SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences
Point cloud data provide three-dimensional (3D) information about objects in the real world, containing rich semantic features. Therefore, the task of semantic segmentation of point clouds has been widely applied in fields such as robotics and autonomous driving. Although existing research has made...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/11/1927 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849722452253343744 |
|---|---|
| author | Bin Guo Chunjing Yao Hongchao Ma Jie Wang Junhao Xu |
| author_facet | Bin Guo Chunjing Yao Hongchao Ma Jie Wang Junhao Xu |
| author_sort | Bin Guo |
| collection | DOAJ |
| description | Point cloud data provide three-dimensional (3D) information about objects in the real world, containing rich semantic features. Therefore, the task of semantic segmentation of point clouds has been widely applied in fields such as robotics and autonomous driving. Although existing research has made unprecedented progress, achieving real-time semantic segmentation of point clouds on airborne devices still faces challenges due to excessive computational and memory requirements. To address this issue, we propose a novel sequence convolution semantic segmentation architecture that integrates Convolutional Neural Networks (CNN) with a sequence-to-sequence (seq2seq) structure, termed SeqConv-Net. This architecture views point cloud semantic segmentation as a sequence generation task. Based on our unique perspective of spatially ordered sequences, we use Recurrent Neural Networks (RNN) to encode elevation information, then input the structured hidden states into a CNN for planar feature extraction. The results are combined with the RNN’s encoded outputs via residual connections and are fed into a decoder for sequence prediction in a seq2seq manner. Experiments show that the SeqConv-Net architecture achieves 75.5% mean Intersection Over Union (mIOU) accuracy on the DALES dataset, with the total processing speed from data preprocessing to prediction being several to tens of times faster than existing methods. Additionally, SeqConv-Net can balance accuracy and speed by adjusting the hyperparameters and using different RNNs and CNNs, providing a new solution for real-time point cloud semantic segmentation in airborne environments. |
| format | Article |
| id | doaj-art-3a34ae1aacb04ef0b1c8b6ed9787c1d4 |
| institution | DOAJ |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-3a34ae1aacb04ef0b1c8b6ed9787c1d42025-08-20T03:11:20ZengMDPI AGRemote Sensing2072-42922025-06-011711192710.3390/rs17111927SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered SequencesBin Guo0Chunjing Yao1Hongchao Ma2Jie Wang3Junhao Xu4School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430072, ChinaPoint cloud data provide three-dimensional (3D) information about objects in the real world, containing rich semantic features. Therefore, the task of semantic segmentation of point clouds has been widely applied in fields such as robotics and autonomous driving. Although existing research has made unprecedented progress, achieving real-time semantic segmentation of point clouds on airborne devices still faces challenges due to excessive computational and memory requirements. To address this issue, we propose a novel sequence convolution semantic segmentation architecture that integrates Convolutional Neural Networks (CNN) with a sequence-to-sequence (seq2seq) structure, termed SeqConv-Net. This architecture views point cloud semantic segmentation as a sequence generation task. Based on our unique perspective of spatially ordered sequences, we use Recurrent Neural Networks (RNN) to encode elevation information, then input the structured hidden states into a CNN for planar feature extraction. The results are combined with the RNN’s encoded outputs via residual connections and are fed into a decoder for sequence prediction in a seq2seq manner. Experiments show that the SeqConv-Net architecture achieves 75.5% mean Intersection Over Union (mIOU) accuracy on the DALES dataset, with the total processing speed from data preprocessing to prediction being several to tens of times faster than existing methods. Additionally, SeqConv-Net can balance accuracy and speed by adjusting the hyperparameters and using different RNNs and CNNs, providing a new solution for real-time point cloud semantic segmentation in airborne environments.https://www.mdpi.com/2072-4292/17/11/1927point cloudsemantic segmentationrecurrent neural networkconvolutional neural networkseq2seq |
| spellingShingle | Bin Guo Chunjing Yao Hongchao Ma Jie Wang Junhao Xu SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences Remote Sensing point cloud semantic segmentation recurrent neural network convolutional neural network seq2seq |
| title | SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences |
| title_full | SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences |
| title_fullStr | SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences |
| title_full_unstemmed | SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences |
| title_short | SeqConv-Net: A Deep Learning Segmentation Framework for Airborne LiDAR Point Clouds Based on Spatially Ordered Sequences |
| title_sort | seqconv net a deep learning segmentation framework for airborne lidar point clouds based on spatially ordered sequences |
| topic | point cloud semantic segmentation recurrent neural network convolutional neural network seq2seq |
| url | https://www.mdpi.com/2072-4292/17/11/1927 |
| work_keys_str_mv | AT binguo seqconvnetadeeplearningsegmentationframeworkforairbornelidarpointcloudsbasedonspatiallyorderedsequences AT chunjingyao seqconvnetadeeplearningsegmentationframeworkforairbornelidarpointcloudsbasedonspatiallyorderedsequences AT hongchaoma seqconvnetadeeplearningsegmentationframeworkforairbornelidarpointcloudsbasedonspatiallyorderedsequences AT jiewang seqconvnetadeeplearningsegmentationframeworkforairbornelidarpointcloudsbasedonspatiallyorderedsequences AT junhaoxu seqconvnetadeeplearningsegmentationframeworkforairbornelidarpointcloudsbasedonspatiallyorderedsequences |