Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer
Change detection extracts change areas in bitemporal remote sensing images, and plays an important role in urban construction and coordination. However, due to image offsets and brightness differences in bitemporal remote sensing images, traditional change detection algorithms often have reduced app...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/18/4524 |
| _version_ | 1797482856501477376 |
|---|---|
| author | Liegang Xia Jun Chen Jiancheng Luo Junxia Zhang Dezhi Yang Zhanfeng Shen |
| author_facet | Liegang Xia Jun Chen Jiancheng Luo Junxia Zhang Dezhi Yang Zhanfeng Shen |
| author_sort | Liegang Xia |
| collection | DOAJ |
| description | Change detection extracts change areas in bitemporal remote sensing images, and plays an important role in urban construction and coordination. However, due to image offsets and brightness differences in bitemporal remote sensing images, traditional change detection algorithms often have reduced applicability and accuracy. The development of deep learning-based algorithms has improved their applicability and accuracy; however, existing models use either convolutions or transformers in the feature encoding stage. During feature extraction, local fine features and global features in images cannot always be obtained simultaneously. To address these issues, we propose a novel end-to-end change detection network (EGCTNet) with a fusion encoder (FE) that combines convolutional neural network (CNN) and transformer features. An intermediate decoder (IMD) eliminates global noise introduced during the encoding stage. We noted that ground objects have clearer semantic information and improved edge features. Therefore, we propose an edge detection branch (EDB) that uses object edges to guide mask features. We conducted extensive experiments on the LEVIR-CD and WHU-CD datasets, and EGCTNet exhibits good performance in detecting small and large building objects. On the LEVIR-CD dataset, we obtain F1 and IoU scores of 0.9008 and 0.8295. On the WHU-CD dataset, we obtain F1 and IoU scores of 0.9070 and 0.8298. Experimental results show that our model outperforms several previous change detection methods. |
| first_indexed | 2024-03-09T22:38:32Z |
| format | Article |
| id | doaj.art-3e8f86dfb44643b4ad382532f1ffcd2b |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T22:38:32Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-3e8f86dfb44643b4ad382532f1ffcd2b2023-11-23T18:44:03ZengMDPI AGRemote Sensing2072-42922022-09-011418452410.3390/rs14184524Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a TransformerLiegang Xia0Jun Chen1Jiancheng Luo2Junxia Zhang3Dezhi Yang4Zhanfeng Shen5College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, ChinaCollege of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaCollege of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, ChinaCollege of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaChange detection extracts change areas in bitemporal remote sensing images, and plays an important role in urban construction and coordination. However, due to image offsets and brightness differences in bitemporal remote sensing images, traditional change detection algorithms often have reduced applicability and accuracy. The development of deep learning-based algorithms has improved their applicability and accuracy; however, existing models use either convolutions or transformers in the feature encoding stage. During feature extraction, local fine features and global features in images cannot always be obtained simultaneously. To address these issues, we propose a novel end-to-end change detection network (EGCTNet) with a fusion encoder (FE) that combines convolutional neural network (CNN) and transformer features. An intermediate decoder (IMD) eliminates global noise introduced during the encoding stage. We noted that ground objects have clearer semantic information and improved edge features. Therefore, we propose an edge detection branch (EDB) that uses object edges to guide mask features. We conducted extensive experiments on the LEVIR-CD and WHU-CD datasets, and EGCTNet exhibits good performance in detecting small and large building objects. On the LEVIR-CD dataset, we obtain F1 and IoU scores of 0.9008 and 0.8295. On the WHU-CD dataset, we obtain F1 and IoU scores of 0.9070 and 0.8298. Experimental results show that our model outperforms several previous change detection methods.https://www.mdpi.com/2072-4292/14/18/4524change detectionremote sensing imageimage edge detectiontransformer |
| spellingShingle | Liegang Xia Jun Chen Jiancheng Luo Junxia Zhang Dezhi Yang Zhanfeng Shen Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer Remote Sensing change detection remote sensing image image edge detection transformer |
| title | Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer |
| title_full | Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer |
| title_fullStr | Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer |
| title_full_unstemmed | Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer |
| title_short | Building Change Detection Based on an Edge-Guided Convolutional Neural Network Combined with a Transformer |
| title_sort | building change detection based on an edge guided convolutional neural network combined with a transformer |
| topic | change detection remote sensing image image edge detection transformer |
| url | https://www.mdpi.com/2072-4292/14/18/4524 |
| work_keys_str_mv | AT liegangxia buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer AT junchen buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer AT jianchengluo buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer AT junxiazhang buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer AT dezhiyang buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer AT zhanfengshen buildingchangedetectionbasedonanedgeguidedconvolutionalneuralnetworkcombinedwithatransformer |