A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images
Deep learning has proven to be highly successful at semantic segmentation of remote sensing images (RSIs); however, it remains challenging due to the significant intraclass variation and interclass similarity, which limit the accuracy and continuity of feature recognition in land use and land cover...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/11/2811 |
| _version_ | 1797596847628353536 |
|---|---|
| author | Wei Zheng Jiangfan Feng Zhujun Gu Maimai Zeng |
| author_facet | Wei Zheng Jiangfan Feng Zhujun Gu Maimai Zeng |
| author_sort | Wei Zheng |
| collection | DOAJ |
| description | Deep learning has proven to be highly successful at semantic segmentation of remote sensing images (RSIs); however, it remains challenging due to the significant intraclass variation and interclass similarity, which limit the accuracy and continuity of feature recognition in land use and land cover (LULC) applications. Here, we develop a stage-adaptive selective network that can significantly improve the accuracy and continuity of multiscale ground objects. Our proposed framework can learn to implement multiscale details based on a specific attention method (SaSPE) and transformer that work collectively. In addition, we enhance the feature extraction capability of the backbone network at both local and global scales by improving the window attention mechanism of the Swin Transfer. We experimentally demonstrate the success of this framework through quantitative and qualitative results. This study demonstrates the strong potential of the prior knowledge of deep learning-based models for semantic segmentation of RSIs. |
| first_indexed | 2024-03-11T02:58:47Z |
| format | Article |
| id | doaj.art-a431645f13c34c52a4b321074f807dbc |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T02:58:47Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-a431645f13c34c52a4b321074f807dbc2023-11-18T08:29:01ZengMDPI AGRemote Sensing2072-42922023-05-011511281110.3390/rs15112811A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing ImagesWei Zheng0Jiangfan Feng1Zhujun Gu2Maimai Zeng3School of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaSchool of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaPearl River Water Resources Research Institute, Pearl River Water Resources Commission, Guangzhou 510610, ChinaPearl River Water Resources Research Institute, Pearl River Water Resources Commission, Guangzhou 510610, ChinaDeep learning has proven to be highly successful at semantic segmentation of remote sensing images (RSIs); however, it remains challenging due to the significant intraclass variation and interclass similarity, which limit the accuracy and continuity of feature recognition in land use and land cover (LULC) applications. Here, we develop a stage-adaptive selective network that can significantly improve the accuracy and continuity of multiscale ground objects. Our proposed framework can learn to implement multiscale details based on a specific attention method (SaSPE) and transformer that work collectively. In addition, we enhance the feature extraction capability of the backbone network at both local and global scales by improving the window attention mechanism of the Swin Transfer. We experimentally demonstrate the success of this framework through quantitative and qualitative results. This study demonstrates the strong potential of the prior knowledge of deep learning-based models for semantic segmentation of RSIs.https://www.mdpi.com/2072-4292/15/11/2811remote sensingsemantic segmentationposition awarenessattention networkland use and land cover (LULC) |
| spellingShingle | Wei Zheng Jiangfan Feng Zhujun Gu Maimai Zeng A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images Remote Sensing remote sensing semantic segmentation position awareness attention network land use and land cover (LULC) |
| title | A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images |
| title_full | A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images |
| title_fullStr | A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images |
| title_full_unstemmed | A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images |
| title_short | A Stage-Adaptive Selective Network with Position Awareness for Semantic Segmentation of LULC Remote Sensing Images |
| title_sort | stage adaptive selective network with position awareness for semantic segmentation of lulc remote sensing images |
| topic | remote sensing semantic segmentation position awareness attention network land use and land cover (LULC) |
| url | https://www.mdpi.com/2072-4292/15/11/2811 |
| work_keys_str_mv | AT weizheng astageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT jiangfanfeng astageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT zhujungu astageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT maimaizeng astageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT weizheng stageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT jiangfanfeng stageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT zhujungu stageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages AT maimaizeng stageadaptiveselectivenetworkwithpositionawarenessforsemanticsegmentationoflulcremotesensingimages |