智能科学与技术学报2023，Vol.5Issue(4)：477-485,9.DOI:10.11959/j.issn.2096-6652.202342
动态环境下基于语义信息与几何约束的视觉SLAM系统
Visual SLAM based on semantic information and geometric constraints in dynamic environment
摘要
Abstract
Most existing visual SLAM systems assume that the external environment is static,ignoring the influence of dynamic objects on the SLAM system.This assumption largely affects the accuracy and robustness of autonomous naviga-tion.To address this issue,a dynamic SLAM system was proposed,which combined semantic information based on ob-ject detection and geometric information from multi-view geometry constraints by defining and discriminating the dy-namic feature points in the system based on the moving probability.Experiment results on the public TUM dataset and our robot in real environment showed that,when comparing with ORB-SLAM2,the absolute trajectory error could be re-duced larger than 94%,and the average relative position and attitude errors were reduced at least 41%and 40%,respec-tively,in high dynamic environments.It means that the proposed SLAM system effectively removes dynamic feature points,thus improving the localization accuracy and robustness of the visual SLAM system within high dynamic environ-ments.关键词
动态SLAM/深度学习/目标检测/极线几何Key words
dynamic SLAM/deep learning/object detection/epipolar geometry分类
信息技术与安全科学引用本文复制引用
李嘉铭,解明扬,张民,王从庆..动态环境下基于语义信息与几何约束的视觉SLAM系统[J].智能科学与技术学报,2023,5(4):477-485,9.基金项目
启元实验室创新基金项目(No.S20210201102) （No.S20210201102）
南京航空航天大学科研与实践创新计划(No.xcxjh20220342)The Qiyuan Lab Innovation Fund Project(No.S20210201102),Research and Practice Innovation Program of NUAA(No.xcxjh20220342) （No.xcxjh20220342）
