Recognition of handwritten Latin characters with diacritics using CNN

• Autorzy: Lukasik Edyta , Charytanowicz Małgorzata , Milosz Marek , Tokovarov Michail , Kaczorowska Monika , Czerwinski Dariusz , Zientarski Tomasz

Identyfikatory

DOI

10.24425/bpasts.2020.136210

• Języki publikacji: EN

Abstrakty

EN

Convolutional Neural Networks (CNN) have achieved huge popularity in solving problems in image analysis and in text recognition. In this work, we assess the effectiveness of CNN-based architectures where a network is trained in recognizing handwritten characters based on Latin script. European languages such as Dutch, French, German, etc., use different variants of the Latin script, so in the conducted research, the Latin alphabet was extended by certain characters with diacritics used in Polish language. To evaluate the recognition results under the same conditions, a handwritten Latin dataset was also developed. The proposed CNN architecture produced an accuracy of 96% for the extended character set. This is comparable to state-of-the-art results found in the domain of identifying handwritten characters. The presented approach extends the usage of CNN-based recognition to different variants of the Latin characters and shows it can be successfully used for a set of languages based on that script. It seems to be an effective technique for a set of languages written using the Latin script.

Słowa kluczowe

EN

handwritten documents; diacritic; neural networks; character recognition; deep learning

PL

dokument odręczny; znaki diakrytyczne; sieci neuronowe; rozpoznawanie znaków; głęboka nauka

• Wydawca: Polska Akademia Nauk, Wydział IV Nauk Technicznych
• Czasopismo: Bulletin of the Polish Academy of Sciences. Technical Sciences
• Rocznik: 2021
• Tom: Vol. 69, nr 1
• Strony: art. no. e136210
• Opis fizyczny: Bibliogr. 25 poz., rys., tab.

Twórcy

autor

Lukasik Edyta

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

• https://orcid.org/0000-0003-3644-9769

autor

Charytanowicz Małgorzata

• Systems Research Institute Polish Academy of Sciences, ul. Newelska 6, 01-447 Warsaw, Poland
• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

• https://orcid.org/0000-0002-1956-3941

autor

Milosz Marek

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

• https://orcid.org/0000-0002-5898-815X

autor

Tokovarov Michail

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

autor

Kaczorowska Monika

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

autor

Czerwinski Dariusz

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

autor

Zientarski Tomasz

• Lublin University of Technology, ul. Nadbystrzycka 38D, 20-618 Lublin, Poland

• https://orcid.org/0000-0002-1693-5316

Bibliografia

• [1] E. Lukasik and T. Zientarski, “Comparative analysis of selected programs for optical text recognition”, J. Comput. Sci. Inst. 7, 191‒194 (2018).
• [2] P. Kusaj, M. Kosyra, and M. Charytanowicz, “Web-Page Classification Based on Wikipedia Structure. Recent Developments” in Mathematics and Informatics, Contemporary Mathematics and Computer Science 2, Part II, A. Zapała (red.), pp. 89‒102, Wydawnictwo KUL, 2016.
• [3] D. Połap and M. Woźniak, “Flexible neural network architecture for handwritten signatures recognition”, Int. J. Electron. Telecommun. 62, 197–202 (2016).
• [4] M. Milosz and J. Gazda, “Effectiveness of artificial neural networks in recognising handwriting characters”, J. Comput. Sci. Inst. 7, 210‒214 (2018).
• [5] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition”. Proc. IEEE 86(11), 2278‒2324 (1998).
• [6] A. Pal and D. Singh, “Handwritten English character recognition using neural network”, Int. J. Comput. Sci. Commun. 1(2), 141‒144 (2010).
• [7] B.K. Verma, “Handwritten Hindi character recognition using multilayer perceptron and radial basis function neural network”, IEEE International Conference on Neural Network 4, 2111‒2115 (1995).
• [8] D. Singh, S.K. Singh, and M. Dutta, “Hand written character recognition using twelve directional feature input and neural network”, Int. J. Comput. Appl. 1(3), 94‒98 (2010).
• [9] Y. Perwej and A. Chatirvedi, “Neural networks for handwritten English alphabet recognition”, Int. J. Comput. Appl. 20(7), 1–5 (2011).
• [10] J. Pradeep, E. Srinivasan, and S. Himavathi, “Neural network based handwritten character recognition system without feature extraction”, 2011 International Conference on Computer, Communication and Electrical Technology (ICCCET), Tamilnadu, 2011, pp. 40‒44.
• [11] A.M. Obaid, H.M. El Bakry, M.A. Elodusuky, and A.I. Shehab, “Handwritten text recognition system based on neural network”, Int. J. Adv. Res. Comput. Sci. Technol. 4(1), 72‒77 (2016).
• [12] V. Lebedev and V. Lempitsky. “Speeding-up convolutional neural networks: A survey”, Bull. Pol. Ac.: Tech. 66(6), 799‒810 (2018).
• [13] D. Firmani, P. Merialdo, E. Nieddu, and S. Scardapane, “In codice ratio: OCR of handwritten Latin documents using deep convolutional networks”, in AI* CH@ AI* IA, 2017, pp. 9‒16.
• [14] F.P. Such, D. Peri, F. Brockler, P. Hutkowski, and R. Ptucha. “Fully convolutional networks for handwriting recognition”. In: 2018 16th International Conference on Frontiers in Handwriting Recognition (ICFHR), IEEE, 2018, pp. 86‒91.
• [15] P. Grother, “NIST special database 19 handprinted forms and characters database”, National Institute of Standards and Technology, Tech. Rep., 1995.
• [16] M. Lutf, X. You, Y. Cheung, and C.L.P. Chen, “Arabic font recognition based on diacritics features”, Pattern Recognit. 47, 672–684 (2014).
• [17] K.E. Gajoui, F.A. Allah, and M. Oumsis, “Diacritical Language OCR based on neural network: Case of Amazigh language”. Procedia Comput. Sci. 73, 298‒305 (2015).
• [18] J. Náplava, M. Straka, P. Straňák, and J. Hajič, “Diacritics Restoration Using Neural Networks”, Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC), 2018.
• [19] D. Grzelak, K. Podlaski, and G. Wiatrowski, “Analyze the effectiveness of an algorithm for identifying Polish characters in handwriting based on neural machine learning technologies”, Journal of King Saud University – Computer and Information Sciences, 2019, doi: 10.1016/j.jksuci.2019.08.001.
• [20] G. Cohen, S. Afshar, J. Tapson, and A. van Schaik, ”EMNIST: an extension of MNIST to handwritten letters”. Retrieved from: http://arxiv.org/abs/1702.05373, 2017.
• [21] M. Tokovarov, M. Kaczorowska, and M. Milosz, “Development of Extensive Polish Handwritten Characters Database for Text Recognition Research”, Adv. Sci. Technol. Res. J. 14(3), 30–38 (2020), doi:10.12913/22998624/122567.
• [22] M. Charytanowicz and P. Kulczycki, “An Image Analysis Algorithm for Soil Structure Identification“; in: Intelligent Systems’2014, pp. 681‒692, D. Filev, J. Jablkowski, J. Kacprzyk, I. Popchev, L. Rutkowski, V. Sgurev, E. Sotirova, P. Szynkarczyk, S. Zadrozny (eds.), Springer, Berlin, 2014.
• [23] The Polish Handwritten Characters Database, [Online]. https:// cs.pollub.pl/phcd/?lang=en.
• [24] D.P. Kingma and J.L. Ba, “Adam: A method for stochastic optimization”. arXiv:1412.6980v9, 2014.
• [25] M. Abadi et al., “Tensorflow: A system for large-scale machine learning,” in 12th Symposium on Operating Systems Design and Implementation, 2016, pp. 265‒283.

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa Nr 461252 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2021).