Optical Character Recognition for English Handwritten Text Using Recurrent Neural Network
R. Parthiban, R. Ezhilarasi, D. Saravanan
Abstract
Manually written Text Recognition is an innovation that is genuinely necessary right now of today.Before appropriate execution of this innovation we have depended on composing writings with own may leads to some mistakes. It is hard to maintain safely and gathering that information with effectiveness.Difficult work is required so as to keep up appropriate association of the data.Recurrent neural networkis utilized to discover the arrangement of character.Today we have OCRs effectively accessible for the English language.We can discover OCRs for formal texted English also yet OCRs for written by hand content are uncommon.Furthermore,those which are accessible don't have a goodaccuracy.We expect to make such an OCR which gives us an impressiverecognition exactness for manually written Text using recurrent neural network.The proposed model is implemented using Conda, used with TensorflowFramework. The purpose of Recurrent neural network is to improve accuracy.