基于聚类建模的三维人脸识别技术研究

Research of 3D Face Recognition Technology Based on the Cluster Modeling

由于信息采集困难、环境条件约束、实现方式和手段不足等原因,三维人脸识别技术还很不成熟.为此提出将聚类技术应用到三维人脸建模过程中来提高建模的效果和效率.首先定义了三维人脸相似性,提出了圆锥曲线相似性定义方法.其次基于三维人脸聚类建模提出了人脸识别系统的新框架,设计了与新系统对应的识别策略.实验证明,基于聚类建模的人脸识别系统在进行人脸识别时所用的时间远远少于采用传统形变模型的方法所用的时间,而且对人脸样本的数量不敏感.

Abstract： As the information collection difficulties, environmental conditions, lack of ways and means to achieve and so on, the realization of 3D face recognition is still immature. This paper applied cluster analysis techniques to 3D face modeling, 3D face similarity should come first in the framework, we reasonably defined similarity by Conic Affinity and designed a no- vel recognition strategy for our new framework of face recognition system based on cluster modeling, experiments show that clustering based modeling of face recognition system dur- ing the time when the spending is far lower than the traditional deformable model approach, but also to face the namber of samples is not sensitive.