Artificial Neural Network Based Automatic Number Plate Recognition System

Abstract

This paper deals with Automatic Number Plate Recognition (ANPR) using Artificial Neural Network. The ANPR system includes steps like pre-processing, localization, character segmentation and character recognition. The developed system first detects the vehicle and then captures the vehicle image. The captured image is pre-processed in order to enhance it for further processing. In localization the license plate region is located and cropped from the complete image. In character segmentation images of individual alpha-numeric characters are extracted from the localized plate. In this paper, we proposed Neural Network based character recognition. Scaled Conjugate Gradient Backpropogation algorithm is used for training the neural network. This system is implemented in MATLAB R2014b.