Real-time Traffic Surveillance and Detection using Deep Learning and Computer Vision Techniques

This research paper presents a web app-based ML/DL model that performs traffic surveillance and detection with high accuracy and efficiency. The paper aims to improve the existing traffic system in India by proposing a Deep Learning/Computer Vision based solution for traffic surveillance and detection. The system includes four models: helmet detection, license plate detection, vehicle classification, and speed detection, implemented on a web app with three detection options. YOLOv5 and YOLOv7 are employed, achieving state-of-the-art performance in detecting objects, and evaluation on a large dataset shows accurate and efficient real-time detection, making it a valuable tool for traffic flow analysis. The proposed system demonstrates the potential of deep learning in developing advanced traffic surveillance and detection systems, enhancing the overall traffic management system in India. In conclusion, the proposed Deep Learning/Computer Vision based solution offers an efficient and accurate approach to traffic surveillance and detection, addressing the challenges faced by the current traffic system in India. The use of YOLOv5 and YOLOv7 demonstrates the potential of deep learning techniques in developing advanced traffic surveillance and detection systems, which can enhance the overall traffic management system in India.