摘要
针对单线激光雷达进行特征提取时存在对噪声敏感和稳定性差等问题,提出一种基于前瞻窗口的单线激光雷达特征提取方法。该方法首先使用双边滤波滤除噪声,保留了激光点云边缘特征信息;使用哈里斯(Harris)角点检测提取特征角点;并通过带有前瞻窗口的区域搜索方法完成聚类。最后将方法应用到实际的基于机器人操作系统(Robot Operation System,ROS)的移动机器人环境特征提取实验中,实验结果表明,该方法能有效提取环境特征,且有较好的稳定性和准确性。
Aiming at the problems of noise sensitivity and low stability in feature extraction of laser ranger finder,a feature extraction method based on forward-looking window is proposed.This method uses bilateral filtering to filter out noise and retains the edge feature information of the laser point cloud.And the method extracts characteristic corners by Harris corner detection method,besides the method completes the clustering by the area search method with a preview window.Finally,the method is applied to the actual ROS-based mobile robot environment feature extraction experiment.The experimental results show that the method can effectively extract environmental features,and which has good stability and accuracy.
作者
马小陆
巩朝光
王兵
郑睿
Ma Xiaolu;Gong Chaoguang;Wang Bing;Zheng Rui(School of Electrical and Information Engineering,Anhui University of Technology,Maanshan 243002,China;School of Physics and Electronic Information,Anhui Normal University,Wuhu 241002,China)
出处
《湖南文理学院学报(自然科学版)》
CAS
2021年第3期49-57,共9页
Journal of Hunan University of Arts and Science(Science and Technology)
基金
国家自然科学基金(61472282)
安徽省科技重大专项(202003a05020028)
安徽省教育厅高校科学研究重大项目(KJ2019ZD05)
安徽高校自然科学研究重点项目(KJ2019A0065)
特种重载机器人安徽省重点实验室开放课题资助项目(TZJQR004-2020)。
关键词
机器人
特征提取
前瞻窗口
双边滤波
robot
feature extraction
preview window
bilateral filtering