EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Enhancing the Extraction of Low-Quality Old English Documents Using Optical Character Recognition

Nicholas Enrico, Gede Putra Kusuma

JOIV : International Journal on Informatics Visualization · 30.03.2026 · 10 (2) · 670

DOI: 10.62527/joiv.10.2.3375 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JOIV : International Journal on Informatics Visualization, online erschienen am 30.03.2026. Verfasst von Nicholas Enrico und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study seeks to improve the accuracy of Optical Character Recognition (OCR) in extracting text from low-quality old English documents. To accomplish this, a combination of pre-processing and post-processing techniques is applied. Pre-processing techniques, including background removal, grayscale conversion, image resizing, sharpening, blurring, binarization, and noise reduction, were applied to enhance image quality before OCR processing. The OCR tool used in this study is Tesseract 4.1, which processes every single character of the entire document. Post-processing techniques, such as removing punctuation, converting to lowercase, tokenizing, and correcting text using Levenshtein distance, Hamming distance, Jaro-Winkler distance, and Sørensen-Dice coefficient, were applied to improve OCR results. These techniques are evaluated for effectiveness using the Noise OCR Dataset (NOD). The performance of these models was evaluated using the Word Error Rate (WER) metric. Among the 40 data samples tested, the model that combined pre-processing set 3 with post-processing using Levenshtein Distance had the lowest WER, at 8.654%. These outcomes demonstrate the success of the proposed method in boosting OCR accuracy, making it a useful tool for preserving and accessing historical information from low-quality documents. Based on the results of this study, several areas for further enhancement emerge. Future research could focus on applying similar pre-processing and post-processing methods to different language datasets. Additionally, exploring alternative OCR technologies could yield valuable insights into how different tools affect OCR's overall performance on low-quality documents.

Zitieren

APA

Enrico, N., & Kusuma, G. P. (2026). Enhancing the Extraction of Low-Quality Old English Documents Using Optical Character Recognition. JOIV : International Journal on Informatics Visualization, 10(2), 670. https://doi.org/10.62527/joiv.10.2.3375

Vancouver

Enrico N, Kusuma GP. Enhancing the Extraction of Low-Quality Old English Documents Using Optical Character Recognition. JOIV : International Journal on Informatics Visualization. 2026;10(2):670. doi:10.62527/joiv.10.2.3375

BibTeX
@article{Enrico2026,
  author = {Enrico, Nicholas and Kusuma, Gede Putra},
  title = {Enhancing the Extraction of Low-Quality Old English Documents Using Optical Character Recognition},
  journal = {JOIV : International Journal on Informatics Visualization},
  year = {2026},
  volume = {10},
  number = {2},
  pages = {670},
  doi = {10.62527/joiv.10.2.3375}
}
Kostenlos starten →