<![CDATA[Japanese manga has evolved into a worldwide cultural phenomenon. Despite its popularity, many official editions of manga remain in raw format, namely untranslated Japanese versions. This creates difficulties for non-Japanese especially Indonesian readers in understanding the storyline. The swift progress of Artificial Intelligence and Machine Learning provides innovative solutions to overcome this barrier. One promising approach is the use of Optical Character Recognition technology to detect textual elements in manga images and translate them automatically with the help of object detection models such as You Only Look Once version 8. Building on this approach, the present study develops a web-based manga translation application by applying the Software Development Life Cycle method, which includes requirement analysis, design, implementation, testing, and maintenance. The system integrates You Only Look Once version 8 for panel detection, Manga Optical Character Recognition for Japanese text recognition, and translation libraries for automatic text replacement. The novelty of this research lies in offering an end-to-end translation solution that supports both image and portable document format inputs, while also providing downloadable outputs for improved usability. In this way, the study contributes to the fields of Optical Character Recognition and machine translation, while delivering a functional tool that enables Indonesian readers to access Japanese manga without language barriers.]]>
Copyrights © 2025
