摘要
提出了一种多动态目标的监控模型,设计了目标监控优先级的计算方法和动态扫描队列的实现算法,并利用粗糙集理论的约简算法求得最小目标监控规则集,实现了仅利用单个PT(pan-tilt)摄像机对多个动态目标的实时监控。所进行的实际机器人动态环境实验,验证了提出方法的有效性和实用性。通过对实际和计划监控几率的比值分析说明了监控策略的合理性。
A multi-object monitoring model is proposed,with designs of algorithm and for the method of deployed cameras' dynamic scanning.A minimum object monitoring rule-set is derived from a simplified algorithm of the rough-set theory.It brings about simultaneous real time multi-objects monitoring with a single PT (pan-tilt) video camera. Through a dynamic environmental experiment with deployed robots, the efficiency and practicability of the strategy is proved. The monitored ratio out of the succeeded monitoring multi-objects vs planned ones and the analysis thereof show that the monitoring strategy is rational.
出处
《中国民航学院学报》
2004年第6期1-4,10,共5页
Journal of Civil Aviation University of China
基金
国家高技术发展计划资助项目(863-2002AA735041).
关键词
自主移动机器人
多目标监控
视觉伺服
粗糙集
autonomous mobile robot
multi-object monitoring
visual servoing
rough set
