A Deep Learning Based Optical Character Recognition Model for Old Turkic

Seyed Hossein Taheri; Houman Kosarirad; Isabel Adrover Gallego; Nedasadat Taheri

doi:10.4108/airo.8460

Authors

Seyed Hossein Taheri Islamic Azad University, Tehran
Houman Kosarirad University of Nebraska–Lincoln
Isabel Adrover Gallego University of Nebraska–Lincoln
Nedasadat Taheri University of Nebraska–Lincoln https://orcid.org/0009-0009-8946-9115

DOI:

https://doi.org/10.4108/airo.8460

Keywords:

Deep learning, optical character recognition, convolutional neural network

Abstract

This study presents the development and evaluation of a deep learning-based optical character recognition (OCR) model specifically designed for recognizing Old Turkic script. Utilizing a convolutional neural network (CNN), the project aimed to achieve high classification accuracy across a dataset comprising 38 distinct Old Turkic characters. To enhance the model’s robustness and generalization capabilities, sophisticated data augmentation techniques were employed, generating 760 augmented images from the original 38 characters. The model was rigorously trained and validated, achieving an overall ac- curacy of 96.34%. Evaluation metrics such as precision, recall, and F1-scores were systematically analyzed, showing superior performance in most classes while identifying areas for further optimization. The results underscore the effectiveness of CNN architectures in specialized OCR tasks, demonstrating their potential in preserving and digitizing historical scripts. This study not only advances the field of document analysis and OCR but also contributes to the digital preservation and accessibility of ancient scripts.

Downloads

References

[1] M. V. Vavulin, “Documentation of old turkic runic inscriptions of the altai mountains using photogrammetric technology,” The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 42, pp. 257–261, 2017.

[2] A. Bakırcı, “A deep learning based translation system from ottoman turkish to modern turkish,” Unpublished master thesis. I˙stanbul: Gebze Technical University, 2019.

[3] S. Kirmizialtin and D. Wrisley, “Automated transcription of non-latin script periodicals: a case study in the ottoman turkish print archive,” arXiv preprint arXiv:2011.01139, 2020.

[4] J. Premi et al., “Cnn based digital alphanumeric archaeolinguistics apprehension for ancient script detection,” Turkish Journal of Computer and Mathematics Education (TURCOMAT), vol. 12, no. 6, pp. 5320– 5326, 2021.

[5] S. R. Narang, M. Kumar, and M. K. Jindal, “Deepnetdevanagari: a deep learning model for devanagari ancient character recognition,” Multimedia Tools and Applications, vol. 80, pp. 20 671–20 686, 2021.

[6] Patel, Chirag, Atul Patel, and Dharmendra Patel. "Optical character recognition by open source OCR tool tesseract: A case study." International journal of computer applications 55.10 (2012).

[7] Li, Minghao, et al. "Trocr: Transformer-based optical character recognition with pre-trained models." Proceedings of the AAAI conference on artificial intelligence. Vol. 37. No. 11. 2023.

[8] Rezanezhad, Vahid, Konstantin Baierer, and Clemens Neudecker. "A hybrid CNN-transformer model for historical document image binarization." Proceedings of the 7th International Workshop on Historical Document Imaging and Processing. 2023.

[9] Li, Minghao, et al. "Trocr: Transformer-based optical character recognition with pre-trained models." Proceedings of the AAAI conference on artificial intelligence. Vol. 37. No. 11. 2023.

[10] Dosovitskiy, Alexey, et al. "An image is worth 16x16 words: Transformers for image recognition at scale." arXiv preprint arXiv:2010.11929 (2020).

[11] Egan, Gabriel. "Introduction to a special issue on computational methods for literary-historical textual scholarship." (2019).

[12] Clanuwat, Tarin, Alex Lamb, and Asanobu Kitamoto. "Kuronet: Pre-modern Japanese kuzushiji character recognition with deep learning." 2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019.

[13] “Old turkic script,” 2024, accessed: 2024-05-05. [Online]. Available: https://en.wikipedia.org/wiki/OldT urkicscript.

[14] Hedjam, R., & Cheriet, M. (2014). Historical document image restoration using multispectral imaging system. Pattern Recognition, 47(6), 2022-2030.

[15] Hedjam, R., Nafchi, H. Z., Kalacska, M., & Cheriet, M. (2015). An investigation on multispectral imaging for historical document image binarization. IEEE Transactions on Image Processing, 24(9), 3113-2025.

[16] Yang, Y., Sun, S., Li, W., & Wang, J. (2019). Deep Learning for Document Image Enhancement. IEEE Transactions on Image Processing, 28(5), 2420-2435.

[17] Jindal, A., & Arora, C. (2021). "Recognition of historical scripts using shallow convolutional neural networks." International Journal on Document Analysis and Recognition (IJDAR), 24(2), 89-98.

[18] Patel, Chirag, Atul Patel, and Dharmendra Patel. "Optical character recognition by open source OCR tool tesseract: A case study." International journal of computer applications 55.10 (2012).

[19] Holley, Rose. "How good can it get? Analysing and improving OCR accuracy in large scale historic newspaper digitisation programs." D-Lib Magazine 15.3/4 (2009).

[20] Dosovitskiy, A., et al. (2021). "An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale." International Conference on Learning Representations (ICLR).

[21] He, K., Zhang, X., Ren, S., & Sun, J. (2016). "Deep Residual Learning for Image Recognition." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770-778.

[22] Li, M., Lv, T., Cui, L., Lu, S., & Tang, C. (2022). "TrOCR: Transformer-based Optical Character Recognition." IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(11), 8810-8823.

[23] Clanuwat, T., Kitamoto, A., Lamb, A., Yamamoto, K., & Ha, D. (2020). "KuroNet: Pre-modern Japanese Kuzushiji character recognition with deep learning." NeurIPS 2020

[24] Kiessling, B., & Malyshev, A. (2019). "Cuneiform script recognition using convolutional neural networks." Proceedings of the 3rd Workshop on Ancient Document Processing, ICDAR 2019.

[25] Shorten, C., & Khoshgoftaar, T. M. (2019). "A survey on Image Data Augmentation for Deep Learning." Journal of Big Data, 6(1), 1-48.