A neural network based intelligent intruders detection and tracking system using CCTV images

Abstract

This paper reports the development of a neural network based intelligent intruders detection and tracking system using closed-circuit television (CCTV) images. It examines the techniques and algorithms used to identify a potential intruder and methods to eliminate other non-threatening objects. Once the presence of an intruder is determined, the object will be monitored and tracked. The tracked information can be used to further identify any suspicious behaviour in the sparse and complex environments. The traditional approach to intelligent scene monitoring (ISM) is examined and compared with the artificial neural network (ANN) approach. The ANN approach demonstrates how a system can learn how to distinguish suspicious movements from non-suspicious movements. The proposal has a potential to be used as an intelligent surveillance system.