Distributed resource-aware task assignment for complex monitoring scenarios in Visual Sensor Networks

Smart camera networks usually have limited resources. This includes not only processing power and memory but in many applications also the amount of available energy. To find a trade-off between available resources and the current requirements of the application, the configuration of the network is very important. Due to rapid changes in the visual sensor network's environment, reconfiguration must be dynamic and performed online. The reconfiguration must also be able to deal with heterogeneous tasks assigned within the network where the tasks require varying levels of dynamic actions by the cameras. In this paper we present a fully distributed algorithm for a combined spatial coverage and object handover problem. In our approach we focus on (i) assigning heterogeneous tasks to cameras in order to cover a certain area of interest (ii) hand over moving objects between cameras to allow seamless tracking and (iii) perform (i) and (ii) in a combined fashion to maximize the overall quality of surveillance while minimizing the resources required during operation.