流形和非流形方法对配准算法的影响分析  

Analysis and research on the influence of manifold and non-manifold methods on registration algorithm

作　　者：裴东 高文辉 任琪 但唐旭 PEI Dong;GAO Wen-hui;REN Qi;DAN Tang-xu(Eegineering Research Center of Gansu Province for Intelligent Information Technlolgy and Application,College of Physics and Electronic Engineering,Northwest Normal University,Lanzhou 730070,Gansu,China)

机构地区：[1]西北师范大学物理与电子工程学院/甘肃省智能信息技术与应用工程研究中心,甘肃兰州730070

出　　处：《西北师范大学学报（自然科学版）》2022年第1期37-43,共7页Journal of Northwest Normal University(Natural Science)

基　　金：国家自然科学基金资助项目(61961037)。

摘　　要：针对最近点迭代算法(Iterative closest point,ICP)求解步骤繁琐导致的迭代速率下降问题,提出了一种利用李群流形空间扰动的变换率求解雅可比矩阵的方法.该方法首先在点云配准前对点云数据进行随机降采样,其次使用K-D树搜索的方法进行2帧点云的搜索匹配,然后利用非流形的方法对ICP进行迭代求解,最后得到使用流形的方式优化雅可比矩阵的求解方式.在仿真和真实环境中进行了验证测试,文中方法相较于传统ICP迭代速率提升了57%,结果表明,所提方法极大地降低了点云配准消耗的时间,有效提高了ICP匹配效率.To solve the problem that the iterative rate decreases due to the complexity of the iterative closest point(ICP)solution,a method to solve the Jacobian matrix by using the transformation rate of Lie group manifold space perturbation is proposed.In this method,the point cloud data is randomly sampled before point cloud registration,then the K-D tree search method is used to search and match two frames of a point cloud,and then the non-manifold method is used to iteratively solve ICP.Finally,the solution of the Jacobian matrix is optimized by the manifold method.Verification tests were carried out in simulation and real environments,and the iteration rate of the proposed method was improved by 57%compared with the traditional ICP method.The results show that the proposed method can greatly reduce the time consumption of point cloud registration and effectively improve the ICP matching efficiency.

关 键 词：流形和非流形 迭代最近点算法 点云配准 雅可比矩阵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]