A Novel Approach to an Automatic License Plate Recognition (ALPR) System

Abstract

This paper presents an algorithm for the automatic license plate recognition system using various approaches. Automatic licence plate recognition (ALPR) technique consists of two modules, a license plate locating module and a license number identification module. In this paper, an experimental composition is performed with PCA (Principle Component Analysis), OCR (Optical Character Recognition) and SVM (Support Vector Machine) which are based on original binary images of the characters. The training via support vector machine has yielded the accurate outputs for the recognition of various input images taken in the algorithm. The proposed algorithm is efficient enough over a calculated dataset of licensed number plates.