An ensembled neural network classifier for vehicle classification using ILD

Abstract

Vehicle classification is required to study various parameters related to traffic. It is impossible to estimate the density of vehicles, number of vehicle types etc. without vehicles classification. This paper reviews various neural network algorithms using single loop detector that can be used in real time traffic management system to classify vehicles. Since it is evident that neural network is a weak classifier, we propose a model which uses an ensembled neural network algorithm using ILD to achieve high speed and accuracy than the traditional method. © 2012 Springer-Verlag.