A Cost Effective Approach to Real Time Video-Surveillance of Outdoor Scenes

Abstract

Real time interpretation of outdoor scenes for video-surveillance applications is a complex task. These kind of systems are supposed to work day and night, 365 days per year. under changing environmental conditions, variable weather and lighting conditions, with a minimum ratio of false alarms. Video-surveillance systems should be able to detect moving objects of different size moving at different speeds, and provide this information in real time to those persons who may need it at a reasonable price. This paper describes a system which provides a cost-effective solution to these kinds of problems based on the idea of processing sequences of images in real time.