Adaptive Traffic Control Systems

Abstract

This project presents an AI-driven Adaptive Traffic Control System (ATCS) that optimizes traffic signal timing using real-time video analytics and vehicle density estimation. The system captures live road footage, processes vehicle flow patterns and lane-wise congestion levels, and extracts meaningful traffic indicators such as queue length, vehicle count, and average waiting time. A machine learning model analyzes these features to dynamically assign optimal green-time durations and classify intersections into Low, Moderate, or High congestion states. Implemented using Streamlit and Python, the system generates a detailed PDF report summarizing intersection performance and predicted traffic load.