EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Implementation of the YOLOv11s Model for Braille Text Detection in an Android-Based Application

Zeva Lorensia Naibaho, Wiyli Yustanti, Widi Aribowo, Hisham A. Shehadeh

Journal of Artificial Intelligence and Data Science · 2026

DOI: 10.67561/jaids.v2i2.1 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Artificial Intelligence and Data Science, 2026. Verfasst von Zeva Lorensia Naibaho und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Braille is how a blind person reads and gets access to information, but anyone who hasn't learned the system usually struggles to understand it, and that gap can get in the way of communication and support. There's also the detection side of things, Braille dots are small and sit close together, which makes picking them out automatically genuinely difficult. This study builds a system for detecting Grade 1 Braille characters using YOLO11s, meant to run on an Android device. The dataset pulls together 4,086 Braille images across 42 character classes, alphabet letters included along with various Braille symbols. Multiple training setups were tested, different optimizers, learning rates, epoch counts, image sizes, mixing and matching until the best combination came out on top. The selected model was then deployed onto Android via TensorFlow Lite, with real-time testing carried out under different lighting and angles. That winning setup landed on 0.903 precision, 0.853 recall, mAP@0.5 at 0.901, and mAP@0.5–0.95 at 0.655. On the test set, the numbers shift a bit, precision drops to 0.884, recall to 0.852, mAP@0.5 to 0.87, mAP@0.5–0.95 to 0.64. Real-time testing also gave promising results across different lighting and positions. Beyond just detection, the study also shows how the detected text can be turned into speech using Text-to-Speech, pointing to the system's potential for real-world use.

Zitieren

APA

Naibaho, Z. L., Yustanti, W., Aribowo, W., & Shehadeh, H. A. (2026). Implementation of the YOLOv11s Model for Braille Text Detection in an Android-Based Application. Journal of Artificial Intelligence and Data Science. https://doi.org/10.67561/jaids.v2i2.1

Vancouver

Naibaho ZL, Yustanti W, Aribowo W, Shehadeh HA. Implementation of the YOLOv11s Model for Braille Text Detection in an Android-Based Application. Journal of Artificial Intelligence and Data Science. 2026. doi:10.67561/jaids.v2i2.1

BibTeX
@article{Naibaho2026,
  author = {Naibaho, Zeva Lorensia and Yustanti, Wiyli and Aribowo, Widi and Shehadeh, Hisham A.},
  title = {Implementation of the YOLOv11s Model for Braille Text Detection in an Android-Based Application},
  journal = {Journal of Artificial Intelligence and Data Science},
  year = {2026},
  doi = {10.67561/jaids.v2i2.1}
}
Kostenlos starten →