I. Introduction
Vehicle detection and classification is a paramount activity in both civilian [1], [2] and military [3] transportation system for surveillance application. This information enhances the pedestrian’s safety [4] and restricts the movement of unauthorized vehicles at security checkpoints [5]. The passive acoustic modality based moving target classification is a challenging task, particularly in the outdoor environment [6]. The moving vehicle produces a wide-band signal, which is the combination of many source factors i.e., engine noise, tyre noise, exhaust noise, body vibration, environmental and Doppler noise effect [7], [8]. Apart from these challenges, the acoustic modality has following advantages like day-night workability, cost effective, and less sensitive from the terrain which inspires researchers towards this emerging field [9], [10].