Traffic Signal Violation Detection System

Abstract

Traffic signal violations are a major cause of accidents and traffic congestion. This project presents an automated Traffic Signal Violation Detection System using Deep Learning-based Object Detection. The system leverages SSD MobileNet V1, a pre-trained Convolutional Neural Network (CNN), to detect and classify traffic signals in real-time. Using the TensorFlow Object Detection API, the model identifies traffic lights and determines violations based on detected signals. The approach integrates image processing, real-time object detection, and violation recognition, providing an intelligent traffic monitoring solution. The proposed system enhances road safety, reduces human intervention, and supports smart city traffic management systems.