Abstract
It is highly desirable to develop an offline handwritten character recognition system for the Hindi script with an acceptable accuracy level. This script has many characters, including numerals and different shape variations of these characters. Moreover, the Hindi script also has compound characters as well as characters with modifiers. It makes the task of character recognition more complex. Even the shape of characters written by an individual may have significant variation at different times. Hence, incorporating all these issues is very challenging. This paper proposes the augmented data set to overcome these problems up to an extent. The augmented data set introduces the hypothetical changes in the original data set to increase the variations in the data samples, which may increase the variety of samples. Thus, the accuracy of the recognition system improves. For the classification of characters, Convolutional Neural Networks (CNNs) are employed. Based on CNN, classification models are more suitable for recognition tasks where the images are taken as inputs. So, utilizing deep learning algorithms for recognition of Hindi script characters with the projected scheme will improve the accuracy of the system.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
A. Krizhevsky, I. Sutskever, G.E. Hinton, ImageNet classification with deep convolutional neural networks. Commun. ACM 60(6), 84–90 (2017)
K. Peymani, M. Soryani, From machine-generated to handwritten character recognition: a deep learning approach, in 3rd International Conference on Pattern Recognition and Image Analysis (IPRIA), Shahrekord, pp. 243–247. https://doi.org/10.1109/PRIA.2017.7983055 (2017)
S. Banerjee, B.R. Ghosh, A. Kundu, Handwritten character recognition from bank cheque. Int. J. Comput. Sci. Eng. EISSN: 2347–2693, 4(1), 099–104 (2016)
G. Katiyar, S. Mehfuz, A hybrid recognition system for off-line handwritten characters. Springer Plus 5(1), 1–18 (2016)
P.M. Kamble, R.S. Hegadi, Geometrical features extraction and KNN based classification of handwritten marathi characters, in World Congress on Computing and Communication Technologies (WCCCT), pp. 219–222 (2017)
A. Indian, K. Bhatia, Offline handwritten hindi ‘SWARs’ recognition using a novel wave based feature extraction method. Int. J. Comput. Sci. Issues 14(4), 8–14 (2017)
H.D. Anuvadiya, A.A. Abhangi, A research on improve handwritten character recognition by using convolutional neural network. Int. J. Adv. Eng. Res. Dev. 5(05), 20–25 (2018)
A. Indian, K. Bhatia, Off-line handwritten hindi consonants recognition system using zernike moments and genetic algorithm, in Proceedings of the IEEE Conference International Conference on System Modeling & Advancement in Research Trends, (SMART-2018), pp. 10–16 (2018)
Z. Alom, P. Sidike, M. Hasan, T.M. Taha, V.K. Asari1, Handwritten bangla character recognition using the state-of-the-art deep convolutional neural networks. Hindawi Computat. Intell. Neurosci. 2018, Article ID 6747098, pp 1–13. https://doi.org/10.1155/2018/6747098 (2018)
S. Puria, S.P. Singh, An efficient devnagari character classification in printed and handwritten documents using SVM, in International Conference on Pervasive Computing Advances and Applications—PerCAA 2019, Procedia Computer Science, vol. 152, pp. 111–121 (2019)
A. Indian, K. Bhatia, Offline handwritten hindi numerals recognition using zernike moments. Int. J. Tomograp. Simul. 32(2), 68–82 (2019)
J. Memon, M. Sami, R.A. Khan, Handwritten Optical Character Recognition (OCR): a comprehensive Systematic Literature Review (SLR), Cornell University (2020)
M. Jaderberg, K. Simonyan, A. Vedaldi, A. Zisserman, Synthetic data and artificial neural networks for natural scene text recognition (2014)
C.E. Nwankpa, W. Ijomah, A. Gachagan, S. Marshall, Activation functions: comparison of trends in practice and research for deep learning. arXiv:1811.03378v1 [cs.LG] 8.(2018)
H. Wu, X. Gu, Max-pooling dropout for regularization of convolutional neural networks. (2015)
N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, R. Salakhutdinov, Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15, 1929–1958 (2014)
I. Sergey, C. Szegedy, Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv:1502.03167 (2015)
J. Brownlee, A gentle introduction to pooling layers for convolutional neural networks, in A Web Article on Deep Learning for Computer Vision, April 22 (2019)
https://in.mathworks.com/help/deeplearning/ref/nnet.cnn.layer.batchnormalizationlayer.htm
B. Baranidharan, A. Kandpal, A. Chakravorty, Hindi handwritten character recognition using CNN. Int. J. Adv. Sci. Technol. 29(6), 58–66 (2020)
P. K. Sonawane, S. Shelke, Handwritten devanagari character classificationusing deep learning, in IEEE International Conference on Information, Communication, Engineering and Technology (ICICET), Zeal College of Engineering and Research, Narhe, Pune, India, pp. 1–4, Aug. 29–30 (2018)
N. Aneja, S. Aneja, Transfer learning using CNN for handwritten devanagari character recognition, in Accepted for publication in IEEE International Conference on Advances in Information Technology (ICAIT) (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Indian, A., Bhatia, K., Kumar, K. (2022). Offline Handwritten Hindi Character Recognition Using Deep Learning with Augmented Dataset. In: Agrawal, R., He, J., Shubhakar Pilli, E., Kumar, S. (eds) Cyber Security in Intelligent Computing and Communications. Studies in Computational Intelligence, vol 1007. Springer, Singapore. https://doi.org/10.1007/978-981-16-8012-0_11
Download citation
DOI: https://doi.org/10.1007/978-981-16-8012-0_11
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-8011-3
Online ISBN: 978-981-16-8012-0
eBook Packages: EngineeringEngineering (R0)