Intelligent Traffic Light Control Using YOLO for Vehicle Detection and Flow Optimization

Abstract

Traffic congestion is fast becoming one of the most important problems as cities grow in population and cars. Traffic congestion not only increases fuel use and air pollution but also adds time and stress to drivers. Though it seems to be everywhere, megacities are the most impacted. This calls for real-time road traffic density calculation to improve signal control and efficient traffic management. Traffic controllers are among the key determinants of traffic flow. Traffic control optimisation is therefore more necessary to better fit this growing need. The paper estimates traffic density over all lanes using traffic cameras and YOLO object identification algorithms, then modifies red and green signal timing.