Hole detection and filling in road surface point clouds from vehicle-mounted LiDAR

doi:10.13474/j.cnki.11-2246.2025.1124

Abstract

Abstract: Vehicle-mounted LiDAR scans of urban roads often contain holes in the road surface point cloud due to occlusions by vehicles,pedestrians,and roadside objects.These gaps degrade the completeness and accuracy of downstream 3D modeling and spatial analysis.We propose an automated pipeline for hole detection and repair in road surface point clouds.Firstly,noise and non-road points are removed via filtering and clustering techniques to isolate the road surface.Next,a multi-scale Alpha Shape algorithm extracts the 2D road boundary and identifies boundary breaks caused by occlusion; NURBS curves then restore a continuous boundary.Finally,the area within the repaired boundary is partitioned into a regular grid,empty cells are clustered to locate holes,and a quadratic surface is fitted to surrounding points to interpolate and fill each hole.Experiments on ten diverse urban road segments demonstrate that our method achieves 96.5% hole-filling success for common occluders,produces smooth,seamless transitions with the original data,and preserves critical geometric features such as curb and corner shapes.The proposed approach reliably repairs small-scale occlusions in road surface point clouds and is broadly applicable to most urban driving scenarios.

Key words: vehicle-mounted LiDAR, road surface point clouds, point cloud hole filling, Alpha Shape algorithm, NURBS curve fitting

CLC Number:

P237

CHEN Jianping, SHI Jian, DAI Xiangxi, ZHANG Qinyu, HAN Wenquan. Hole detection and filling in road surface point clouds from vehicle-mounted LiDAR[J]. Bulletin of Surveying and Mapping, 2025, 0(11): 154-158.

References

[1] 刘如飞,卢秀山,岳国伟,等.一种车载激光点云数据中道路自动提取方法[J].武汉大学学报(信息科学版),2017,42(2): 250-256.
[2] JAAKKOLA A,HYYPPÄ J,HYYPPÄ H,et al.Retrieval algorithms for road surface modelling using laser-based mobile mapping[J].Sensors,2008,8(9):5238-5249.
[3] ZHONG Mianqing,SUI Lichun,WANG Zhihua,et al.Pavement crack detection from mobile laser scanning point clouds using a time grid[J].Sensors,2020,20(15):4198.
[4] ELHASHASH M,ALBANWAN H,QIN Rongjun.A review of mobile mapping systems:from sensors to applications[J].Sensors,2022,22(11):4262.
[5] LIU Haibin,WU Chao,WANG Huanjie.Real time object detection using LiDAR and camera fusion for autonomous driving[J].Scientific Reports,2023,13:8056.
[6] 方莉娜,杨必胜.车载激光扫描数据的结构化道路自动提取方法[J].测绘学报,2013,42(2):260-267.
[7] 闫利,张毅.基于法向量模糊聚类的道路面点云数据滤波[J].武汉大学学报(信息科学版),2007,32(12):1119-1122.
[8] YADAV M,LOHANI B,SINGH A K.Road surface detection from mobile lidar data[J].ISPRS Annals of the Photogrammetry,Remote Sensing and Spatial Information Sciences,2018,IV-5:95-101.
[9] 杨必胜,魏征,李清泉,等.面向车载激光扫描点云快速分类的点云特征图像生成方法[J].测绘学报,2010,39(5):540-545.
[10] 吴学群,宁津生,杨芳.车载激光扫描数据分类支持下的路面数据提取[J].测绘通报,2018(2):107-110.
[11] BALADO J,MARTÍNEZ-SÁNCHEZ J,ARIAS P,et al.Road environment semantic segmentation with deep learning from MLS point cloud data[J].Sensors,2019,19(16):3466.
[12] 陈天珩,于斌,马羊,等.复杂激光点云数据中路面自动化提取技术[J].南京理工大学学报,2022,46(1):48-54.
[13] 孙伟利.基于车载LiDAR技术的公路三维建模与应用[D].北京:首都师范大学,2013.
[14] 张振华.基于车载LiDAR数据的道路三维建模方法研究[D].赣州:江西理工大学,2018.
[15] 刘嘉煌,邓兴升.公路路面LiDAR数据空洞插补的MLSPF方法[J].测绘工程,2023,32(6):44-53.
[16] WANG Jianning,OLIVEIRA M M.Filling holes on locally smooth surfaces reconstructed from point clouds[J].Image and Vision Computing,2007,25(1):103-113.
[17] LIEPA P.Filling holes in meshes[C]//Proceedings of 2003 Eurographics/ACM SIGGRAPH symposium on Geometry processing.Goslar:Eurographics Association,2003:200-205.
[18] CHEN Jingdao,YI J S K,KAHOUSH M,et al.Point cloud scene completion of obstructed building facades with generative adversarial inpainting[J].Sensors,2020,20(18):5029.
[19] HU Wei,FU Zeqing,GUO Zongming.Local frequency interpretation and non-local self-similarity on graph for point cloud inpainting[J].IEEE Transactions on Image Processing,2019,28(8):4087-4100.
[20] XIE Yizhou,XIE Xiangning,WANG Yuran,et al.Terrain point cloud inpainting via signal decomposition[J].Computers ＆ Graphics,2024,120:103915.
[21] FEI Ben,YANG Weidong,CHEN Wenming,et al.Comprehensive review of deep learning-based 3D point cloud completion processing and analysis[J].IEEE Transactions on Intelligent Transportation Systems,2022,23(12):22862-22883.
[22] 肖信峰,余敏,高飞,等.车载点云数据高速公路路面提取方法[J].遥感信息,2021,36(6):147-152.