A Neural Approach to Cursive Handwritten Character Recognition Using Features Extracted from Binarization Technique

Abstract

The feature extraction is one of the most crucial steps for an Optical Character Recognition (OCR) System. The efficiency and accuracy of the OCR System, in recognizing the off-line printed characters, mainly depends on the selection of feature extraction technique and the classification algorithm employed. This chapter focuses on the recognition of handwritten characters of Roman Script by using features which are obtained by using binarization technique. The goal of binarization is to minimize the unwanted information present in the image while protecting the useful information. Various preprocessing techniques such as thinning, foreground and background noise removal, cropping and size normalization etc. are also employed to preprocess the character images before their classification. A multi-layered feed forward neural network is proposed for classification of handwritten character images. The difference between the desired and actual output is calculated for each cycle and the weights are adjusted during error back-propagation. This process continues till the network converges to the allowable or acceptable error. This method involves the back propagation-learning rule based on the principle of gradient descent along the error surface in the negative direction. Very promising results are achieved when binarization features and the multilayer feed forward neural network classifier is used to recognize the off-line cursive handwritten characters.