Forecasting traffic signal intervals based on vehicle count

Traffic congestion is a critical problem in urban areas, leading to delays, increased pollution, and inefficient transport systems. Traditional traffic signal systems use static timings that fail to adapt to real-time traffic conditions. This paper proposes a smart traffic management system that dynamically forecasts traffic signal intervals based on real-time vehicle count data using machine learning algorithms. The system also integrates computer vision techniques to detect emergency vehicles and prioritize their movement. Ultrasonic sensors count vehicles, while an Arduino board and camera module interface with a central processing unit to analyze traffic patterns. Using algorithms like Linear Regression and Random Forest, the system predicts optimal green signal durations. Emergency vehicle detection using OpenCV ensures dynamic signal adjustments in critical situations. The proposed system demonstrates improved traffic flow, reduced waiting times, and enhanced emergency vehicle response efficiency, offering a scalable and cost effective solution for smart cities.