人脸动画方法综述  

Review on methods of facial synthesis

作　　者：潘红艳[1] 柳杨华[1] 徐光祐[1] 

机构地区：[1]清华大学计算机科学与技术系

出　　处：《计算机应用研究》2008年第2期327-331,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(60273005)

摘　　要：在分析人脸动画实现的困难点的基础上,介绍了人脸动画的三大类实现方法:基于生理模型、基于三维模型和基于图像的方法。探讨了这些方法的特点,并指出了它们各自的适用领域和局限性,提出了各类人脸动画方法需要根据实际需求相互灵活结合使用的观点。This paper analyzed difficulties of realization of facial synthesis and introduced three kind of realistic methods about facial synthesis： method based on biometric model, method based on 3D facial synthesis model and method based on image. Discussed characteristic of the methods, then pointed out applicable field and localization of each method. Put forward the point of view that all kinds of methods of facial synthesis should combined with actual command.

关 键 词：人脸动画 基于生理模型 基于三维模型 基于图像 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]