Pendekatan Denoising Hibrid untuk Pengenalan Aksara Jawa: Upaya Integrasi Teknologi dan Pelestarian Budaya Digital

Aji Priyambodo; Prihati Prihati

doi:10.29408/edumatic.v9i3.32376

Authors

Aji Priyambodo Program Studi Sistem dan Teknologi Informasi, Institut Teknologi dan Bisnis Semarang https://orcid.org/0000-0001-9974-762X
Prihati Prihati Program Studi Sistem dan Teknologi Informasi, Institut Teknologi dan Bisnis Semarang https://orcid.org/0009-0001-5867-8047

DOI:

https://doi.org/10.29408/edumatic.v9i3.32376

Keywords:

digital cultural preservation, hybrid denoising, heritage computer vision, javanese script, structural-preserving denoising

Abstract

The digitization of Javanese manuscripts is hindered by image degradation such as bleed-through, stains, blur, and uneven illumination that obscures fine graphemic details and reduces the reliability of optical character recognition (OCR). This challenge is significant for computational heritage preservation, as degraded visual inputs threaten the accuracy and longevity of digital cultural archives. This study applies and evaluates a Structural-Preserving Denoising (SPD) approach designed to retain graphemic contours during noise reduction. Using a comparative-experimental design, six denoising methods (Wavelet, Bilateral, Non-Local Means/NLM, and three hybrids) were tested as independent variables, with PSNR, SSIM, and OCR accuracy as dependent variables. A dataset of 4,632 degraded Javanese character images (64×64 pixels) was processed using Histogram of Oriented Gradients and classified via a cosine-kernel Support Vector Machine. Analyses included comparative evaluation, error analysis, and Pearson correlation. Results show that Bilateral Filtering achieved the best balance of structural preservation and recognition performance (SSIM 0.9108; PSNR 29.85 dB, accuracy 96.97%). The strong correlation between visual quality and OCR accuracy indicates that maintaining graphemic structure is crucial for robust recognition. The study contributes a theoretical SPD framework and a practical preprocessing pipeline applicable to the digital preservation of traditional scripts.

References

Anggraeny, F. T., Via, Y. V., & Mumpuni, R. (2023). Image preprocessing analysis in handwritten Javanese character recognition. Bulletin of Electrical Engineering and Informatics, 12(2), 860–867. https://doi.org/10.11591/eei.v12i2.4172

Anuradha, I., Liyanage, C., & Weerasinghe, R. (2021). Estimating the Effects of Text Genre, Image Resolution and Algorithmic Complexity needed for Sinhala Optical Character Recognition. International Journal on Advances in ICT for Emerging Regions (ICTer), 14(3), 43–51. https://doi.org/10.4038/icter.v14i3.7231

Bataineh, B., Tounsi, M., Zamzami, N., Janbi, J., Abu-ain, W. A. K., AbuAin, T., & Elnazer, S. (2025). A Comprehensive Review on Document Image Binarization. Journal of Imaging, 11(5), 133. https://doi.org/10.3390/jimaging11050133

Boudraa, M., Bennour, A., Mekhaznia, T., Alqarafi, A., Marie, R. R., Al-Sarem, M., & Dogra, A. (2024). Revolutionizing Historical Manuscript Analysis: A Deep Learning Approach with Intelligent Feature Extraction for Script Classification. Acta Informatica Pragensia, 13(2), 251. https://doi.org/10.18267/j.aip.239

Damayanti, F., Yuniarno, E. M., & Suprapto, Y. K. (2024). Deep Learning-Based Inpainting for Reconstructing Severely Damaged Handwritten Javanese Characters. Register: Jurnal Ilmiah Teknologi Sistem Informasi, 10(2), 160–174. https://doi.org/10.26594/register.v10i2.4929

El Khayati, M., Kich, I., & Taouil, Y. (2024). CNN-based Methods for Offline Arabic Handwriting Recognition: A Review. Neural Processing Letters, 56(2), 115. https://doi.org/10.1007/s11063-024-11544-w

Ge, Y., Zhang, Y., & Roh, S.-B. (2025). FAIR-Net: A Fuzzy Autoencoder and Interpretable Rule-Based Network for Ancient Chinese Character Recognition. Sensors, 25(18), 5928. https://doi.org/10.3390/s25185928

Gonwirat, S., & Surinta, O. (2022). DeblurGAN-CNN: Effective Image Denoising and Recognition for Noisy Handwritten Characters. IEEE Access, 10, 90133–90148. https://doi.org/10.1109/ACCESS.2022.3201560

Guo, Y., Davy, A., Facciolo, G., Morel, J.-M., & Jin, Q. (2021). Fast, Nonlocal and Neural: A Lightweight High Quality Solution to Image Denoising. IEEE Signal Processing Letters, 28, 1515–1519. https://doi.org/10.1109/LSP.2021.3099963

Kaneko, H., Ishibashi, R., & Meng, L. (2023). Deteriorated Characters Restoration for Early Japanese Books Using Enhanced CycleGAN. Heritage, 6(5), 4345–4361. https://doi.org/10.3390/heritage6050230

Ke, W., Hou, Q., Liu, Y., Song, X., & Wei, J. (2024). SARN: Script-Aware Recognition Network for scene multilingual text recognition. Expert Systems with Applications, 250, 123753. https://doi.org/10.1016/j.eswa.2024.123753

Maheswari, S. U., Maheswari, P. U., & Aakaash, G. R. S. (2024). An intelligent character segmentation system coupled with deep learning based recognition for the digitization of ancient Tamil palm leaf manuscripts. Heritage Science, 12(1), 342. https://doi.org/10.1186/s40494-024-01438-4

Mao, J., Sun, L., Chen, J., & Yu, S. (2025). Overview of Research on Digital Image Denoising Methods. Sensors, 25(8), 2615. https://doi.org/10.3390/s25082615

Pratama, N. Z. S., & Ratnasari, C. I. (2024). Sinau: Javanese Educational Games for Early Childhood as an Effort to Preserve Javanese Heritage. Edumatic: Jurnal Pendidikan Informatika, 8(2), 645–654. https://doi.org/10.29408/edumatic.v8i2.27889

Rahman, M. G., & Islam, M. Z. (2022). A Framework for Supervised Heterogeneous Transfer Learning Using Dynamic Distribution Adaptation and Manifold Regularization. IEEE Transactions on Services Computing, 1–18. https://doi.org/10.1109/TSC.2022.3213238

Rosalina, R., Afriliana, N., Utomo, W. H., & Sahuri, G. (2024). Deep learning utilization in Sundanese script recognition for cultural preservation. Indonesian Journal of Electrical Engineering and Computer Science, 36(3), 1759. https://doi.org/10.11591/ijeecs.v36.i3.pp1759-1768

Ruldeviyani, Y., Suhartanto, H., Sotardodo, B. A., Fahreza, M. H., Septiano, A., & Rachmadi, M. F. (2024). Character recognition system for pegon typed manuscript. Heliyon, 10(16), e35959. https://doi.org/10.1016/j.heliyon.2024.e35959

Salehudin, M. A. M., Basah, S. N., Yazid, H., Basaruddin, K. S., Safar, M. J. A., Som, M. H. M., & Sidek, K. A. (2023). Analysis of Optical Character Recognition using EasyOCR under Image Degradation. Journal of Physics: Conference Series, 2641(1), 012001. https://doi.org/10.1088/1742-6596/2641/1/012001

Siahaan, D., Sutramiani, N. P., Suciati, N., Duija, I. N., & Darma, I. W. A. S. (2022). DeepLontar dataset for handwritten Balinese character detection and syllable recognition on Lontar manuscript. Scientific Data, 9(1), 761. https://doi.org/10.1038/s41597-022-01867-5

Sun, L., Li, H., & Muhammad, G. (2023). A Metaverse text recognition model based on character-level contrastive learning. Applied Soft Computing, 149, 110969. https://doi.org/10.1016/j.asoc.2023.110969

Zhang, H., Gao, G., & Bao, F. (2025). Multi-scene Traditional Mongolian OCR and Its Public Dataset. Data Intelligence, 2025. https://doi.org/10.3724/2096-7004.di.2025.0119

Zhang, Y., Zhu, Y., Peng, D., Zhang, P., Yang, Z., Yang, Z., Yao, C., & Jin, L. (2025). HierCode: A lightweight hierarchical codebook for zero-shot Chinese text recognition. Pattern Recognition, 158, 110963. https://doi.org/10.1016/j.patcog.2024.110963