A Study on the Development of a Webtoon Background Asset Production Platform Utilizing Generative AI Technology

Abstract

This study proposes a generative AI-based platform aimed at automating the creation of background assets in the webtoon production process. The platform focuses on automating the production of background assets, which, while not critical to the narrative, are essential elements of webtoons. By doing so, it seeks to alleviate the intensive workload of webtoon creators and enhance production efficiency. The platform was implemented using React and TypeScript, leveraging Stable Diffusion technology and the LoRA model to generate background assets for webtoons. The research methodology included a literature review, platform development, and experimental application to actual webtoons. The findings indicate that the proposed platform enables the generation of backgrounds solely through prompt input, without requiring additional software installation or the purchase of modeling tools. Furthermore, it efficiently handles both image generation and refinement simultaneously, demonstrating high operational efficiency. However, limitations were observed in the accuracy of foreground objects and human representation. As such, complementary use with existing 3D modeling tools is deemed effective.