Traffic Density Estimation for Traffic Management Applications Using Neural Networks

Abstract

Traffic density is one of the elemental variables used in molding road traffic kinetics. Current density estimation techniques include loop detectors and sensors which are dependent on the crowd-sourcing of traffic data, which suffers from limited coverage and high cost. This article proposes a unique method to estimate traffic density based on neural network and mathematical modelling which uses surveillance feed from cameras. The proposed method can save both transportation costs and journey time, thus helping in better traffic management. The result analysis shows that the proposed method works well for varying traffic flow conditions and dynamic conditions.