运动规划
能见度
计算机科学
树(集合论)
路径(计算)
机器人
人工智能
计算机视觉
可见性图
任务(项目管理)
实时计算
数学
工程类
地理
数学分析
气象学
正多边形
程序设计语言
系统工程
几何学
作者
Wen Xing,Aiguo Song,Lifeng Zhu
标识
DOI:10.1109/icra48506.2021.9560801
摘要
This paper proposes a new path planning strategy - the Rapid Visible Tree (RVT) algorithm to guide a robot to its goal in a complex environment without dangerous collisions. By fusing the visibility information with the classic tree-based searching method, RVT only takes the noisy points locally acquired from the environment as input and computes the visible region at each location to decide the growing direction of the path tree. Compared with traditional methods, RVT is more efficient, lightweight, and robust. We demonstrate that the RVT algorithm can not only complete the path planning task in real-time but also explore the unknown environment in simulated or real scenes.
科研通智能强力驱动
Strongly Powered by AbleSci AI