Kinect驱动的人脸动画合成技术研究  

Research on Synthesis Technology of Facial Animation Driven by Kinect

作　　者：李俊龙[1] 章登义[1] 黄珺[2] 

机构地区：[1]武汉大学计算机学院,武汉430079 [2]武汉大学测绘遥感信息工程国家重点实验室,武汉430079

出　　处：《计算机工程》2015年第3期237-241,共5页Computer Engineering

基　　金：教育部博士点基金资助项目(20120141120006);武汉科技计划基金资助项目(2013010501010146)

摘　　要：三维人脸动画合成技术可以应用于虚拟现实、角色控制等多个领域。为此,提出一种基于Kinect的人脸动画合成方法。人脸跟踪客户端利用Kinect对用户的脸部表情进行跟踪识别,得到人脸表情动画参数,通过socket发送给人脸动画合成服务器,由人脸动画合成服务器查找基于MPEG-4标准的人脸动画定义表,控制人脸模型的变形,从而实时合成与用户表情相对应的三维人脸动画。实验结果表明,该方法能够在满足实时性要求的条件下合成高质量的三维人脸动画,同现有技术相比,结构简单、容易部署且具有较好的可扩展性。The technique of 3D human facial animation can be applied in a wide range of fields such as virtual reality,avatar control and so on. This paper proposes a real-time facial animation system based on Kinect. In this system,facial expression parameters are inferred by facial tracking client with Kinect and sent to the facial animation synthesizing server through socket. The server looks up Face Def Tables based on MPEG-4 facial animation standard to control the morphing of 3D face model.3D facial animations which emulate expressions of the user are then generated in real time.Experimental results show that the method can rebuild quality realistic 3D facial animations in real time. Compared with existing technologies,this system with simple structure is easy to deploy and is also extendibility.

关 键 词：Kinect驱动 人脸跟踪 人脸动画 合成 MPEG-4标准 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]