License Plate Recognition Part II: Wavelet Transform and Euclidean Distance Method

• Main Authors: Brahim Belhaouari Samir, BBS, Daud, Hanita, raja, syahira, aissa , boudjella
• Format: Conference or Workshop Item
• Published: 2012
• Subjects: QA Mathematics
• Online Access: http://eprints.utp.edu.my/8617/1/ICIAS_Samir.pdf; http://eprints.utp.edu.my/8617/

In this paper, we have developed car licence plate recognition for parking system, based on symelets 8 wavelet transform for features extraction and k-Nearest Neighbours for classification. The system is implemented and simulated in MATLAB, and its performance is tested on real image using Euclidean distance method. The proposed system consists of three steps:1) image acquisition module which captures the image of a moving car regions, 2) image processing module for extraction and classification, and 3)database that handles the car information. This method gives better recognition rate of 96% at the threshold Tc=0.4. More importantly, the present results show that the number of wavelet coefficient decreases as the threshold increases. Thus, the computation time can be reduced considerably which will make this method faster and robust. Increasing the level from 5 to 6, does not affect significantly the accuracy and the computation time.