Scholarlink

Given the rapid development of 5G transmission, information technology, and the evolution of people’s usage habits, there has been a significant increase in demand for visual communication, primarily in the form of videos and animated graphics. Coupled with advancements in AR/VR devices, this trend is gradually transitioning towards the era of Internet 3.0 and the metaverse. The demand for mapping text into images/videos in cyberspace is consequently growing. Therefore, it is of great significance to explore a low-cost pathway for generating a textual mapping metaverse space. Drawing inspiration from the three-dimensional spatial axonometric drawing techniques in architecture, this paper addresses the lack of image-based historical records in the study of ancient cities. It proposes a method for generating vectorized, axonometric images from text, with clear scene relationships. Furthermore, by introducing the concept of “narrative”, the generated scene graphs are extrapolated into animations, providing a technical pathway for lightweight generation of metaverse scenes from textual content.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

In the Context of Metaverse: Text to Animated Axonometric Space of Ancient Cities Generation

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年

查看全文
  • Mengyao Li,
  • Zheng Zhang,
  • Xin Yan,
  • Keyang Tang,
  • Jiaxin Ke

摘要

Given the rapid development of 5G transmission, information technology, and the evolution of people’s usage habits, there has been a significant increase in demand for visual communication, primarily in the form of videos and animated graphics. Coupled with advancements in AR/VR devices, this trend is gradually transitioning towards the era of Internet 3.0 and the metaverse. The demand for mapping text into images/videos in cyberspace is consequently growing. Therefore, it is of great significance to explore a low-cost pathway for generating a textual mapping metaverse space. Drawing inspiration from the three-dimensional spatial axonometric drawing techniques in architecture, this paper addresses the lack of image-based historical records in the study of ancient cities. It proposes a method for generating vectorized, axonometric images from text, with clear scene relationships. Furthermore, by introducing the concept of “narrative”, the generated scene graphs are extrapolated into animations, providing a technical pathway for lightweight generation of metaverse scenes from textual content.