Public security involves visual surveillance of mass crowd, in public facilities and mass gatherings. Crowd analysis depends heavily on the experience and concentration of the observer.Video monitoring by security personnel is a daunting task due to (i) the high number of camera`s (ii) human fatige and distraction and (iii) the complexity of the crowd behaviour.
This project aims at developing a crowd behaviour analysis system that can analyse abnormalities accurately and in real time and create an alert.

Runtime: 2018 - 2020