现代电子技术2024，Vol.47Issue(1)：177-180,4.DOI:10.16652/j.issn.1004-373x.2024.01.031
基于改进3D-NDT机器人自定位算法
Robot self-localization algorithm based on improved 3D-NDT
许振权 1徐红伟1
作者信息
- 1. 中国计量大学 机电工程学院, 浙江 杭州 310000
- 折叠
摘要
Abstract
In the process of robot self-localization,the traditional 3D-NDT point cloud registration has the defects of poor registration effect,large error and time-consuming registration without giving the initial rotation matrix,so a relatively efficient improved 3D-NDT point cloud registration algorithm is proposed.ISS feature points are extracted from the input point cloud,and the fast point feature histogram(FPFH)of these feature points is calculated first,and then the feature points are matched according to the direct correspondence estimation.The incorrect correspondence relationship is removed by RANSAC(random sample consensus)to obtain the initial rotation matrix.The obtained initial rotation matrix is substituted into the 3D-NDT algorithm for matching and the final matching results are obtained.The scene point cloud in the indoor and outdoor is used for the test.The experimental results show that the improved 3D-NDT algorithm can output better matching results,and the matching accuracy is improved.However,the algorithm is complex and needs to be further optimized.关键词
点云配准/ISS特征点/机器人/自定位/特征直方图/场景点云Key words
point cloud registration/ISS feature point/robot/self-localization/feature histogram/scene point cloud分类
信息技术与安全科学引用本文复制引用
许振权,徐红伟..基于改进3D-NDT机器人自定位算法[J].现代电子技术,2024,47(1):177-180,4.
