Design and Implementation of a Handwriting Text Recognition System

Abstract

Handwritten characters are difficult to recognize due to diversity of human handwriting style, variation in angle, size and shape of letters. In this work, an extensively employed method of handwritten text recognition is used to transform handwritten data into electronic format or digital form, preprocess, extract distinctive features, and classify written characters. The proposed approach is to design a Graphical User Interface (GUI) that corresponds to the ability of human beings to identify and verify handwritten characters by training datasets of 80 labeled digit images with Machine Learning, Deep Learning and Convolution Neural Network (CNN) for character recognition. The proposed recognition system performs excellently for the cursive handwriting with 80-90% accuracy.