Cloud-computing-based framework for multi-camera topology inference in smart city sensing system

Abstract

This paper proposes a cloud-computing-based algorithmic framework which is scalable and adaptive to online smart city video sensing system. One of the most cost-expensive works in such a system is to infer the topology structure of video camera network, thus spatio-temporal relationship inference for large-scale camera network is simulated on a cloud-computing platform to validate the proposed framework. The simulation results and time complexity analysis demonstrate the effectiveness and scalability of the proposed approach.