监控系统中的多摄像头协同算法

给出了一个分布式多摄像头监控系统结构,描述了一种使用路径模型和目标外形变化估计的数据融合方案,提出了一种新的在分布式监控系统中的多摄像头协同算法(PDA,PriorityandDistanceAlgorithm)。提出的算法基于多摄像头数据融合结果,根据任务优先级、目标与摄像头之间的距离及目标的可见性情况,分配摄像头给目标,其特点是使系统中有高优先级并距离摄像头最近的可见目标优先分配摄像头。实验结果表明提出算法能协同多摄像头可靠地跟踪人。

Algorithm for Cooperative Multi-camera Surveillance

An architecture of distributed multi-camera surveillance system is given,a scheme of data fusion that uses path modeling and change estimation of object appearance is described,and a novel algorithm for cooperative multiple cameras in distributed surveillance system is proposed.Based on data fusion,this algorithm allocates cameras to objects according to tasks' priority,the object-to-camera distance and the visibility of objects.It attempts to assign cameras to the object with high priority and the nearest distance to the camera,which can be seen without occlusion.Experiment results show that the proposed algorithm can coordinate cameras to track people reliably.