On-line integration and reasoning of multi-sensor data to enhance infrastructure surveillance

Modern security systems employed in infrastructure surveillance applications include a set of different sensing technologies integrated by appropriate management systems. Such systems are still highly dependent from human operators for supervision and intervention. One of the challenging goals of the research community in this field is the automatic detection of both natural and malicious threat scenarios. In this paper we describe a framework for correlating events detected by sensor networks to provide early warning and decision support capabilities. To this aim, we propose an overall system architecture for the integration of the DETECT and SeNsIM frameworks: DETECT (Decision Triggering Event Composer & Tracker) is a new system which is able to recognize complex events by a model-based correlation of primitive events; SeNsIM (Sensor Networks Integration and Management) is a middleware for the integration of heterogeneous sensor networks. The paper also provides example application scenarios in the railway domain.