Implementation of Digital Transformation for Korean Traditional Heritage

Jae-Ho Lee, Hye-Bin Lee, Hee-Kwon Kim, Chan-Woo Park

Abstract: In this paper, we introduce technologies for transforming Korean traditional heritage into digital assets and applying them to various platforms. To transform low resolution and distorted historic images to high-resolution usable digital assets, we have trained deep learning models accustomed to analyzing traditional heritages. We have also established steps and specifications for acquiring high-quality 2D/3D models of traditional heritage for the generation of digital assets. Korean natural language processing and object detection models for analyzing relics tagged with historic information were also implemented for extracting relations and creating an intelligent search system. With the attained digital assets, a web-based intelligent database system was built regarding of an intuitive UI that allows museum curators to easily upload and retrieve the assets needed. Arranging and presenting the attained 3D digital heritage through VR/AR/WEB platforms were achieved through adequate transformation of data formats. By demonstrating and storing traditional relics thorough digital platforms, we expect further development of Korean traditional heritage preservation and education.

External IDs:dblp:conf/fcv/LeeLKP22