An Automated Way of Vehicle Theft Detection in Parking Facilities by Identifying Moving Vehicles in CCTV Video Stream

Abstract

Security of parked cars against theft is a long existing concern where image and video processing can offer solutions. The car owners or parking lot operators are worried about having the vehicles stolen from parking lots, so they use CCTV cameras in parking lots to detect theft. The increased use of CCTV and video surveillance indicates their success as theft deterrents but a major drawback of the system is that non-automated human monitoring of vehicles can have human errors or lapses due to human fatigue. This paper presents an automated way of detecting vehicle theft as it happens. This procedure is based on moving object detection using Canny Edge Detection method and eventually notifying the security personnel or the parking lot operator about the movement. The first step is to detect the edges through Canny method and then finding the edge change ratio to finally determine a movement. Canny is one of the modern Edge detection techniques and choosing this one over other methods is because of the double thresholding and its better performance, which makes the method described in this paper efficient and useful. This paper proves the effectiveness of the described method.