PhD forum: Decision-theoretic coordination and control for active multi-camera surveillance

In this thesis, we present novel decision-theoretic multi-agent approaches for controlling and coordinating multiple active cameras in surveillance. Decision-theoretic approaches models the interaction between active camera network and the uncertain surveillance environment effectively. The goal of the surveillance is to maximize the number of targets observed in active cameras with guaranteed image resolution. We enumerate the practical issues in active camera surveillance and discuss how these issues are addressed in our decision-theoretic approaches. The existing camera control approaches have serious limitations in terms of scalability in number of targets. Where as in our approaches, the scalability in number of targets has been improved by exploiting the structure and properties that are present in our surveillance problem. We proposed two novel decision-theoretic frameworks: Markov Decision Process (MDP) and Partially Observable Markov Decision Process (POMDP) frameworks for coordinating active cameras in fully observable and partially observable surveillance settings.