Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Jurnal Nasional Pendidikan Teknik Informatika (JANAPATI)
Hendra Kusuma
Electrical Engineering Department, Institut Teknologi Sepuluh Nopember, East Java, Indonesia
Computer Science & IT Education Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Real-Time Deep Learning and OCR for Medication Recognition in Visually Impaired Users: A Systematic Literature Review Qory Hidayati; Hendra Kusuma; Muhammad Attamimi
Jurnal Nasional Pendidikan Teknik Informatika: JANAPATI Vol. 15 No. 2 (2026)
Publisher : Prodi Pendidikan Teknik Informatika Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/janapati.v15i2.112489

Abstract

Medication misidentification remains a safety risk for visually impaired and low-vision users who manage medicines independently. This systematic literature review synthesizes eligible studies on real-time medication recognition systems on real-time medication recognition systems that combine deep learning (DL), optical character recognition (OCR), and accessible feedback mechanisms. Following PRISMA 2020, searches were conducted in IEEE Xplore, Scopus, PubMed, Web of Science, ACM Digital Library, and ScienceDirect for studies published from 2019 to July 2025. Fifty studies met the eligibility criteria. Of these, 30 reported quantitative accuracy and/or latency metrics suitable for comparative extraction. The evidence was synthesized across platform type, recognition technique, evaluation dimension, multimodal integration, and deployment feasibility. Reported accuracies ranged from approximately 70% to 99%, but direct comparison remains limited by differences in datasets, number of medicine classes, image conditions, and metric definitions. Latency evidence indicates that several mobile and embedded systems support sub-second to near-real-time feedback, while cloud-assisted systems may require longer response times. Mobile, wearable, embedded, and multimodal approaches show complementary strengths, yet persistent gaps remain in public datasets, real-world validation with visually impaired users, robust multilingual OCR, transparent quality reporting, and healthcare-system integration. This review contributes a structured synthesis of DL-OCR techniques, accessibility features, and deployment constraints for assistive medication recognition.
Co-Authors Muhammad Attamimi Qory Hidayati