Моделирование событий в видеопотоке с помощью стохастических сетей Петри

Abstract

A model of the events on the basis of a modified Petri nets for events recognition task in the video stream. A new type of transition - transition to time stochastic modeling variativnoj duration of the event. On the basis of the relationship temporal logic suggested an idea of ​​temporal relations between the components of the event. For the probabilistic predictions of future events proposed to build a Markov chain based on Petri net mark. Testing of the proposed model based on a training sample.