Abstract

The decipherment and transcription of ancient manuscripts present substantial challenges due to their frequently fragmented, deteriorated, and stylistically heterogeneous nature. This paper introduces an innovative methodology for ancient text recognition through the application of Convolutional Neural Networks (CNNs) implemented on a Raspberry Pi 3B+ for enhanced sensor integration. The approach leverages advanced deep learning paradigms to preprocess and analyze historical documents, thereby enhancing the precision of text extraction and interpretation. The CNN architecture proposed is trained on a heterogeneous corpus encompassing diverse scripts and languages, thus improving its capacity to generalize across disparate ancient writing systems. Techniques such as data augmentation, transfer learning, and customized network architectures are employed to address challenges such as reduced resolution and significant variability in textual appearance. Empirical results demonstrate that CNN-based methodologies significantly surpass conventional optical character recognition (OCR) approaches, achieving notable advancements in both accuracy and efficiency. This research provides a robust framework for the automated decipherment of ancient texts using the Raspberry Pi 3B+, potentially advancing further historical and philological scholarship.

References

• Anusha Leela Somashekharaiah and Abhay Deshpande. 2024. Preprocessing techniques for recognition of ancient Kannada epigraphs. International Journal of Electrical and Computer Engineering (IJECE), Vol 14, No 1, February 2024.
• Aİ.Topçuand and UğurTöreyin, Neural Machine Translation Approaches for Post OCR Text Processing, 30th Signal Processing and Communications Applications Conference (SIU),Safranbolu, Turkey, 2022.
• A.İ.TopçuandU and UğurTöreyin, Neural Machine Translation Approaches for Post OCR Text Processing, 30th Signal Processing and Communications Applications Conference (SIU), Safranbolu, Turkey, 2022
• .A. Kashyap and A. Kumara B, OCR of Kannada Characters using Deep Learning. Trends in Electrical ,Electronics, Computer Engineering Conference (TEECCON), Bengaluru, India, 2022, pp. 35-38,doi:10.1109/TEECCON54414.2022.9854842.
• .A. Spruck, M. Hawesch, A. Maier, C. Riess, J. Seiler and A, 3D Rendering Frame work for Data Augmentation in Optical Character Recognition, 2021 International Symposiumon Signals, Circuits and Systems (ISSCS), Iasi, Romania, 2021, pp. 1-4,doi:10.1109/ISSCS52333.2021.9497438.
• . A. Spruck, M. Hawesch, A. Maier, C. Riess, J. Seiler and A. Kau, 3D Rendering Frame work for Data Augmentation in Optical Character Recognition, 2021 International Symposiumon Signals, Circuits and Systems (ISSCS), Iasi, Romania, 2021, pp. 1-4,doi:10.1109/ISSCS52333.2021.9497438.
• .Chandrakala,H and Thippeswamy,G, Deep Convolutional Neural Networks for Recognition of Historical Handwritten Kannada Characters. 10.1007/978-981-13-9920-6_7, 2020.
• .Bannigidad, Parashuram, Gudada and Chandrashekar, Historical Kannada Handwritten Character Recognition using K-Nearest Neighbour Technique, 2019.
• Panchal, Amoli, Chintan, Shah and Bhargav, Image Binarization Techniques for Degraded Document Images: A Review Binarization techniques, 2017.
• . W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, and C. Fu, et al. 2017. SSD: Single Shot Multi Box Detector. Proceedings of Computer Vision and Pattern Recognition.
• A. Soumya and G. H. Kumar, Recognition of Ancient Kannada Epigraphs using Fuzzy-based Approach. International Conference on Contemporary Computing and Informatics. Mysore, India, 2014, pp. 657-662, doi: 10.1109/IC3I.2014.7019645.
• K. R. Shah and D. Dattatray Badgujar, Devnagari Handwritten Character Recognition (DHCR) for Ancient Documents. IEEE Conference on Information & Communication Technologies. Thuckalay, India, 2013, pp. 656-660, doi: 10.1109/CICT.2013.6558176.
• C. Naveena and V. N. Manjunath Aradhya, Handwritten Character Segmentation for Kannada Scripts. World Congression Information and Communication Technologies. Trivandrum, India, 2012, pp.144149, doi:10.1109/WICT.2012.6409065.
• T. Sari, A. Kefali and H. Bahi, An MLP for Binarizing Images of Old Manuscripts. International Conference on Frontiers in Handwriting Recognition. Bari, Italy, 2012, pp. 247-251, doi: 10.1109/ICFHR.2012.176.
• Indira, K. & Selvi, S. Sethu, Kannada Character Recognition System,2010 - A Review.
• M. I. Shah and C. Y. Suen, Word Spotting in Gray Scale Handwritten Pashto Documents. 12th International Conference on Frontiersin Handwriting Recognition. Kolkata, India,2010, pp.136-141, doi:10.1109/ICFHR.2010.28.
• Fischer A, Wthrich M Liwicki M, Frinken V, Bunke H, Viehhauser G and Stolz M, Automatic Transcription of Handwritten Medieval Documents. Proceedings of Virtual Systems and Multimedia, 137{142 (2009).
• Sauvola J. J. and Pietikainen M. 2000. Adaptive Document Image Binarization. Pattern Recognition 33(2),225{236 (2000).
• T. Wakahara and Y. Kimura, Affine-Invariant Gray-Scale Character Recognition using GAT Correlation. Proceedings 15th International Conference on Pattern Recognition. ICPR-2000, Barcelona, Spain, pp. 417-421 vol.4, doi: 10.1109/ICPR.2000.902947.