采用特征预处理ICP算法的机器人运动环境建图  

Map Creation of Robot Motion Environment Using Feature Preprocessing ICP Algorithm

作　　者：申鸿 李平[1] 贾丙佳 SHEN Hong;LI Ping;JIA Bingjia(College of Information Science and Engineering,Huaqiao University,Xiamen 361021,China)

机构地区：[1]华侨大学信息科学与工程学院,福建厦门361021

出　　处：《华侨大学学报（自然科学版）》2022年第2期229-236,共8页Journal of Huaqiao University(Natural Science)

基　　金：国家自然科学基金资助项目(61603144);福建省自然科学基金资助项目(2018J01095);福建省高校产学合作科技重大项目(2013H6016);华侨大学中青年教师科技创新资助计划项目(ZQN-PY509);华侨大学研究生科研创新基金资助项目(18014082046)。

摘　　要：针对机器人运动环境建图中迭代最近点(ICP)算法的扫描配准过程存在关联点对对应困难、迭代初值要求高的问题,首先,提出一种特征预处理的粗配准方法,以保证配准过程的迭代初值较小;然后,基于扫描点过滤思想对粗配准后存在较大误差的数据进行过滤,剔除测量噪声,提高点对关联准确率.结果表明:经过特征预处理的粗配准及噪声剔除后的改进ICP算法能够有效地进行扫描配准,解决机器人运动环境建图存在的问题.For the scanning registration process of iterative closest point(ICP)algorithm in map creation of robot motion environment,there are some problems,such as difficult correspondence of correlation point pair and high requirements for the iterative initial values,firstly,a coarse registration method with feature preprocessing is proposed,in order to ensure the iterative initial values of the registration process are relatively small.Then,based on the idea of scanning point filtering,the data with relatively large errors after coarse registration are filtered to eliminate measurement noise and improve the correlation accuracy of point pair.The results show that the improved ICP algorithm after coarse registration of feature preprocessing and noise elimination can effectively perform scanning registration,and solve the existing problems in map creation of robot motion environment.

关 键 词：建图 特征预处理 迭代最近点算法 移动机器人 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]