Abstract: In 3D face synthesis,the lip movement is the most expressive.But in Uyghur lip animation,there has not been a realization method of 3D lip animation system.Therefore,this paper proposes a lip animation system.On the basis of the standard model and according to the current mainstream thinning algorithm,it uses the formula transformation to get the new transformation and model driven method,realizing fast refinement and the motion of the refined model.Experimental results show that the system can achieve the generation of facial details and smooth movement of the transition,so that the detail of the model is full,with a strong sense of reality.

Key words: Candide model, facial animation, Loop subdivision, feature extraction, texture mapping, MPEG-4 standard

摘要： 在三维人脸合成中,最具有表现力的是唇部运动,而在维吾尔语唇形动画中,完整的基于维吾尔语的三维唇部动画系统的实现方法研究较少。为此,提出一种唇部动画系统。在标准模型的基础上,根据目前主流的细化算法,采用公式变换得到新的变换及模型驱动方式,实现模型的快速细化和细化后模型的运动。实验结果表明,该系统能实现人脸细节的生成及平滑流畅的运动过渡,使模型细节饱满,具有较强的真实感。

关键词: Candide模型, 人脸动画, Loop细分, 特征提取, 纹理映射, MPEG-4标准

CLC Number: 

• TP391.41