detection (CD)
channel-spatial feature cooperation (CSFC)
convolution neural network (CNN)
remote sensing image
Transformer
title A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
title_full A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
title_fullStr A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
title_full_unstemmed A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
title_short A Transformer-Based Multiscale Difference Enhancement Network for Change Detection
title_sort transformer based multiscale difference enhancement network for change detection
topic Change detection (CD)
channel-spatial feature cooperation (CSFC)
convolution neural network (CNN)
remote sensing image
Transformer
url https://ieeexplore.ieee.org/document/10876581/
work_keys_str_mv AT mengyangpan atransformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT hangyang atransformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT chengkangyu atransformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT mingqingli atransformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT anpingdeng atransformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT mengyangpan transformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT hangyang transformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT chengkangyu transformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT mingqingli transformerbasedmultiscaledifferenceenhancementnetworkforchangedetection
AT anpingdeng transformerbasedmultiscaledifferenceenhancementnetworkforchangedetection