Intelligent recognition of underground space targets based on LiDAR point clouds

doi:10.13474/j.cnki.11-2246.2026.0410

Abstract

Abstract: Point cloud semantic segmentation has become a key technology for achieving multi-semantic visualization in underground alleyway health monitoring.To address the low segmentation accuracy of target edges in alleyway scenarios,this paper constructs a large-scale point cloud semantic segmentation dataset for alleyway scenarios ALSD,and proposes a segmentation method adapted to this scenario.Based on backbone feature extraction,it introduces a local multi-scale neighborhood considering 3D curvature and a global spatial feature enhancement module,combined with an attention mechanism to improve the representation ability of small components and complex boundaries.An evaluation index system for alleyway scenarios is also established.On the ALSD dataset of real alleyway scenarios,we systematically analyzed the impact of training set size,input dimension,and hyperparameter settings on model performance.Experiments show that the proposed method achieves IoUs of 0.858,0.883,0.933,and 0.822 for pipes,supports,ground,and columns,respectively,with an mIoU of 0.865 and an overall accuracy (OA)of 98.7%.Compared with typical deep learning baseline methods such as PointNet++ and RandLA-Net,the proposed model achieves higher semantic segmentation accuracy on the ALSD dataset,which can provide high-precision 3D semantic support for the structural health monitoring of underground alleyways.

Key words: deep learning, 3D alleyway point cloud, semantic segmentation, LiDAR

CLC Number:

P237

GUO Ming, ZHANG Xiaolan, QIU Gongrun, GUO Shuai, ZHU Li. Intelligent recognition of underground space targets based on LiDAR point clouds[J]. Bulletin of Surveying and Mapping, 2026, 0(4): 65-72.

References

[1] FENG Xingfa,YANG Xian,LI Yongsuo,et al.Influence of underground tunnel excavation on upper-level existing pipelines[J].Advances in Civil Engineering,2022,2022:5469345.
[2] LI Zhe,REN Xiaoqiang,HE Hongming.The damage deterioration mechanism of tunnel structure under the effect of thermo-mechanical coupling and its maintenance[J].International Journal of Heat and Technology,2022,40(4):910-916.
[3] 郭明,唐星雨,刘运明,等.融合时序InSAR和LiDAR技术的海岛地铁沿线地面形变分析[J].光学精密工程,2023,31(13):1988-1999.
[4] WANG Qiling,WANG Mingnian,YUAN Yong,et al.Thermomechanical behavior of tunnel linings in the geothermal environment:field tests and analytical study[J].Tunnelling and Underground Space Technology,2023,137:105109.
[5] GUO Ming,YAN Bingnan,WANG Guoli,et al.Iterative automatic global registration algorithm for multi-view point cloud of underground tunnel space[J].Measurement and Control,2021,54(3/4):385-395.
[6] TIAN Lin,LI Qingquan,HE Li,et al.Image-range stitching and semantic-based crack detection methods for tunnel inspection vehicles[J].Remote Sensing,2023,15(21):5158.
[7] DU Meng,FAN Jia,HUANG Yuchun,et al.Mosaicking of mountain tunnel images guided by laser rangefinder[J].Automation in Construction,2021,127:103708.
[8] SZREK J,JAKUBIAK J,ZIMROZ R.A mobile robot-based system for automatic inspection of belt conveyors in mining industry[J].Energies,2022,15(1):327.
[9] CHENG Peng,GUO Ming,WANG Haibo,et al.Fusion segmentation network guided by adaptive sampling radius and channel attention mechanism module for MLS point clouds[J].Applied Sciences,2023,13(1):281.
[10] GUO Ming,ZHANG Xiaolan,CHENG Peng,et al.Multi-region group sampling radius semantic segmentation network guided by spatial information for highway[J].Journal of Nondestructive Evaluation,2025,44(2):52.
[11] DU Liming,ZHONG Ruofei,SUN Haili,et al.Study of the integration of the CNU-TS-1 mobile tunnel monitoring system[J].Sensors,2018,18(2):420.
[12] CUI Hao,LI Jian,MAO Qingzhou,et al.STSD:a large-scale benchmark for semantic segmentation of subway tunnel point cloud[J].Tunnelling and Underground Space Technology,2024,150:105829.
[13] LIN Wei,SHEIL B,ZHANG Pin,et al.Seg2Tunnel:a hierarchical point cloud dataset and benchmarks for segmentation of segmental tunnel linings[J].Tunnelling and Underground Space Technology,2024,147:105735.
[14] SALAMANCA S,MERCHÁN P,ESPACIO A,et al.Segmentation of 3D point clouds of heritage buildings using edge detection and supervoxel-based topology[J].Sensors,2024,24(13):4390.
[15] SALTORI C,GALASSO F,FIAMENI G,et al.Compositional semantic mix for domain adaptation in point cloud segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(12):14234-14247.
[16] POUX F,MATTES C,SELMAN Z,et al.Automatic region-growing system for the segmentation of large point clouds[J].Automation in Construction,2022,138:104250.
[17] SERNA A,MARCOTEGUI B,HERNÁNDEZ J.Segmentation of façades from urban 3D point clouds using geometrical and morphological attribute-based operators[J].ISPRS International Journal of Geo-Information,2016,5(1):6.
[18] LI Dilong,LU Chenghui,CHEN Zixing,et al.Graph neural networks in point clouds:a survey[J].Remote Sens,2024,16:2518.
[19] LIN Hongbin,ZHENG Wu,PENG Xiuping.Orientation-encoding CNN for point cloud classification and segmentation[J].Machine Learning and Knowledge Extraction,2021,3(3):601-614.
[20] GUO Yulan,WANG Hanyun,HU Qingyong,et al.Deep learning for 3D point clouds:a survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2021,43(12):4338-4364.
[21] CHARLES R Q,HAO Su,MO Kaichun,et al.PointNet:deep learning on point sets for 3D classification and segmentation[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu:IEEE,2017:77-85.
[22] QI C R,YI L,SU H,et al.Pointnet++:deep hierarchical feature learning on point sets in a metric space[C]//Proceedings of 2017 NIPS Conference on Neural Information Processing Systems.Long Beach:[s.n.],2017.
[23] LI Yangyan,BU Rui,SUN Mingchao,et al.PointCNN:convolution on X-transformed points[C]//Proceedings of 2018 NIPS Conference on Computer Vision and Pattern Recognition.Montreal:[s.n.],2018.
[24] LI Jinyuan,HAO L,CUI Guorong,et al.GACNet:a generative adversarial capsule network for regional epitaxial traffic flow prediction[J].Computers,Materials &Continua,2020,64(2):925-940.
[25] HU Qingyong,YANG Bo,XIE Linhai,et al.RandLA-net:efficient semantic segmentation of large-scale point clouds[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle:IEEE,2020:11105-11114.
[26] XIAO Aoran,YANG Xiaofei,LU Shijian,et al.FPS-Net:a convolutional fusion network for large-scale LiDAR point cloud segmentation[J].ISPRS Journal of Photogrammetry and Remote Sensing,2021,176:237-249.
[27] PHAN A V,LE NGUYEN M,NGUYEN Y L H,et al.DGCNN:a convolutional neural network over large-scale labeled graphs[J].Neural Networks,2018,108:533-543.