Title: Wavelet based intelligent techniques for automatic identification of motor vehicles: a comparative study

Authors: Vijaya Laxmi; D.K. Mohanta; B.M. Karan

Addresses: Department of Electrical and Electronics Engineering, Birla Institute of Technology, Mesra, Ranchi, Jharkhand 835215, India ' Department of Electrical and Electronics Engineering, Birla Institute of Technology, Mesra, Ranchi, Jharkhand 835215, India ' Department of Electrical and Electronics Engineering, Birla Institute of Technology, Mesra, Ranchi, Jharkhand 835215, India

Abstract: The vehicles of neighbouring states are more prone to violate Indian Motor Vehicles Act, 1988. An algorithm has been developed to automatically recognize the vehicles not registered in Jharkhand, a state of India for appropriate legal action using wavelet based intelligent techniques. The image of Vehicle Licence Plate (VLP) is extracted using Extended Hough Transform (EHT). Four features are obtained using four levels of decomposition using Haar wavelet and used as inputs to different classifiers. A comparative assessment regarding the efficacy of all classifiers indicates SVM to be the most suitable classifier for the intended automated VLP identification.

Keywords: EHT; extended Hough transform; MRA; multi-resolution analysis; SVM; support vector machines; vehicle licence plates; wavelets; intelligent identification; automatic identification; motor vehicles; vehicle identification; licence plate recognition.

DOI: 10.1504/IJVAS.2013.052273

International Journal of Vehicle Autonomous Systems, 2013 Vol.11 No.1, pp.42 - 61

Received: 13 Oct 2010
Accepted: 15 May 2011

Published online: 26 Feb 2013 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article