Vehicle Classification under Cluttered Background and Mild Occlusion Using Zernike Features

Vehicle recognition and classification in a multi-environment containing cluttered background and occlusion is an important part of machine vision. The goal of this paper is to build a vehicle classifier that identifies a “car” vehicle from “non-car” amidst complex environment taken from university of Illinois at Urbana-Champaign (UIUC) standard database. The image is divided into sub-blocks of equal size without any pre-processing. The zernike moment features are extracted from each sub-block. The features of the objects are fed to the back-propagation neural classifier after normalization. The performance is compared with various categories of blocking models. Quantitative evaluation shows improved results of 85.2%. A critical evaluation of this approach under the proposed standards is presented.