학술논문

A New Deep Learning-Based Mobile Application for Komering Character Recognition
Document Type
Conference
Source
2022 5th International Seminar on Research of Information Technology and Intelligent Systems (ISRITI) Research of Information Technology and Intelligent Systems (ISRITI), 2022 5th International Seminar on. :294-299 Dec, 2022
Subject
Bioengineering
Communication, Networking and Broadcast Technologies
Computing and Processing
General Topics for Engineers
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Deep learning
Training
Handwriting recognition
Image recognition
Computational modeling
Training data
Mobile applications
Komering scripts
image recognition
Deep Learning
Convolutional Neural Network
Android app
Language
Abstract
The Komering script, also known as the Ulu script, is one of the scripts that must be conserved in the South Sumatra region. Some historical manuscripts are composed in the Komering script, which requires transliteration. Using technology, this study intends to preserve and revitalize the Komering script. The application technology developed can assist in reading documents written in the Komering script. The produced mobile application technology product recognizes images using deep learning technology. The Convolutional Neural Network classification algorithm is used to categorize Komering characters. The trained deep learning model can classify 336 Komering characters with a 99% accuracy rate. The learned deep learning model is subsequently incorporated into the Android application to facilitate the program's portability.