摘要
随着3D激光扫描技术的发展,点云数据的应用越来越广泛.然而点云配准一直是点云数据预处理过程中的一个关键问题.目前ICP算法是实现点云配准的主流算法.然而面对数据量大、噪声大的点云数据时,ICP算法在执行的配准效率和执行效果上不够理想.本文通过PCA算法,提取了点云数据集的方向向量,根据源数据与目标数据的方向向量,初步设定了旋转矩阵R的值.此外,定义了源数据与目标数据的曲面距离,在此基础上改进了传统的ICP算法.将改进后的ICP算法成功的应用到点云数据配准中来,提高了点云数据的配准效果,并压缩了算法的执行时间.
With the development of 3 D laser scanning technology,the application of point cloud data becomes increasingly wider.However,point cloud registration has always been a critical problem in point cloud data preprocessing.Now,The ICP is the popular algorithm for point cloud registration.However,in the face of large amount volume data or noise’s disturbance,the ICP is not ideal for registration efficiency and effectiveness.In this paper,the direction vectors of point cloud data set are extracted by PCA algorithm.And the value of rotation matrix R is preliminarily set according to the direction vectors of source data and target data.Besides,this paper defines the surface distance between the source data and the target data.On this basis,the traditional ICP algorithm is improved.The improved ICP algorithm is successfully applied to point cloud data registration,and improved the registration efficiency and compressed the running time in point cloud registration.
作者
乔世权
张坤
QIAO Shi-quan;ZHANG Kun(School of Information Science and Engineering,Hebei University of Science and Technology,Shijiazhuang 050018,China)
出处
《数学的实践与认识》
北大核心
2019年第8期135-143,共9页
Mathematics in Practice and Theory
基金
河北省教育厅青年基金(QN2014174)
关键词
点云数据
PCA算法
ICP算法
配准
point cloud data
PCA algorithm
ICP algorithm
registration
