一种稳健的激光点云拟合空间圆柱面方法  

A Robust Method of Fitting Cylindrical Surface with Laser Point Cloud

作　　者：罗成文 黎东 聂胜 隋立春[3] LUO Chengwen;LI Dong;NIE Sheng;SUI Lichun(Key Laboratory of Digital Earth Science,Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100094,China;University of Chinese Academy of Sciences,Beijing 100049,China;College of Geology Engineering and Geomatics,Chang’an University,Xi’an 710061,China)

机构地区：[1]中国科学院空天信息创新研究院数字地球重点实验室,北京100094 [2]中国科学院大学,北京100049 [3]长安大学地质工程与测绘学院,陕西西安710061

出　　处：《测绘地理信息》2024年第3期31-34,共4页Journal of Geomatics

基　　金：国家重点研发计划(2021YFF0704600)。

摘　　要：针对激光点云三维建模应用中完整圆柱面及局部圆柱面拟合结果稳健性问题,提出一种稳健的空间圆柱面拟合方法。首先采用RANSAC(random sample consensus)算法拟合直线和圆,确定圆柱模型参数的初值,然后将RANSAC算法结合最小二乘法(least square,LS)实现空间圆柱面拟合。利用仿真实验测试了圆柱面完整度在100%、50%及25%3种情况下且包含各种水平噪声的圆柱面点云,拟合的圆柱半径最大误差仅为0.2 mm。结果表明,与传统RANSAC算法和LS法相比,所提方法不仅具有非常高的拟合精度,而且对数据完整性、噪声等具有较高的稳健性。Aiming at the problem of data fitting of complete and partial cylindrical point cloud and the robustness of cylin⁃drical fitting results in 3D modeling application of Laser point cloud,a robust spatial cylindrical fitting method is proposed.Firstly,the random sample consensus(RANSAC)algorithm is used to fit the line and circle,and the initial values of the cylinder model parameters are determined,Then the RANSAC algorithm is combined with the least squares meth⁃od to fit the spatial cylinder surface.Simulation experiments were performed to test the cylindrical point clouds containing various horizontal noises with 100%,50%and 25%cylindri⁃cal integrity,and the maximum error between the fitted cylin⁃drical radius and the real radius was 0.2mm.The results show that,compared with the traditional RANSAC method and the least squares method,the proposed method not only has very high fitting accuracy,but also has high robustness to data integrity and noise.

关 键 词：激光点云 圆柱面拟合 RANSAC算法 最小二乘法 稳健性 

分 类 号：P234[天文地球—摄影测量与遥感]