Distributed deployment algorithms for robotic visual sensor networks in non-convex environment

This paper considers the deployment problem for mobile visual sensor networks equipped with omni-directional communication capability in a non-convex environment. The performance of the visual sensor is assumed to depend not only on the distance but also on the orientation from the sensor to the target. A gradient-based distributed deployment algorithms is designed to maximize the joint detection probabilities of the events in the region of interest by the sensors which introduces discontinuities caused by the obstacles. In addition, a potential function based algorithm is incorporated in order to guarantee the collision avoidance during the deployment of the sensors. The results are validated through numerical simulations.