PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images
Semantic segmentation of very-high-resolution (VHR) remote sensing images plays an important role in the intelligent interpretation of remote sensing since it predicts pixel-level labels to the images. Although many semantic segmentation methods of VHR remote sensing images have emerged recently and...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/17/4219 |
| _version_ | 1827665129655762944 |
|---|---|
| author | Bo Liu Jinwu Hu Xiuli Bi Weisheng Li Xinbo Gao |
| author_facet | Bo Liu Jinwu Hu Xiuli Bi Weisheng Li Xinbo Gao |
| author_sort | Bo Liu |
| collection | DOAJ |
| description | Semantic segmentation of very-high-resolution (VHR) remote sensing images plays an important role in the intelligent interpretation of remote sensing since it predicts pixel-level labels to the images. Although many semantic segmentation methods of VHR remote sensing images have emerged recently and achieved good results, it is still a challenging task because the objects of VHR remote sensing images show large intra-class and small inter-class variations, and their size varies in a large range. Therefore, we proposed a novel semantic segmentation framework for VHR remote sensing images, called Positioning Guidance Network (PGNet), which consists of the feature extractor, a positioning guiding module (PGM), and a self-multiscale collection module (SMCM). First, the PGM can extract long-range dependence and global context information with the help of the transformer architecture and effectively transfer them to each pyramid-level feature, thus effectively improving the segmentation effectiveness between different semantic objects. Secondly, the SMCM we designed can effectively extract multi-scale information and generate high-resolution feature maps with high-level semantic information, thus helping to segment objects in small and varying sizes. Without bells and whistles, the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>m</mi><mi>I</mi><mi>o</mi><mi>U</mi></mrow></semantics></math></inline-formula> scores of the proposed PGNet on the iSAID dataset and ISPRS Vaihingn dataset are 1.49% and 2.40% higher than FactSeg, respectively. |
| first_indexed | 2024-03-10T01:19:08Z |
| format | Article |
| id | doaj.art-8c7774dbecfe49a9a92076e8186c70b4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T01:19:08Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-8c7774dbecfe49a9a92076e8186c70b42023-11-23T14:02:55ZengMDPI AGRemote Sensing2072-42922022-08-011417421910.3390/rs14174219PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing ImagesBo Liu0Jinwu Hu1Xiuli Bi2Weisheng Li3Xinbo Gao4Department of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaDepartment of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaDepartment of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaDepartment of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaDepartment of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaSemantic segmentation of very-high-resolution (VHR) remote sensing images plays an important role in the intelligent interpretation of remote sensing since it predicts pixel-level labels to the images. Although many semantic segmentation methods of VHR remote sensing images have emerged recently and achieved good results, it is still a challenging task because the objects of VHR remote sensing images show large intra-class and small inter-class variations, and their size varies in a large range. Therefore, we proposed a novel semantic segmentation framework for VHR remote sensing images, called Positioning Guidance Network (PGNet), which consists of the feature extractor, a positioning guiding module (PGM), and a self-multiscale collection module (SMCM). First, the PGM can extract long-range dependence and global context information with the help of the transformer architecture and effectively transfer them to each pyramid-level feature, thus effectively improving the segmentation effectiveness between different semantic objects. Secondly, the SMCM we designed can effectively extract multi-scale information and generate high-resolution feature maps with high-level semantic information, thus helping to segment objects in small and varying sizes. Without bells and whistles, the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>m</mi><mi>I</mi><mi>o</mi><mi>U</mi></mrow></semantics></math></inline-formula> scores of the proposed PGNet on the iSAID dataset and ISPRS Vaihingn dataset are 1.49% and 2.40% higher than FactSeg, respectively.https://www.mdpi.com/2072-4292/14/17/4219remote sensing imagessemantic segmentationpositioning guiding moduleself-multiscale collection moduletransformer |
| spellingShingle | Bo Liu Jinwu Hu Xiuli Bi Weisheng Li Xinbo Gao PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images Remote Sensing remote sensing images semantic segmentation positioning guiding module self-multiscale collection module transformer |
| title | PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images |
| title_full | PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images |
| title_fullStr | PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images |
| title_full_unstemmed | PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images |
| title_short | PGNet: Positioning Guidance Network for Semantic Segmentation of Very-High-Resolution Remote Sensing Images |
| title_sort | pgnet positioning guidance network for semantic segmentation of very high resolution remote sensing images |
| topic | remote sensing images semantic segmentation positioning guiding module self-multiscale collection module transformer |
| url | https://www.mdpi.com/2072-4292/14/17/4219 |
| work_keys_str_mv | AT boliu pgnetpositioningguidancenetworkforsemanticsegmentationofveryhighresolutionremotesensingimages AT jinwuhu pgnetpositioningguidancenetworkforsemanticsegmentationofveryhighresolutionremotesensingimages AT xiulibi pgnetpositioningguidancenetworkforsemanticsegmentationofveryhighresolutionremotesensingimages AT weishengli pgnetpositioningguidancenetworkforsemanticsegmentationofveryhighresolutionremotesensingimages AT xinbogao pgnetpositioningguidancenetworkforsemanticsegmentationofveryhighresolutionremotesensingimages |