Intelligent Traffic Management System.

Abstract

The importance of traffic signals is increasing owing to the drastic increase in population. Ensuring road safety is of high priority. In this project, we introduce an Intelligent Traffic Management System (ITMS) capable of managing traffic of varying densities, without the need of a traffic warden to physically monitor a particular intersection. This system is designed to retrieve the live traffic feed from a junction and process the same using the TensorFlow Object Detection API over OpenCV to detect the severity of the traffic based on the number of vehicles detected. Upon determining the number of vehicles, the corresponding signal, based on the traffic intensity is given. (More vehicles detected – Green light for longer duration and vice versa. ) Thus, this system dynamically adapts to the prevailing traffic conditions and grants the corrosponding traffic light sequence for the required duration to maximize the flow of vehicular traffic. The system is designed to ensure smooth traffic flow by decreasing the wait period of vehicles at intersections and automates the process of controlling traffic signal.