The strong need for smart surveillance systems stems from security-sensitive areas e.g. banks. Recent world events also demand practical and robust deployment of video-based solutions for a wide range of scenarios, such as public transport areas. Surveillance cameras are prevalent in many areas with camera outputs usually recorded in tapes or stored in video archives. Traditionally this video data is only used as a forensic tool. The main limitation in the effectiveness of many surveillance systems is the cost of providing adequate human monitoring cover for what is, on the whole, a fairly tedious job.
What is needed is a real-time reactive analysis of surveillance to alert security officers to a burglary in progress or to a suspicious individual wandering about the premises. Such a system would improve personal security and safety in public places such as public transport systems. Visual systems can deal with and monitor complex human movements. Human motion analysis concerns the detection, tracking and recognition of people from image sequences involving humans.
The aim of this project is to develop a visual surveillance system, capable of detecting and tracking people, monitoring their activities and thus identifying individuals through their actions.