Traffic Prediction for Intelligent Transporta...

Traffic Prediction for Intelligent Transportation System

Author Name : A. Sriram, S. Rashpal Singh, M. Sai Harshit, N. Satya Suhas

ABSTRACT

People in today’s era usually have the tendency of using their own private vehicles for commutation rather than using public transit and this result in large number of private vehicles on road. It leads to traffic congestion at every roads. In such scenario one cannot restrictindividual to limit the usage of their private vehicles but we can able to manage traffic flow in a way that it doesn’t alleviate congestion issues. The traditional traffic management approach worksefficiently only if the traffic is less, but if the density of vehicles on a particular side of road increases on one side than other side, this approach fails. Hence, we aim to redesign the traffic signal system from static switching to dynamic signal switching, which canperform instant-timesignal monitoring and handling. There are many projects emerging in order to convert the current transport system of cities to ‘Smart system’, by introducing Intelligent Transport System. Many initiatives are taken to design a system that can perform instant monitoring of traffic signals i.e., the traffic signal switching time will depend on the count of vehicles on each side of the road instead of predefined switching time. The switching time of signal will be decided based on vehicle detection in day-today traffic scenarios with good accuracy. This practice can prove its effectiveness in releasing the congested traffic at an efficient and faster rate.

Keywords: Traffic prediction, Real-time data, historical data, traffic YOLO, deep learning,soft computing