Visual sensor networks—adaptive online configuration of surveillance networks with distributed smart cameras

Abstract

Distributed smart cameras have been increasingly employed to capture dynamic events for tasks such as surveillance and training. When events distribute throughout a large area, many issues may occur, two of which are inevitable and critical, that is, limited computational capabilities and battery energy constraints. These two issues bring both challenge and opportunity to researchers all over the world. Different approaches for camera configurations are presented. In this chapter, we propose an adaptive online configuration for large-scale camera networks. By exploiting tracking-inspired local search and cluster-based cooperating coverage, a distributed coverage-probability-based heuristic algorithm (DCPBHA) is designed with the consideration of limited camera computational capacities and energy constraints. Furthermore, a point corresponding method is built to implement the dynamic configuration, which can minimize the overall energy consumption and balance energy distribution among camera nodes. Simulations show the camera networks can response to dynamic events quickly and last for a longer lifetime, which demonstrate the effectiveness of our proposed methods.