ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention
Semantic segmentation of airborne point clouds is crucial for 3D scene reconstruction and remote sensing in surveying applications. Current deep learning methods for point clouds primarily focus on effectively aggregating local neighborhood information. However, they often overlook the fusion of glo...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10375699/ |
| _version_ | 1827379782639157248 |
|---|---|
| author | Shuowen Huang Qingwu Hu Pengcheng Zhao Jiayuan Li Mingyao Ai Shaohua Wang |
| author_facet | Shuowen Huang Qingwu Hu Pengcheng Zhao Jiayuan Li Mingyao Ai Shaohua Wang |
| author_sort | Shuowen Huang |
| collection | DOAJ |
| description | Semantic segmentation of airborne point clouds is crucial for 3D scene reconstruction and remote sensing in surveying applications. Current deep learning methods for point clouds primarily focus on effectively aggregating local neighborhood information. However, they often overlook the fusion of global context information and elevation features, which are vital for airborne point clouds. In this study, we propose Dense-LGEANet, a novel network with dense connected architecture and multiscale feature supervision based on our designed LGEA module. The key component of our LGEA module is the combination of the graph convolution block and the transformer block with elevation attention. It can effectively fuse local neighborhood information and global context information to improve the accuracy of semantic segmentation of airborne point cloud. Moreover, the designed dense connected network architecture can enhance the feature extraction capability for point cloud objects at different scales by facilitating interactions between multiple up-sampling and down-sampling layers. We have conducted multiple experiments on the public point cloud dataset. Experimental results show that our method can achieve an mIoU of 58.5% and an mF1 of 72.0% on the ISPRS Vaihingen 3D dataset, while an mIoU of 67.2% and an mF1 of 78.3% on the LASDU dataset. It demonstrates the superior performance of our network and the effectiveness of the proposed feature enhancement module and network architecture. |
| first_indexed | 2024-03-08T13:23:04Z |
| format | Article |
| id | doaj.art-e778398c7d9c4d9e8b148a3043c94315 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-03-08T13:23:04Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-e778398c7d9c4d9e8b148a3043c943152024-01-18T00:00:08ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01172877288910.1109/JSTARS.2023.334722410375699ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation AttentionShuowen Huang0https://orcid.org/0009-0007-5085-2569Qingwu Hu1https://orcid.org/0000-0003-0866-6678Pengcheng Zhao2https://orcid.org/0000-0002-1581-634XJiayuan Li3https://orcid.org/0000-0002-9850-1668Mingyao Ai4https://orcid.org/0000-0001-6343-7195Shaohua Wang5https://orcid.org/0009-0000-3070-6678School of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Hubei Luojia Laboratory, Wuhan University, Wuhan, ChinaSemantic segmentation of airborne point clouds is crucial for 3D scene reconstruction and remote sensing in surveying applications. Current deep learning methods for point clouds primarily focus on effectively aggregating local neighborhood information. However, they often overlook the fusion of global context information and elevation features, which are vital for airborne point clouds. In this study, we propose Dense-LGEANet, a novel network with dense connected architecture and multiscale feature supervision based on our designed LGEA module. The key component of our LGEA module is the combination of the graph convolution block and the transformer block with elevation attention. It can effectively fuse local neighborhood information and global context information to improve the accuracy of semantic segmentation of airborne point cloud. Moreover, the designed dense connected network architecture can enhance the feature extraction capability for point cloud objects at different scales by facilitating interactions between multiple up-sampling and down-sampling layers. We have conducted multiple experiments on the public point cloud dataset. Experimental results show that our method can achieve an mIoU of 58.5% and an mF1 of 72.0% on the ISPRS Vaihingen 3D dataset, while an mIoU of 67.2% and an mF1 of 78.3% on the LASDU dataset. It demonstrates the superior performance of our network and the effectiveness of the proposed feature enhancement module and network architecture.https://ieeexplore.ieee.org/document/10375699/Airborne laser scanning (ALS)graph convolutionpoint cloudsemantic segmentationtransformer |
| spellingShingle | Shuowen Huang Qingwu Hu Pengcheng Zhao Jiayuan Li Mingyao Ai Shaohua Wang ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Airborne laser scanning (ALS) graph convolution point cloud semantic segmentation transformer |
| title | ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention |
| title_full | ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention |
| title_fullStr | ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention |
| title_full_unstemmed | ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention |
| title_short | ALS Point Cloud Semantic Segmentation Based on Graph Convolution and Transformer With Elevation Attention |
| title_sort | als point cloud semantic segmentation based on graph convolution and transformer with elevation attention |
| topic | Airborne laser scanning (ALS) graph convolution point cloud semantic segmentation transformer |
| url | https://ieeexplore.ieee.org/document/10375699/ |
| work_keys_str_mv | AT shuowenhuang alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention AT qingwuhu alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention AT pengchengzhao alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention AT jiayuanli alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention AT mingyaoai alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention AT shaohuawang alspointcloudsemanticsegmentationbasedongraphconvolutionandtransformerwithelevationattention |