The project addresses automatic behavioral analysis through video processing, focused on crowded scenarios, for transportation hubs. A system performing human behavioral analysis, detaching what is usual from what is not, would fill the gap and provide a reactive, and hopefully pro-active control task, preventing terroristic attacks or crime situations in public place. In order to learn what is normal or not, we propose to use statistical inference enriched with contextual information.
E.g. , "starting to run" in an exit zone could be abnormal and suspicious, but becomes normal if the person is trying to reach a closing gate. We propose to apply the paradigm "learn-and-predict" -by modeling the normal activity in the hub with tools of automatic and semi-automatic classification and annotation, applying innovative methods of people tracking in crowd, and statistic pattern of activity recognition.

The new tools will be integrated in existing solutions and included in available CCTV systems, without the need of redesigning installed video surveillance systems on different scenarios, e.g. airports, harbors or railway station