Vehicle Overtaking Detection Using Computer Vision Techniques

Traffic surveillance plays a crucial role in road safety and traffic management. This paper studies the application of Artificial Intelligence (AI) in developing a traffic surveillance system capable of detecting illegal vehicles overtaking on the road. The proposed method employs the YOLO algorithm for object detection, along with Deep SORT tracker for vehicle tracking. Canny edge detection, Hough transform are combined to be utilized for automated lane detection and point-line distance is used for overtaking violation identification. The proposed point-line distance approach works by calculating the perpendicular distance from the center point of a vehicle to the defined lane marking. A predefined distance threshold is set, allowing the system to determine whether a vehicle has crossed the lane marking, which indicates that the vehicle has performed illegal overtaking when condition is met. Upon detecting a vehicle as overtaking, the system will send an alert message to the traffic authorities to alert the authorities about the occurrence of an overtaking event. The main goal of the proposed system is to improve road safety and traffic management by addressing several challenges such as high accident rate and high costing. Thus, the system offers an efficient and cost-effective solution for traffic surveillance in detecting overtaking events on the road.