Open AccessHandwriting Text Recognition using Neural Networks
Author(s) -
H Parikshith,
Naga Rajath S M,
Sudhish Reddy D,
C M Sindhu,
P Ravi
Publication year - 2019
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.b7705.129219
Subject(s) - computer science , handwriting , convolutional neural network , artificial intelligence , classifier (uml) , artificial neural network , intelligent character recognition , recurrent neural network , deep learning , handwriting recognition , neocognitron , natural language processing , task (project management) , speech recognition , time delay neural network , pattern recognition (psychology) , feature extraction , character recognition , management , economics , image (mathematics)
Handwritten text recognition is a laborious task because humans can write a similar message in numerous ways or due to huge diversity in individual’s style of writing. The performance of text recognition systems implemented as neural networks has better results and accuracy than normal traditional classifiers. In this paper we explore the methods used to recognize and detect handwritten text or words in different languages. The major method used to recognize text is the Convolutional neural network (CNN) as a deep learning classifier. The other techniques used are Recurrent Neural Network (RNN) and a custom developed model called deep-writer, which is a variant of CNN architecture.