| 253 | 6 | 36 |
| 下载次数 | 被引频次 | 阅读次数 |
依赖于平面拟合或局部几何特征区分地面障碍物的方法广泛应用于自动驾驶领域，但是在具有倾斜地形或稀疏数据的情况下，性能会降低。在倾斜地形场景中，利用激光雷达(LiDAR)提供的点云数据，通过地面点云的3D投影与直线拟合，利用栅格地图方法降低计算复杂度，实现地面点云的分割。针对非地面的点云集合，利用SLR聚类算法处理，通过设定强度特征阈值在垂直方向区分地面点与非地面点，并对扫描到的障碍物地面分类。通过实验分析，提出的算法较其他地面点云分割算法，一方面在倾斜地形上具有更好的建图效果，另一方面SLR聚类算法处理后的强度特征在X、Y、Z三个方向覆盖范围更精确，如在X方向相较于快速地面分割算法平均提高了44.0%,相较于添加了栅格地图的算法平均提高了40.1%。
Abstract：The methods that rely on plane fitting or local geometric features to distinguish ground obstacles are widely used in the field of automatic driving, but their performance could be reduced in the case of sloping terrain or sparse data.In a sloping terrain scene, the point cloud data provided by LiDAR, the grid map method through 3D projection and line fitting of the ground point cloudare used to reduce the computational complexity and realize the segmentation of the ground point cloud. For the collection of non-ground point clouds, SLR clustering algorithm is used to process. Ground points and non-ground points are distinguished in the vertical direction by setting the intensity feature threshold, and the scanned obstacle ground is classified.Through experimental analysis, the proposed algorithm has better mapping effect than other ground point cloud segmentation algorithms in sloping terrain. On the other hand, the intensity features processed by SLR clustering algorithm are more accurate in the coverage of X, Y and Z directions. For example, in the X direction, compared with the fast ground segmentation algorithm, the average increase is 44.0%; compared with the algorithm with grid map, the average increase is 40.1%.
[1]陈建，于殿泓，惠鑫鑫，等．基于二维激光雷达多车道车辆检测与分车算法的研究[J]．传感技术学报，2022,35(11):1513-1520．
[2]周梦飞，王子洋，杨家富．应用于无人车的三维点云分类方法研究综述[J]．传感技术学报，2022,35(7):931-937．
[3]THRUN S,MONTEMERLO M,DAHLKAMP H,et al.Stanley:The Robot that Won the DARPA Grand Challenge[J].Journal of Field Robotics,2006,23 (9):661-692．
[4]HIMMELSBACH M,HUNDELSHAUSEN F V,WUENSCHEH J.Fast Segmentation of 3D Point Clouds for Ground Vehicles[C]∥2010 IEEE Intelligent Vehicles Symposium.La Jolla:IEEE,2010:560-565．
[5]张子腾，盛传贞，蔚保国，等．基于约束优化的多传感器车载定位方法[J]．无线电工程，2022,52 (10):1781-1787．
[6]RODRíGUEZ-GARAVITO C H,PONZ A,GARCíA F,et al.Automatic Laser and Camera Extrinsic Calibrationfor Data Fusion Using Road Plane[C]∥17th International Conference on Information Fusion (FUSION).Salamanca:IEEE 2014:1-6．
[7]HU X,RODRIGUEZ F S A,GEPPERTH.A Multi-modal System for Road Detection and Segmentation[C]∥2014IEEE Intelligent Vehicles Symposium Proceedings.Dearborn:IEEE,2014:1365-1370．
[8]ZERMAS D,IZZAT I,PAPANIKOLOPOULOS N.Fast Segmentation of 3D Point Clouds:A Paradigm on Li DARData for Autonomous Vehicle Applications[C]∥2017IEEE International Conference on Robotics and Automation(ICRA).Singapore:IEEE,2017:5067-5073．
[9]NITSCH J,AGUILAR J,NIETO J,et al.3D Ground Point Classification for Automotive Scenarios[C]∥2018 21st International Conference on Intelligent Transportation Systems (ITSC).Maui:IEEE,2018:2603-2608．
[10]RIEKEN J,MATTHAEI R,MAURER M.Benefits of Using Explicit Ground-plane Information for Grid-based Urban Environment Modeling[C]∥2015 18th International Conference on Information Fusion (Fusion).Washington D.C.:IEEE,2015:2049-2056．
[11]邹兵，陈鹏，刘登洪．一种基于栅格投影的快速地面点云分割算法[J]．城市勘测，2021(3):112-116．
[12]JIMéNEZ V,GODOY J,ARTU NNEDO A,et al.Ground Segmentation Algorithm for Sloped Terrain and Sparse LiDAR Point Cloud[J].IEEE Access,2021,9:132914-132927．
[13]SHAN T,ENGLOT B.Le GO-LOAM:Lightweight and Ground-optimized Li DAR Odometry and Mapping on Variable Terrain[C]∥2018 IEEE/RSJ International Conference on Intelligent Robots and Systems(IROS).Madrid:IEEE,2018:4758-4765．
[14]LIM H,HWANG S,MYUNG H.ERASOR:Egocentric Ratio of Pseudo Occupancy-based Dynamic Object Removal for Static 3D Point Cloud Map Building[J].IEEE Robotics and Automation Letters,2021,6 (2):2272-2279．
[15]WEI X,LV J,SUN J,et al.Ground-SLAM:Ground Constrained Li DAR SLAM for Structured Multi-floor Environments[EB/OL].(2021-05-05)[2023-04-10].https:∥arxiv.org/abs/2103.03713．
[16]LIM H,OH M,MYUNG H.Patchwork:Concentric Zonebased Region-wise Ground Segmentation with Ground Likelihood Estimation Using a 3D Li DAR Sensor[J].IEEE Robotics and Automation Letters,2021,6 (4):6458-6465．
[17]RUMMELHARD L,PAIGWAR A,NEGRE A,et al.Ground Estimation and Point Cloud Segmentation Using Spatiotemporal Conditional Random Field[C]∥2017IEEE Intelligent Vehicles Symposium(IV).Los Angeles:IEEE,2017:1105-1110.
基本信息:
DOI：
中图分类号:TN958.98;U463.6
引用信息:
[1]张清宇,崔丽珍,李敏超等.倾斜地面3D点云快速分割算法[J].无线电工程,2024,54(02):447-456.
基金信息:
国家自然科学基金(62261042); 内蒙古自治区科技计划项目(2022YFSH0051)~~