A Simulation-Based Approach to Evaluate the Performance of Automated Surveillance Camera Systems for Smart Cities

Abstract

A surveillance camera is the typical device that makes up a surveillance camera system in a modern city. It will still be a representative surveillance unit in future scenarios such as in smart cities. Furthermore, as the demand for public safety increases, a massive number of surveillance cameras will be in use in the future, and an automated system that controls surveillance cameras intelligently will also be required. Meanwhile, installing a surveillance system without any verification system might not be cost-effective, so a simulation that evaluates the system’s performance is required in advance. For this reason, we introduce how to simulate a surveillance area and evaluate surveillance performance in this paper to assess a surveillance system consisting of large amounts of surveillance cameras. Our simulator defined the surveillance area as a pair of two-dimensional planes, which depend on various camera-related configurations. Both surveillance areas are used to determine if the moving object belongs to the coverage of a surveillance camera. In addition, our simulator adopts several performance indices to evaluate a surveillance camera system in terms of target detection and quality. The simulation study provides comprehensive results on how various components of the surveillance system affect the performance of the surveillance system, leading to the conclusion that building a sophisticated scheme to control a large number of surveillance cameras can provide a cost-effective and reliable surveillance system for smart cities.