摘要
论文提出了一种新的基于三维人脸形变模型,并兼容于MPEG-4的三维人脸动画模型。采用基于均匀网格重采样的方法建立原型三维人脸之间的对齐,应用MPEG-4中定义的三维人脸动画规则,驱动三维模型自动生成真实感人脸动画。给定一幅人脸图像,三维人脸动画模型可自动重建其真实感的三维人脸,并根据FAP参数驱动模型自动生成人脸动画。
A novel 3D facial animation model is proposed in this paper to generate realistic facial animations.The 3D facial animation model is constructed based on morphable model and compatible with MPEG-4.Uniform mesh resampling method is put forward to align prototypical 3D faces and facial animation principle defined in MPEG-4 is adopted to drive the 3D facial animation model.The animation model can automatically reconstruct realistic 3D face of specific person given a single image and be driven by FAP parameters to generate 3D facial animations automatically.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第6期33-35,44,共4页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:60375007)
北京市自然科学基金资助项目(编号:Kz200310005002)
北京市教委资助项目(编号:Km200310005025)