In collaboration with BAE Systems, CSIT leads a project on Video-based Semantic Analysis of Crowd Behaviour.  The project aims to deliver a generic automated video-based system for semantically analysing behaviour, normal and unusual, in both static and dynamic crowds, such as sports stadiums and railway stations respectively. The use of intelligent video analytics technology provides enhanced situational awareness to analysts in a safety control room / security operations centre.  Enabling staff to make timely decisions in the event of violent disorder, stampedes or terrorist attacks in public spaces. 

Central to this work is the ability to identify subjects and re-identify them as they move from one camera’s view into another non-overlapping camera’s field of vision. Seminal work in this area was published by CSIT researchers[1] in 2017.

 

[1] Video Person Re-Identification for Wide Area Tracking based on Recurrent Neural Networks. / McLaughlin, Niall; Martinez del Rincon, Jesus; Miller, Paul.

In: IEEE Transactions on Circuits and Systems for Video Technology, 26.07.2017.