Real-Time Surveillance System Using Pattern Matching

Police department in every country has been trying to decrease crime´s number, but in fact, the crime´s number always grows every year. Instead of catching the criminal, we can minimize the crime by lowering the opportunity of criminal´s action, so we propose the system to detect if there is someone entering the room by comparing pattern every couple of time. There are many pattern matching techniques, but the simple one is by counting its distance using Normal Euclidean. If there is a different pattern from previous pattern, the system will create an alarm. For this purpose, we use a capture application whose ability to capture an image periodically and the system that compare every image from previous capture. Testing result showed that our proposed system could detect every suspect object entering the area of our webcam or Closed-circuit Television (CCTV).