Towards Efficient Vehicle Classification in Intelligent Transportation Systems

Abstract

The classification of vehicles is an important task in Intelligent Transportation Systems (ITS) for applications such as analyzing traffic, checking for fraud, tracking targets, and other security applications. In the recent years, automated systems to recognize makes and models of oncoming vehicles are gaining attention, utilizing existing infrastructure of traffic cameras. To this end, we present an unexplored approach for vehicle make and model recognition (VMMR) and demonstrate its highly accurate and real-time performance, using a recently published real-world dataset. The encouraging results of our approach pave the way towards efficient large-scale and distributed vehicular surveillance in ITS.