SISTEM PENGOLAHAN VIDEO UNTUK PENGAWASAN BENDA STATIS; VIDEO PROCESSING SYSTEM FOR STATIC OBJECT SURVEILLANCE

AMRULLAH, M. SHOFWAN, Faizah

2016 | Skripsi | FMIPA

The high number of motor vehicle thefts drives security efforts to always be improved. One of the method that have been done is to install CCTV cameras (closed circuit television) for surveillance. Furthermore, IP camera’s technology makes it possible to see images from CCTV cameras online. However, IP camera still needs humans for manual surveillance. Therefore we need a program that can help us to keep track of some static objects automatically. Program is built by using background subtraction and contour detection algorithm. The algorithms are applied plainly, background subtraction is used with the background model solely specified by the user, while contour detection is used with hierarchy of the detected object ignored. Then the surveillance were done by comparing each object that have been detected on each frame. There are variables in the program that can be configured manually by users to help program in detecting objects and detect the disappearance of the objects afterward. After the program is created, we can conclude that this program are able to keep track of some objects and put out a warning when the object disappear. Different environment and place, requires distinct configuration of variables, and no variable values that suits all conditions of surveillance. However, the program that has been made were very sensitive to illumination changes, because the value of learning rate is set to 0. This makes this kind of programs are not suitable for use in places that exposed to illumination changes, such as day and night transitions.

Kata Kunci : background subtraction, contour detection, object detection, object surveillance, CCTV