Computer vision based traffic monitoring system

Abstract

This research paper explores the application of OpenCV (Open-Source Computer Vision) technology in the field of traffic management. With the increasing number of vehicles on roads worldwide, efficient traffic management systems are crucial to ensure smooth and safe traffic flow. The paper begins by introducing the concept of OpenCV, a powerful open-source library widely used for computer vision tasks. It highlights the versatility and robustness of OpenCV in various applications, including traffic management. The research focuses on utilizing OpenCV algorithms for real-time traffic analysis, vehicle detection, tracking, and speed detection. To achieve accurate and reliable results, the research incorporates advanced computer vision techniques such as edge detection, image segmentation, feature extraction, and machine learning. A comparative analysis has been made to find out the best results among different techniques. Firstly, vehicle detection is done using image processing technique and result calculated. Secondly, the system is tested with YOLO V8 which is a pretrained machine learning model. Result shows system developed with YOLO V8 provides much better accuracy and reliability. We got 97% accurate result while doing this research paper using YOLO V8. Thus, whole system is developed with this approach to detect and manage traffic system.