License Plate Recognition Based on K-Means Clustering Algorithm

Abstract

The stolen vehicles are tracked by License Plate Recognition (LPR) system. In image processing technique LPR is used to identify vehicles by their license plates. LPR used in traffic and other various security applications. In this work, LPR tracking system using K-Means (KM) clustering algorithm and Optical Character Recognition (OCR) technique is discussed. LPR system includes pre-processing using median filter, KM segmentation, binarization of KM segmented image; characters are segmented by the license plate region and finally, characters are recognized by OCR technique. The LPR system is tested by different license plate images in different lighting conditions. The experimental research shows the better performance of the LPR system.