earticle

논문검색

Convergence of Internet, Broadcasting and Communication

A Study on a Method for Generating a 3D Head Model from a Single 2D Face Image Using Open Source

초록

영어

Due to the COVID-19 pandemic, society's interest in the metaverse has recently increased rapidly. By using avatars, which are virtual objects that represent users within the metaverse, people enjoy a higher degree of freedom in the virtual world and various interactions with other people become possible. In most metaverse platforms, the user selects one from a limited number of pre-made avatar templates. Afterwards, the user must individually select the avatar's hair style, eyes, nose, and mouth shape from among those presented within the metaverse platform. Creating an avatar that resembles the user is not an easy task because the presented options for facial features and hairstyles are limited. In this paper, we present a method to automatically generate a 3D head model from a user's 2D face image and implement a prototype system. We show that a 3D head model creation system can be built at low cost and in a short time by utilizing open source at each stage of model creation. In this study, the process of applying facial texture to the 3D face shape was omitted and focused on 3D modelling work.

목차

Abstract
1. Introduction
2. Related Works
3. Methodology
3.1 Creating a 3D face shape from a 2D face image
3.2 Combination of a 3D face shape and a basic 3D head model
3.3 Hair segmentation
3.4 Matching 3D hair models
3.5 Combination of a 3D hair model and a 3D head model
4. System Implementation
5. Performance Evaluation
6. Conclusion
References

저자정보

  • Kyungho Kim Undergraduate student, Dept. of Computer Science, Seokyeong University, Korea
  • Woochang Shin Professor, Dept. of Computer Science, Seokyeong University, Korea

참고문헌

자료제공 : 네이버학술정보

    함께 이용한 논문

      ※ 기관로그인 시 무료 이용이 가능합니다.

      • 4,000원

      0개의 논문이 장바구니에 담겼습니다.