面向三维激光SLAM的动态物体检测与跟踪  

Dynamic Object Detection and Tracking for 3D Laser SLAM

作　　者：杨锦芳 李育文 YANG Jinfang;LI Yuwen

机构地区：[1]上海市智能制造及机器人重点实验室 [2]上海大学机电工程与自动化学院

出　　处：《计量与测试技术》2025年第3期98-102,共5页Metrology & Measurement Technique

摘　　要：在现实环境中,检测并剔除动态物体是移动系统实现自主导航、定位和地图生成的关键能力。因此,本文提出一种基于三维激光雷达的动态物体滤除算法,通过4DMOS区分动态物体和静态物体。同时,针对漏检的动态物体,引入单目标跟踪,将环境中的动态特征点尽可能剔除,并通过公开数据集,对算法的有效性进行验证,结果表明:该算法可提高SLAM系统在动态环境下的定位精度和建图效果。In the real environment,detecting and eliminating dynamic objects is the key ability of the mobile system to achieve autonomous navigation,positioning and map generation.Therefore,this paper proposes a dynamic object filtering algorithm based on 3 D LiDAR,which distinguishes dynamic objects from static objects by 4DMOS.At the same time,for missing dynamic objects,single target tracking is introduced,dynamic feature points in the environment are eliminated as much as possible,and the effectiveness of the algorithm is verified by exposing the data set.The results show that the algorithm can improve the localization accuracy and mapping effect of SLAM system in the dynamic environment.

关 键 词：动态物体检测 漏检补偿 单目标追踪 同步定位与建图 

分 类 号：G63[文化科学—教育学]