An Ehanced Digical Character Recogniton using Art Network Classifier

Abstract

Image Classification is one of the major applications of Image processing to place all the related images in a separate group. In some Recognition and Identification applications, Image classification is used to reduce the size of processing dataset. In this present work, the image classification is used to perform the digital character recognition. In this work, A Soft computing technique called Art Network is been implemented through the classification process. The work is divided into two main stages named Training and Recognition process. During the Training Phase, the input imageset is processed and divided in N Classes based upon the tolerance ratio. The Eligibility criterion to specific class is decided by Feature analysis over the image. In second stage, the feature extraction over the input image is performed and based on featured value; the related class is been identified. Now one-to-one comparison is performed on that class to identify the class on which comparison is performed. The effectiveness to the work is estimated based on matching ratio.