Skip to Main Content
In the context of public place surveillance, the evaluation of available technologies shows that the security bottleneck isn't the surveillance hardware, but rather the real-time analysis and correlation of data provided by various sensors. Also, there is an evident lack of global threat management policy. In this paper, we present an implementation of an event based inference technology called complex event processing, which allows fusion of information generated by heterogeneous sensors supporting the goal of providing a global situational view for public place surveillance. We also present a description logic paradigm to support a non programmer in the situation modeling task. An example of panic crowd detection in an airport illustrates the complete process.