基于RANSAC的WTLSD平面拟合算法研究  

Research on algorithm of plane fitting of RANSAC-WTLSD

作　　者：张中岳 周惠兴 王舜 于大海[3] 张玉品 赵亢[3] Zhang Zhongyue;Zhou Huixing;Wang Shun;Yu Dahai;Zhang Yupin;Zhao Kang(School of Mechanical-Electronic and Vehicle Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;School of Civil and Transportation Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;BCEG LIUJIAN Construction Engineering Group Co.,Ltd.,Beijing 100143,China)

机构地区：[1]北京建筑大学机电与车辆工程学院,北京100044 [2]北京建筑大学土木与交通工程学院,北京100044 [3]北京六建集团有限责任公司,北京100143

出　　处：《国外电子测量技术》2022年第6期93-98,共6页Foreign Electronic Measurement Technology

基　　金：北京市自然科学基金项目-市教委联合基金(KZ202110016024);北京建筑大学博士研究生科研能力项目(DG2021015)资助。

摘　　要：为提高点云数据的平面拟合精度,降低点云数据中噪点对平面拟合算法的影响,提出一种基于随机采样一致性(random sample consensus,RANSAC)算法的距离加权整体最小二乘(weighted total least squares based distances,WTLSD)平面拟合算法,即RANSAC-WTLSD平面拟合算法。算法通过RANSAC算法对点云数据进行平面初拟合,基于初拟合平面参数构建初始距离权阵。在经过WTLSD算法对拟合平面参数的迭代计算与距离权阵的反复修正后,求得最终平面拟合参数。通过在仿真点云数据以及实际点云数据中的实验结果表明,该算法比常规平面拟合算法相比,具有更低的单位权中误差,更高的平面拟合精度,具有一定的实际应用价值。该算法适用于小数量级的点云数据平面拟合。In order to improve accuracy of plane fitting and reduce the impact of random error in LiDAR data to accuracy of plane fitting,this paper presents an algorithm of plane fitting based on random sample consensus(RANSAC)and weighted total least squares based distances(WTLSD),which called RANSAC-WTLSD.Firstly,this algorithm uses the algorithm of RANSAC to fit the initial plane of the data of the LiDAR.Then,Getting the distance-weighted matrix based on the parameters of the initial plane.The final parameters of the plane fitting were gotten through the iterative calculating through the algorithm of WTLSD to get the parameters of the plane with the corrected distance-weighted matrix.The experimental results in the simulation point cloud data and the actual point cloud data show that this algorithm has lower error in unit weight and higher precision of plane fitting.This algorithm has practical application value and can be used in plane fitting based on spare point cloud.

关 键 词：平面拟合 点云数据 随机采样一致性 距离加权最小二乘法 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]