Pencarian Informasi Berbasis Teks Dalam Komik Digital Menggunakan OCR

Abstract

This research is aimed to research and develop a web-based application which will enable user to find texts in a digital comic pages using OCR. The methods that will be used to achieve this is a neural network learning approach using YOLOv3 algorithm that will enable a program to detect text in a comic page and a text recognition engine developed by google called Tesseraact that will enable the program to read texts in a comic page to a functional format such as string text. The software development cycle that is used is the incremental model and the user interface will be made using a python based web framework called Flask using Hyper Text Markup Language, Cascading Style Sheet and Javascript for searching functionality