A technique of deviant behaviour detection in video surveillance systems based on complex behaviour analysis

Abstract

This paper deals with deviant behavior detection using video surveillance systems. Deviant behaviour detection is a complex problem and includes not only the recognition of the abnormal patterns of individual's behaviour, but also the problem of general methodology development, which consists of deviant behaviour markers and techniques to identify them. A new complex approach to the object behaviour analysis is presented in this paper as well as the integrated distributed classifier generic architecture is described. The novelty of this research is the consequent subject analysis based on a complex of potentially deviant and socially dangerous behaviour attributes.