SERKET: an open software platform for preventive security in public crowded places and for large events

摘要

Security of public infrastructures and for large events (cultural, sporting) traditionally relies on monolithic video surveillance systems, where security operators are quickly overloaded with information coming from multiple sources, displayed on multiple monitoring screens. Indeed, in such crowded places, it is quite impossible for operators to detect risk-prone situations in real-time and to react accordingly before such situations degenerate: they can only act after an alarm is triggered; surveillance is said to be passive. Conversely, the European project SERKET of the seventh ITEA call (EUREKA program), involving twenty-four industrial and academic partners, aims at developing an innovative software approach where dispersed data coming from heterogeneous sensors - cameras, microphones, human beings, as well as badge readers, IR barriers for intrusion detection, etc. - are automatically processed (signal processing), combined (information fusion) and analysed (threat assessment) so as to provide security personnel with the right information at the right time: security then becomes preventive. The objective of this paper is to glance at the innovations of the SERKET security system, and to focus on key layers of intelligent signal processing, advanced information fusion and situational awareness rendering.