Advertisement
Published in:
2021 | OriginalPaper | Chapter
35. Deep Learning Approach for Prediction of Handwritten Telugu Vowels
Authors : Ch. Prathima, Naresh Babu Muppalaneni
Published in: Intelligent Manufacturing and Energy Sustainability
Publisher: Springer Singapore
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Abstract
Telugu is an ancient language in which many historical and valuable scripts are written. Understanding and digitalization of theses scripts is difficult task. Hence, here we are proposing a machine learning approach which recognizes the handwritten Telugu vowels (Achulu). We employed convolutional neural network (CNN) approach to find the features automatically than the handicraft features. Dataset has been built and deployed in IEEE dataport. The results are impressive and interesting. Handwritten character recognition could be a part of Optical Character Recognition System (OCRS). OCRS can be applied to each printed text and handwritten documents. This paper describes the handwritten Telugu vowels recognition by employing a CNN approach. The dataset is pre-processed and also extracted the features using deep neural network system for training. The model is validated on test dataset, and ~98% of training accuracy ~92% of test accuracy are observed.