Over Speed Detection and Number Plate Recognition

Abstract

In an era of rapidly increasing vehicular traffic, ensuring road safety and environmental sustainability has become a critical challenge. This paper presents an innovative and comprehensive solution for vehicle over-speed detection and traffic monitoring, titled "Integrated Vehicle Speed and Number Plate Detection System Using Machine Learning and IoT". The proposed system combines the power of machine learning, computer vision, and sensor-based IoT technologies to deliver a dual-purpose solution that enhances road safety and environmental awareness. The system integrates high-resolution CCTV cameras with deep learning models such as YOLOv8/SSD for real-time vehicle detection, coupled with speed estimation by analysing consecutive video frames. Upon detecting over speeding, Automatic Number Plate Recognition (ANPR) is performed using Haar Cascade/YOLO models and Optical Character Recognition (OCR) tools like Tesseract or Easy OCR to identify the vehicle. Simultaneously, an Arduino-based embedded system employs infrared (IR) sensors to monitor vehicle speed, ultrasonic sensors to ensure safe proximity by measuring the distance from obstacles, and gas sensors to assess carbon monoxide emissions. When violations such as over speeding or high pollution levels are detected, alerts are triggered via buzzers or LEDs for immediate notification. This integrated solution maintains a comprehensive log of violations, generates real-time alerts for enforcement authorities, and promotes both traffic law compliance and emission control. Designed for scalability and easy implementation, the system holds immense potential for integration into smart city infrastructures, offering a significant step forward in automated traffic control, accident prevention, and environmental protection.