DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module
Extracting road networks from remote sensing images holds critical implications for various applications including autonomous driving, path planning, and road navigation. Despite its importance, the task remains arduous due to the complex backdrops in remote sensing imagery, intricate road geometrie...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10315115/ |
| _version_ | 1827742499957899264 |
|---|---|
| author | Debin Wei Pinru Li Hongji Xie Yongqiang Xu |
| author_facet | Debin Wei Pinru Li Hongji Xie Yongqiang Xu |
| author_sort | Debin Wei |
| collection | DOAJ |
| description | Extracting road networks from remote sensing images holds critical implications for various applications including autonomous driving, path planning, and road navigation. Despite its importance, the task remains arduous due to the complex backdrops in remote sensing imagery, intricate road geometries, and the challenges arising from vegetation and structural obstructions. To address these multifaceted issues, we introduce a specialized model for road extraction in remote sensing images, termed DRCNet. This model employs a pre-trained DenseNet-121 as its encoder and is fortified with both Recurrent Criss-Cross Attention (RCCA) and Convolutional Block Attention Module (CBAM). RCCA facilitates the capture of global contextual information across all pixels, thereby enriching the model’s understanding of global image relationships. Simultaneously, CBAM is integrated within the skip connections to optimize the network’s focus on significant road features. Comprehensive experiments conducted on both the DeepGlobe Road and Massachusetts Road datasets substantiate that DRCNet outperforms other benchmark models in road detection tasks. |
| first_indexed | 2024-03-11T04:02:49Z |
| format | Article |
| id | doaj.art-8b2c12126dd84a5dba42ee4086961672 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-11T04:02:49Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-8b2c12126dd84a5dba42ee40869616722023-11-18T00:01:00ZengIEEEIEEE Access2169-35362023-01-011112687912689110.1109/ACCESS.2023.333212110315115DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention ModuleDebin Wei0https://orcid.org/0000-0002-8950-5341Pinru Li1https://orcid.org/0009-0003-6443-5206Hongji Xie2https://orcid.org/0009-0000-7054-9080Yongqiang Xu3https://orcid.org/0009-0007-0951-2116Communication and Network Laboratory, Dalian University, Dalian, ChinaCommunication and Network Laboratory, Dalian University, Dalian, ChinaCommunication and Network Laboratory, Dalian University, Dalian, ChinaCommunication and Network Laboratory, Dalian University, Dalian, ChinaExtracting road networks from remote sensing images holds critical implications for various applications including autonomous driving, path planning, and road navigation. Despite its importance, the task remains arduous due to the complex backdrops in remote sensing imagery, intricate road geometries, and the challenges arising from vegetation and structural obstructions. To address these multifaceted issues, we introduce a specialized model for road extraction in remote sensing images, termed DRCNet. This model employs a pre-trained DenseNet-121 as its encoder and is fortified with both Recurrent Criss-Cross Attention (RCCA) and Convolutional Block Attention Module (CBAM). RCCA facilitates the capture of global contextual information across all pixels, thereby enriching the model’s understanding of global image relationships. Simultaneously, CBAM is integrated within the skip connections to optimize the network’s focus on significant road features. Comprehensive experiments conducted on both the DeepGlobe Road and Massachusetts Road datasets substantiate that DRCNet outperforms other benchmark models in road detection tasks.https://ieeexplore.ieee.org/document/10315115/Road extractionremote sensing imageDenseNet-121 networkattention module |
| spellingShingle | Debin Wei Pinru Li Hongji Xie Yongqiang Xu DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module IEEE Access Road extraction remote sensing image DenseNet-121 network attention module |
| title | DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module |
| title_full | DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module |
| title_fullStr | DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module |
| title_full_unstemmed | DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module |
| title_short | DRCNet: Road Extraction From Remote Sensing Images Using DenseNet With Recurrent Criss-Cross Attention and Convolutional Block Attention Module |
| title_sort | drcnet road extraction from remote sensing images using densenet with recurrent criss cross attention and convolutional block attention module |
| topic | Road extraction remote sensing image DenseNet-121 network attention module |
| url | https://ieeexplore.ieee.org/document/10315115/ |
| work_keys_str_mv | AT debinwei drcnetroadextractionfromremotesensingimagesusingdensenetwithrecurrentcrisscrossattentionandconvolutionalblockattentionmodule AT pinruli drcnetroadextractionfromremotesensingimagesusingdensenetwithrecurrentcrisscrossattentionandconvolutionalblockattentionmodule AT hongjixie drcnetroadextractionfromremotesensingimagesusingdensenetwithrecurrentcrisscrossattentionandconvolutionalblockattentionmodule AT yongqiangxu drcnetroadextractionfromremotesensingimagesusingdensenetwithrecurrentcrisscrossattentionandconvolutionalblockattentionmodule |