基于三角网格表示的点云匹配算法  

A Registration Algorithm For Point Clouds Based on The Triangular Representation

作　　者：胡凡建[1,2] 

机构地区：[1]湖北第二师范学院物理与机电学院,武汉430205 [2]华中科技大学材料成形与模具技术国家重点实验室,武汉430074

出　　处：《湖北第二师范学院学报》2016年第2期16-21,共6页Journal of Hubei University of Education

摘　　要：针对传统的匹配方法收敛速度慢、对初值敏感的问题,本文提出了一种基于三角网格表示的点云匹配算法。该算法主要包括两个步骤:首先,采用主成份分析对点云进行整体分析,通过奇异值分解计算初始变换参数;然后,采用螺旋运动理论在三角网格模型中定义点-切面距离以构造目标函数,并通过求解一个线性方程组计算最优刚体变换参数。实验结果证明了本文所提算法的有效性。This paper proposes a novel registration algorithm based on the triangular representation,aiming to handle the problems of slow convergence and existing sensibility to initial value in traditional algorithms. The proposed algorithm consists of two major steps： First,Principal Component Analysis is used to analyze the data,then the Singular Value decomposition is used to calculate the initial transformation parameters; Second,the theory of screw motion is employed to define the point-tangent distance in triangles in order to construct the objective function which is calculated to obtain the optimal rigid transformation parameters by solving a series of linear systems. The validity of the proposed algorithm is verified by the experiment.

关 键 词：匹配 刚体变换 主成份分析 螺旋运动 

分 类 号：TH14[一般工业技术—材料科学与工程]