LEARNING THE DISTRIBUTION OF OBJECT TRAJECTORIES FOR EVENT RECOGNITION

The advent in recent years of robust, real-time, model-based tracking techniques for rigid and non-rigid moving objects has made automated s urveillance and event recognition a possibility. A statistically based model of object trajectories is presented which is learnt from the ob servation of long image sequences. Trajectory data is supplied by a tr acker using Active Shape Models, from which a model of the distributio n of typical trajectories is learnt. Experimental results are included to show the generation of the model for trajectories within a pedestr ian scene. We indicate how the resulting model can be used for the ide ntification of atypical events.