基于ISS-BSHOT特征的ICP点云配准方法  

ICP point cloud registration method based on ISS-BSHOT feature

作　　者：王帅帅 柏艳红[1] 王银[1] 孙志毅[1] WANG Shuaishuai;BAI Yanhong;WANG Yin;SUN Zhiyi(School of Electronic Information Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学电子信息工程学院,太原030024

出　　处：《扬州大学学报（自然科学版）》2022年第6期50-55,共6页Journal of Yangzhou University：Natural Science Edition

基　　金：山西省自然科学基金资助项目(201901D111247);山西省科技攻关计划资助项目(201903D121130).

摘　　要：针对点云配准处理过程中配准精度低且耗时长等问题,提出一种基于内部形态描述子(intrinsic shape signatures,ISS)关键点与二进制方向直方图描述子(binary signature of histograms of orientations,BSHOT)相结合的点云配准方法.首先,计算点云分辨率,采用ISS算法提取源点云与目标点云的关键点,并利用BSHOT算法描述关键点邻域,通过汉明距离匹配对应点对;其次,采用随机采样一致性算法删除匹配错误的对应点对,完成粗配准;最后,利用迭代最近点(iterative closest point,ICP)算法完成精配准.实验结果表明,该算法可在保证配准精度的同时显著提高配准效率.Aiming at the problems of low accuracy and long time consumption in point cloud registration process,a point cloud registration method based on intrinsic shape signatures(ISS)key points combined with binary signature of histograms of orientations(BSHOT)is proposed.Firstly,the point cloud resolution is calculated,and combined with the point cloud resolution,the ISS algorithm is used to extract the key points of source point cloud and target point cloud.The BSHOT algorithm is used to describe the neighborhood of the key points,and the corresponding point pairs are matched by Hamming distance.Secondly,the random sampling consistency algorithm is used to delete the matching error of the corresponding point pairs to complete the coarse registration.Finally,the iterative closest point(ICP)algorithm is used to complete the fine registration.Experimental results show that the algorithm can guarantee the registration accuracy and improve the registration efficiency significantly.

关 键 词：内部形态描述子 二进制方向直方图描述子 点云配准 迭代最近点 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]