Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery
Road segmentation from high-resolution (HR) remote sensing images plays a core role in a wide range of applications. Due to the complex background of HR images, most of the current methods struggle to extract a road network correctly and completely. Furthermore, they suffer from either the loss of c...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-09-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/19/10726 |
| _version_ | 1797576231069155328 |
|---|---|
| author | Lin Gao Chen Chen |
| author_facet | Lin Gao Chen Chen |
| author_sort | Lin Gao |
| collection | DOAJ |
| description | Road segmentation from high-resolution (HR) remote sensing images plays a core role in a wide range of applications. Due to the complex background of HR images, most of the current methods struggle to extract a road network correctly and completely. Furthermore, they suffer from either the loss of context information or high redundancy of details information. To alleviate these problems, we employ a dual branch dilated pyramid network (DPBFN), which enables dual-branch feature passing between two parallel paths when it is merged to a typical road extraction structure. A DPBFN consists of three parts: a residual multi-scaled dilated convolutional network branch, a transformer branch, and a fusion module. Constructing pyramid features through parallel multi-scale dilated convolution operations with multi-head attention block can enhance road features while suppressing redundant information. Both branches after fusing can solve shadow or vision occlusions and maintain the continuity of the road network, especially on a complex background. Experiments were carried out on three datasets of HR images to showcase the stable performance of the proposed method, and the results are compared with those of other methods. The OA in the three data sets of Massachusetts, Deep Globe, and GF-2 can reach more than 98.26%, 95.25%, and 95.66%, respectively, which has a significant improvement compared with the traditional CNN network. The results and explanation analysis via Grad-CAMs showcase the effective performance in accurately extracting road segments from a complex scene. |
| first_indexed | 2024-03-10T21:49:21Z |
| format | Article |
| id | doaj.art-32aa6155726e4d35b8412c00f416a4f4 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T21:49:21Z |
| publishDate | 2023-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-32aa6155726e4d35b8412c00f416a4f42023-11-19T14:03:22ZengMDPI AGApplied Sciences2076-34172023-09-0113191072610.3390/app131910726Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing ImageryLin Gao0Chen Chen1School of Information Science and Engineering, Shenyang Ligong University, Shenyang 110159, ChinaSchool of Information Science and Engineering, Shenyang Ligong University, Shenyang 110159, ChinaRoad segmentation from high-resolution (HR) remote sensing images plays a core role in a wide range of applications. Due to the complex background of HR images, most of the current methods struggle to extract a road network correctly and completely. Furthermore, they suffer from either the loss of context information or high redundancy of details information. To alleviate these problems, we employ a dual branch dilated pyramid network (DPBFN), which enables dual-branch feature passing between two parallel paths when it is merged to a typical road extraction structure. A DPBFN consists of three parts: a residual multi-scaled dilated convolutional network branch, a transformer branch, and a fusion module. Constructing pyramid features through parallel multi-scale dilated convolution operations with multi-head attention block can enhance road features while suppressing redundant information. Both branches after fusing can solve shadow or vision occlusions and maintain the continuity of the road network, especially on a complex background. Experiments were carried out on three datasets of HR images to showcase the stable performance of the proposed method, and the results are compared with those of other methods. The OA in the three data sets of Massachusetts, Deep Globe, and GF-2 can reach more than 98.26%, 95.25%, and 95.66%, respectively, which has a significant improvement compared with the traditional CNN network. The results and explanation analysis via Grad-CAMs showcase the effective performance in accurately extracting road segments from a complex scene.https://www.mdpi.com/2076-3417/13/19/10726remote sensing imagerytransformer mechanismdilated convolutionroad segmentationexplanation analysis |
| spellingShingle | Lin Gao Chen Chen Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery Applied Sciences remote sensing imagery transformer mechanism dilated convolution road segmentation explanation analysis |
| title | Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery |
| title_full | Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery |
| title_fullStr | Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery |
| title_full_unstemmed | Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery |
| title_short | Dual Parallel Branch Fusion Network for Road Segmentation in High-Resolution Optical Remote Sensing Imagery |
| title_sort | dual parallel branch fusion network for road segmentation in high resolution optical remote sensing imagery |
| topic | remote sensing imagery transformer mechanism dilated convolution road segmentation explanation analysis |
| url | https://www.mdpi.com/2076-3417/13/19/10726 |
| work_keys_str_mv | AT lingao dualparallelbranchfusionnetworkforroadsegmentationinhighresolutionopticalremotesensingimagery AT chenchen dualparallelbranchfusionnetworkforroadsegmentationinhighresolutionopticalremotesensingimagery |