Research and evaluation of the efficiency of handwritten character recognition methods using convulsional neural networks
DOI:
https://doi.org/10.31498/2225-6733.47.2023.299989Keywords:
recognition of handwritten text, recognition of charactersAbstract
The paper analyzes the possibilities of using deep learning neural networks for the implementation of text processing information systems, substantiates the prospects of this technology and the possibilities of its improvement. The possibility of using the MNIST database of handwritten digits on, as well as the EMNIST database of handwritten letters and numbers, namely the EMNIST Letters set, was considered. The research object was modelled in IDEF0 and IDEF3 notations for the «AS-IS» option. It was found that «Building a dataset for model training» and «Choosing a convolutional neural network architecture» are the most frequently asked questions. Reengineering measures are proposed, namely, the expediency of using, in addition to the well-known EMNIST and MNIST sets, a set of handwritten and italic fonts with Ukrainian glyphs, additionally created as part of the work, is shown. For this purpose, modern IT tools such as the Pillow library, Image Data Generator and the Scikit-Learn package were used to select training and test samples. Also, in addition to the initially proposed simplest CNN architectures of the Lenet type, the use of more complex architectures of the AlexNet and VGG-16 types is proposed. Taking into account the proposed reengineering measures, IDEF0 and IDEF3 diagrams were constructed for the «TO-BE» option. The paper analyzes in detail the results of recognizing handwritten Ukrainian letters and Arabic numerals using 6 different CNN architectures using a synthetic data set for training. The research presented in the work was carried out using a software application developed in the Python programming language using the Scikit Learn package, which provides the user with the ability to recognize handwritten text using a multilayer perceptron. carried out precisely using the developed program. It is justified that, in contrast to simple Lenet-type architectures, it is more appropriate to use more complex options, namely the VGG-16 type architecture. Experimental studies of the influence of the number of CNN parameters of different architectures on the recognition accuracy and training time of the neural network have been carried out. Also analyzed are the results of character recognition when recognizing images that do not belong to the training or test sample
References
Studies in fuzziness and soft computing. Springer Cham, 2017. Vol. 352: Optical character recognition systems for different languages with soft computing / Chaudhuri A., Mandaviya K., Badelia, P., Ghosh S. 248 p. DOI: https://doi.org/10.1007/978-3-319-50252-6.
Li H., Wang P., Shen C. Toward end-to-end car license plate detection and recognition with deep neural networks. IEEE Transactions on Intelligent Transportation Systems. 2018. Vol. 20(3). Pp. 1126-1136. DOI: https://doi.org/10.1109/TITS.2018.2847291.
Rajavelu A., Musavi M.T., Shirvaikar M.V. A neural network approach to character recognition. Neural Networks. 1989. Vol. 2(5). Pp. 387-393. DOI: https://doi.org/10.1016/0893-6080(89)90023-3.
Image character recognition using deep convolutional neural network learned from different lan-guages / Bai J., Chen Z., Feng B., Xu B. 2014 IEEE International Conference on Image Processing ICIP 2014, Paris, France, 27-30 October 2014. Pp. 2560-2564. DOI: https://doi.org/10.1109/ICIP.2014.7025518.
Maitra D.S., Bhattacharya U., Parui S.K. CNN based common approach to handwritten character recognition of multiple scripts. 3th International Conference on Document Analysis and Recognition (ICDAR), Tunis, Tunisia, 23-26 August 2015. Pp. 1021-1025. DOI: https://doi.org/10.1109/ICDAR.2015.7333916.
Bilgin Taşdemir E.F. Online Turkish Handwriting Recognition Using Synthetic Data. Avrupa Bilim Ve Teknoloji Dergisi. 2021. Vol. 32. Pp. 649-656. DOI: https://doi.org/10.31590/ejosat.1039846.
Handwritten Kazakh and Russian (HKR) database for text recognition / D. Nurseitov, K. Bostanbekov, D. Kurmankhojayev, A. Alimova, A. Abdallah, R. Tolegenov. Multimedia Tools and Applications. 2021. Vol. 80. Pp. 33075-33097. DOI: https://doi.org/10.1007/s11042-021-11399-6.
Abdallah A., Hamada M., Nurseitov D. Attention-based fully gated CNN-BGRU for Russian handwritten text. Journal of Imaging. 2020. Vol. 6(12). Pp. 1-23. DOI: https://doi.org/10.3390/jimaging6120141.
Ullah Z., Jamjoom M. An intelligent approach for Arabic handwritten letter recognition using convolutional neural network. PeerJ Computer Science. 2022. Vol. 8. Pp. 1-22. DOI: https://doi.org/10.7717/peerj-cs.995.
Handwritten letter recognition using artificial intelligence / Jeevitha D., Muthu S., Nila I., Santhoshi V. International journal for research in applied science and engineering technology. 2022. Vol. 10. Pp. 2752-2758. DOI: https://doi.org/10.22214/ijraset.2022.42949.
An exploratory study on the handwritten allographic features of multiethnic population with different educational backgrounds / L. Gannetion, K.Y. Wong, P.Y. Lim, K.H. Chang, A.F.L. Abdullah. PloS one. 2022. Vol. 17(10). Pp. 1-16. DOI: https://doi.org/10.1371/journal.pone.0268756.
EMNIST: Extending MNIST to handwritten letters / Cohen G., Afshar S., Tapson J., Van Schaik A. 2017 International Joint Conference on Neural Networks (IJCNN), Anchorage, AK, USA, 14-19 May 2017. Pp. 2921-2926. DOI: https://doi.org/10.48550/arxiv.1702.05373.
Backpropagation applied to handwritten zip code recognition / Y. LeCun et al. Neural Computation. 1989. Vol. 1. Iss. 4. Pp. 541-551. DOI: https://doi.org/10.1162/neco.1989.1.4.541.
Núñez Fernández D., Hosseini S. Real-time handwritten letters recognition on an embedded computer using ConvNets. IEEE Sciences and Humanities International Research Conference (SHIRCON), Lima, Peru, 20-22 November 2018. Pp. 1-4. DOI: https://doi.org/10.1109/SHIRCON.2018.8592981.
Krizhevsky A., Sutskever I., Hinton G. ImageNet classification with deep convolutional neural networks. Communications of the ACM. 2012. Vol. 60. Iss. 6. Pp. 84-90. DOI: https://doi.org/10.1145/3065386.
Going deeper with convolutions / C. Szegedy et al. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, USA, 7-12 June 2015. Pp. 1-9. DOI: https://doi.org/10.1109/CVPR.2015.7298594.
Baldominos A., Sáez Y., Isasi P. A survey of handwritten character recognition with MNIST and EMNIST. Applied Sciences. 2019. Vol. 9(15). Pp. 1-16. DOI: https://doi.org/10.3390/app9153169.
Handwritten indic character recognition using capsule networks / B. Mandal, S. Dubey, S. Ghosh, R. Sarkhel, N. Das. 2018 IEEE Applied Signal Processing Conference (ASPCON), Kolkata, India, 07-09 December 2018. Pp. 304-308. DOI: https://doi.org/10.1109/ASPCON.2018.8748550.
Recognition of isolated characters across different input interfaces using 2D DCNN / K. Yadav, A. Kirupakaran, S. Barlaskar, N. Ahmad, R. Laskar, M. Bhuyan. 2021 IEEE Region 10 Conference (TENCON), Auckland, New Zealand, 07-10 December 2021. Pp. 504-509. DOI: https://doi.org/10.1109/TENCON54134.2021.9707451.
He K., Girshick R., Dollár P. Rethinking imagenet pre-training. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea (South), 27 October - 02 November 2019. Pp. 4918-4927. DOI: https://doi.org/10.1109/ICCV.2019.00502.
Albattah W., Albahli S. Intelligent Arabic handwriting recognition using different standalone and hybrid CNN architectures. Applied Sciences. 2022. Vol. 12. Pp. 1-23. DOI: https://doi.org/10.3390/app121910155.
Performance analysis of state-of-the-art convolutional neural network architectures in bangla handwritten character recognition / T. Ghosh, Min-Ha-Zul Abedin, H. Al Banna, N. Mumenin, M. Abu Yousuf. Pattern recognition and image analysis. 2021. Vol. 31. Pp. 60-71. DOI: https://doi.org/10.1134/S1054661821010089.
Bhardwaj A., Ravendra Singh R. Handwritten devanagari character recognition using deep learning – convolutional neural network (CNN) model. PalArch’s Journal of Archaeology of Egypt/Egyptology. 2020. Vol. 17(6). Pp. 7965-7984.
Handwritten devanagari character recognition using modified Lenet and Alexnet convolution neural networks / Duddela P., Vasanth Kumar M.R., Ramana K., Vidhyacharan B. Wireless Personal Communications. 2022. Vol. 122. Pp. 349-378. DOI: https://doi.org/10.1007/s11277-021-08903-4.
Recognizing arabic handwritten literal amount using convolutional neural networks / A. Korichi, S. Sihem, T. Najiba, Z. Ramzi, A. Oussama. Artificial Intelligence and Its Applications : Proceeding of the 2nd International Conference, EL-Oued, Algeria, 28-30 September 2021. Pp. 153-165. DOI: https://doi.org/10.1007/978-3-030-96311-8_15.
A new Arabic handwritten character recognition deep learning system (AHCR-DLS) / Balaha H., Sabry M., Ali H., Badawy M. Neural Computing and Applications. 2020. Vol. 33. Pp. 6325-6367. DOI: https://doi.org/10.1007/s00521-020-05397-2.
Samra A., Al Amin G., Hadi O. An optimized deep residual network with a depth concatenated block for handwritten characters classification. Computers, Materials & Continua. 2021. Vol. 680. Pp. 1-28. DOI: https://doi.org/10.32604/cmc.2021.015318.
Classification of handwritten names of cities and handwritten text recognition using various deep learning models / D. Nurseitov, K. Bostanbekov, M. Kanatov, A. Alimova, A. Abdallah, G. Abdimanap. Advances in Science, Technology and Engineering Systems Journal (ASTESJ). 2021. Vol. 5. Iss. 5. Pp. 934-943. DOI: https://doi.org/10.48550/arXiv.2102.04816.
Vovchuk O., Kyrychenko M. Recognition of handwritten cyrillic letters using PCA. 2019. 6 p.
Economic efficiency of innovative projects of CNN modified architecture application / V. Khavalko, V. Mykhailyshyn, R. Zhelizniak, I. Kovtyk, A. Mazur. CEUR Workshop Proceedings. Vol. 2654: Proceedings of the International workshop on cyber hygiene (CybHyg-2019) co-located with 1st International conference on cyber hygiene and conflict management in global information networks (CyberConf 2019), Kyiv, Ukraine, November 30, 2019. Pp. 182-193.
Розпізнавання рукописних українських літер та цифр з використанням синтетичного набору даних та згорткових нейронних мереж / Є. Чичкарьов, О. Зінченко, О. Балалаєва, А. Сергієнко, О. Ковальов. Grail of Science. 2023. Vol. 23. Pp. 241-253. DOI: https://doi.org/10.36074/grail-of-science.23.12.2022.36.
Study and observation of the variations of accuracies for handwritten digits recognition with various hidden layers and epochs using convolutional neural network / Arif R.B., Siddique M.A.B., Khan M.M.R., Oishe M.R. 4th International Conference on Electrical Engineering and Information & Communication Technology (iCEEiCT), Dhaka, Bangladesh, 13-15 September 2018. Pp. 112-117. DOI: https://doi.org/10.1109/CEEICT.2018.8628078.
Siddique F., Sakib S., Siddique M.A.B. Handwritten digit recognition using convolutional neural network in python with tensorflow and observe the variation of accuracies for various hidden layers. Preprints. 2019. Pp. 1-6. DOI: https://doi.org/10.20944/preprints201903.0039.v1.
LeCun Y., Cortes C., Burges C.J.C. The MNIST database of handwritten digits. URL: http://yann.lecun.com/exdb/mnist (дата звернення: 30.05.2023).
Grother P.J. NIST special database 19 – handprinted forms and characters database. National Institute of Standards and Technology (NIST), Tech. Rep., 1995. DOI: http://doi.org/10.18434/T4H01C.
Downloads
Published
How to Cite
Issue
Section
License
The journal «Reporter of the Priazovskyi State Technical University. Section: Technical sciences» is published under the CC BY license (Attribution License).
This license allows for the distribution, editing, modification, and use of the work as a basis for derivative works, even for commercial purposes, provided that proper attribution is given. It is the most flexible of all available licenses and is recommended for maximum dissemination and use of non-restricted materials.
Authors who publish in this journal agree to the following terms:
1. Authors retain the copyright of their work and grant the journal the right of first publication under the terms of the Creative Commons Attribution License (CC BY). This license allows others to freely distribute the published work, provided that proper attribution is given to the original authors and the first publication of the work in this journal is acknowledged.
2. Authors are allowed to enter into separate, additional agreements for non-exclusive distribution of the work in the same form as published in this journal (e.g., depositing it in an institutional repository or including it in a monograph), provided that a reference to the first publication in this journal is maintained.
