A reasoning engine for intruders' localization in wide open areas using a network of cameras and RFIDs

Wide open areas represent challenging scenarios for surveillance systems, since sensory data can be affected by noise, uncertainty, and distractors. Therefore, the tasks of localizing and identifying targets (e.g., people) in such environments suggest to go beyond the use of camera-only deployments. In this paper, we propose an innovative system relying on the joint use of cameras and RFIDs, allowing us to “map” RFID tags to people detected by cameras and, thus, highlighting potential intruders. To this end, sophisticated filtering techniques preserve the uncertainty of data and overcome the heterogeneity of sensors, while an evidential fusion architecture, based on Transferable Belief Model, combines the two sources of information and manages conflict between them. The conducted experimental evaluation shows very promising results.