Border Intruder Detection System Using Digital Image Processing Techniques

Now a day, it has become very necessary to secure each and every public place, like stations, bus stands and also other general areas from intruders. For this video surveillance is playing a very important role to secure the areas. In our paper, we present an intelligent framework to detect the multiple events in Region of interest with this video surveillance. From the region of interest we get a set of variables, create classes like humans or vehicles, attributes of the classes like speed or locality, and create notions to detect and understand the activities in the circumstances. Now all the knowledge got from it is used to understand the any complex activities of the objects. The objects may be pet, human or any vehicle and are detected on the basis of ratio of height and width given by horizontal and vertical clamps. After calculation if objects falls into range than a boundary box is set around that object and a signal is given to the users. Finally the proposed algorithm is giving very good results and intruders are found out and marked by a boundary box in the frames given by video of CCTV or any camera footage.