Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet
Deep learning algorithms offer an effective solution to the inefficiencies and poor results of traditional methods for building a footprint extraction from high-resolution remote sensing imagery. However, the heterogeneous shapes and sizes of buildings render local extraction vulnerable to the influ...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/15/3766 |
| _version_ | 1797586066670092288 |
|---|---|
| author | Geding Yan Haitao Jing Hui Li Huanchao Guo Shi He |
| author_facet | Geding Yan Haitao Jing Hui Li Huanchao Guo Shi He |
| author_sort | Geding Yan |
| collection | DOAJ |
| description | Deep learning algorithms offer an effective solution to the inefficiencies and poor results of traditional methods for building a footprint extraction from high-resolution remote sensing imagery. However, the heterogeneous shapes and sizes of buildings render local extraction vulnerable to the influence of intricate backgrounds or scenes, culminating in intra-class inconsistency and inaccurate segmentation outcomes. Moreover, the methods for extracting buildings from very high-resolution (VHR) images at present often lose spatial texture information during down-sampling, leading to problems, such as blurry image boundaries or object sticking. To solve these problems, we propose the multi-scale boundary-refined HRNet (MBR-HRNet) model, which preserves detailed boundary features for accurate building segmentation. The boundary refinement module (BRM) enhances the accuracy of small buildings and boundary extraction in the building segmentation network by integrating edge information learning into a separate branch. Additionally, the multi-scale context fusion module integrates feature information of different scales, enhancing the accuracy of the final predicted image. Experiments on WHU and Massachusetts building datasets have shown that MBR-HRNet outperforms other advanced semantic segmentation models, achieving the highest intersection over union results of 91.31% and 70.97%, respectively. |
| first_indexed | 2024-03-11T00:17:04Z |
| format | Article |
| id | doaj.art-0466a0bdb7ae4289886bc678775834a4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T00:17:04Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-0466a0bdb7ae4289886bc678775834a42023-11-18T23:30:30ZengMDPI AGRemote Sensing2072-42922023-07-011515376610.3390/rs15153766Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNetGeding Yan0Haitao Jing1Hui Li2Huanchao Guo3Shi He4School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, ChinaSchool of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, ChinaSchool of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, ChinaSchool of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, ChinaSchool of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, ChinaDeep learning algorithms offer an effective solution to the inefficiencies and poor results of traditional methods for building a footprint extraction from high-resolution remote sensing imagery. However, the heterogeneous shapes and sizes of buildings render local extraction vulnerable to the influence of intricate backgrounds or scenes, culminating in intra-class inconsistency and inaccurate segmentation outcomes. Moreover, the methods for extracting buildings from very high-resolution (VHR) images at present often lose spatial texture information during down-sampling, leading to problems, such as blurry image boundaries or object sticking. To solve these problems, we propose the multi-scale boundary-refined HRNet (MBR-HRNet) model, which preserves detailed boundary features for accurate building segmentation. The boundary refinement module (BRM) enhances the accuracy of small buildings and boundary extraction in the building segmentation network by integrating edge information learning into a separate branch. Additionally, the multi-scale context fusion module integrates feature information of different scales, enhancing the accuracy of the final predicted image. Experiments on WHU and Massachusetts building datasets have shown that MBR-HRNet outperforms other advanced semantic segmentation models, achieving the highest intersection over union results of 91.31% and 70.97%, respectively.https://www.mdpi.com/2072-4292/15/15/3766building footprint extractionremote sensing imageryboundary refinementmulti-scale context fusionintra-class inconsistency |
| spellingShingle | Geding Yan Haitao Jing Hui Li Huanchao Guo Shi He Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet Remote Sensing building footprint extraction remote sensing imagery boundary refinement multi-scale context fusion intra-class inconsistency |
| title | Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet |
| title_full | Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet |
| title_fullStr | Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet |
| title_full_unstemmed | Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet |
| title_short | Enhancing Building Segmentation in Remote Sensing Images: Advanced Multi-Scale Boundary Refinement with MBR-HRNet |
| title_sort | enhancing building segmentation in remote sensing images advanced multi scale boundary refinement with mbr hrnet |
| topic | building footprint extraction remote sensing imagery boundary refinement multi-scale context fusion intra-class inconsistency |
| url | https://www.mdpi.com/2072-4292/15/15/3766 |
| work_keys_str_mv | AT gedingyan enhancingbuildingsegmentationinremotesensingimagesadvancedmultiscaleboundaryrefinementwithmbrhrnet AT haitaojing enhancingbuildingsegmentationinremotesensingimagesadvancedmultiscaleboundaryrefinementwithmbrhrnet AT huili enhancingbuildingsegmentationinremotesensingimagesadvancedmultiscaleboundaryrefinementwithmbrhrnet AT huanchaoguo enhancingbuildingsegmentationinremotesensingimagesadvancedmultiscaleboundaryrefinementwithmbrhrnet AT shihe enhancingbuildingsegmentationinremotesensingimagesadvancedmultiscaleboundaryrefinementwithmbrhrnet |