Telugu Text Extraction and Recognition Using Convolutional and Recurrent Neural Networks
A. Ram Bharadwaj1, A. Venugopal2, Ch. Surya Kiran3, M. V. Nageswara Rao4
1A. Ram Bhardwaj, Department of ECE, GMRIT, Rajam (Andhra Pradesh), India.
2A. Venugopal, Department of ECE, GMRIT, Rajam (Andhra Pradesh), India.
3Ch. Surya Kiran, Department of ECE, GMRIT, Rajam (Andhra Pradesh), India.
4Dr. M. V. Nageswara Rao, Department of ECE, GMRIT, Rajam (Andhra Pradesh), India.
Manuscript received on 18 June 2019 | Revised Manuscript received on 25 June 2019 | Manuscript published on 30 June 2019 | PP: 1449-1451 | Volume-8 Issue-5, June 2019 | Retrieval Number: E7422068519/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Recognizing words form images is the most atomic aspect of an OCR system. Inspired by the recent success of deep-learning based techniques in computer vision & sequence prediction, an end-to-end trainable CNN-RNN based neural network model for recognizing Telugu words from images is presented. This model can also be extended to other languages.
Keywords: Recurrent Neural Networks -Rnn, Optical Character Recognition -Ocr, Convolutional Neural Networks -Cnn
Scope of the Article: Neural Information Processing