摘要
为了使自主移动机器人在结构化和半结构化环境中能快速有效地提取道路的可行区域,采用全局搜索及双阈值的算法.该算法首先采用基于八邻域的全局搜索法搜索激光数据点,再结合角度和高度差双阈值对数据点进行归类并检测道路边界,最后利用障碍物检测原理获取障碍物.实验结果表明:该算法能够检测出路边及障碍物边界,此过程只对机器人感兴趣区域的数据点进行检测,从而能够提取出路面上的障碍物与可行区域.该算法具有实用性好、可行性高、使用范围广等优点,能够为自主移动机器人提供安全可行的区域,进而为路径规划提供参考依据.
An algorithm of global searching and dual thresholds is proposed, which can be used to extract feasible motion region quickly and effectively for autonomous mobile robot in structured or semi-structured road environment. Firstly, the algorithm uses global searching approach of the 8- neighbor to collect laser scanning data. Secondly, by combining dual thresholds of angle and height, different laser scanning data are categorized and road boundary is detected. Finally, obstacles are obtained according to the obstacle detection principle. The experiment results show that the algorithm can extract obstacles and road outline. This process is suitable for detecting the data of the region which the robot is interested in. And then, obstacles and feasible motion region are extracted. The algorithm has the advantages of good practicability, high feasibility and wide application scope. The algorithm can provide safe and feasible motion region for autonomous mobile robot, and furnish information for path planning.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2011年第B09期88-92,共5页
Journal of Southeast University：Natural Science Edition
基金
国家自然科学基金资助项目(60675043)
浙江省自然科学基金资助项目(GK090906011
Y1090426)
浙江省科技计划基金资助项目(2007C21051)
关键词
可行区域提取
激光测距仪
八邻域搜索
双阈值
feasible motion region extraction
laser range finder
8-neighbor searching
dual thresh old
