| 摘要 |
An avatar is generated by i) providing: a) a three-dimensional representation (2) of the structure of the character, the structure representation changing in real time to represent movement of the character, b) a three-dimensional representation of the surface (1) of the character which is mapped onto said structure representation, and c) a two-dimensional representation of frequently changing portions of said surface, eg the portions used to generate facial expressions and ii) combining said representations a), b) and c) to generate a combined representation of the moving character. Because the frequently changing portions are represented only two-dimensionally, less processing power and ROM is needed to display the avatar.
|
