Summary
In anime, LOD changes according to the intended scene direction. In this study, LOD represent the amount of information defined in the design, such as not only the number of polygons but also color, line thickness, the number of shadow gradations, and the degree of abstraction.
For example, characters with higher importance, such as protagonists, are often drawn with a higher level of detail. Additionally, objects close to the camera are often drawn with large changes in LOD within their internal structure. Furthermore, smaller characters are often drawn with lower detail compared to larger ones. In this study, these elements are referred to as presence weight, depth exaggeration, and sense of size, respectively.
Example of visual expressions using changes in LOD ．
By designing anime stylized LOD (Anime LOD) appropriately, it becomes possible to effectively convey depth or emphasize specific objects. Applying this technique can make visual expressions of 3DCG anime more attractive. However, research and discussions on Anime LOD are still in their early stages. In this study, therefore, we focus on character representation to establish an anime stylized LOD control model.
Results using proposed method.
Members
| Name | Affiliation | Web site |
|---|---|---|
 Masahiro Aono | Keio University |
Publications
Presentation
Unrefereed
- Masahiro Aono，Issei Fujishiro：”Proposal of an LOD control model toward anime production”，in Proceedings of Expressive Japan 2025 (The Institute of Image Information and Television Engineers Technical Report, Vol. 49, No. 10, pp.377-380, AIT2025-146), Tokyo Polytechnic University, March 10, 2025, Excellent Poster Awards, StudioGOONEYS Awards (in Japanese).
- Masahiro Aono, Issei Fujishiro: “Proposal of an LOD control model for anime production,” in Proceedings of the 87th National Convention of International Processing Society of Japan, Vol. 2, pp. 503―504 (6S-07), The University of Ritsumeikan, March 15, 2025, Student encouragement awards (in Japanese).
Grants
Grant-in-Aid for Challenging Research (Exploratory): 23K18468 (2023-)