Devnagari Character Recognition using Optical Character Recognition (OCR)

Abstract

In the age of digital applications, optical character recognition (OCR) is a useful tool. There are many different languages and scripts used around the world, with more than 300 million individuals in India speak Hindi as their first language. OCR previously created for the Devnagari script has a very low recognition rate. In this paper, we propose an OCR system based on a Convolutional Neural Network (CNN) model. It increases the handwritten data's efficiency or accuracy by 98.32%. We do several steps for getting this efficiency like preprocessing of data, grayscale images, binarization, reshaping, normalization, modelling, and Evaluation of the data with the help of plotting.