License plate detection using discrete wavelet transform

Abstract

License Plate Detection (LPD) technique is developed and applied for purpose of facilitating the surveillance, law enforcement, access control and intelligent transportation monitoring with least human intervention. It occupies a significant importance as an intelligent management and monitoring system in real time. In this research work a novel edge detection paradigm is envisaged to work for various styles of license plate images. Wavelet transform is considered as landmark in the field of edge detection due to the feature that it represents a signal in terms of functions those are localized in both frequency and time domain. A wide range of research on LPD systems are available in literature. But performance of these existing algorithms fails due to slow processing time, changing illumination level and background conditions. Considering condition of number plates in Asian countries make it even harder for the exiting systems to show promising results because of the non-standard plate designs being followed. Fourier transform being global in nature can neither localize sharp transients nor differentiate between true and false edges. Similarly, the classical edge detectors do not yield adequate edge maps of the images. The purpose of this work is to develop an algorithm that must meet the odds by applying methodology for LPD based on two-dimensional discrete wavelet transformation. The performance of proposed scheme is evaluated using different discrete wavelet families which are Haar wavelet, Daubechies wavelet, Coifman wavelet and Symlet wavelet in terms of PSNR, processed time and computational load. The proposed methodology promises a fast processing time with better performance in edge detection as compared to the exiting algorithms for Asian condition.