人脸显性特征的融合构造方法及识别

Face Recognition Based on Explicit Facial Features by Fusion Construction Method

目前的人脸识别研究中,面部几何特征没有得到很好的利用.本文阐述了几何特征对于人脸识别的重要性,在此基础上提出了一种提取面部几何特征的新方法;通过融合几何信息和纹理信息构造出一种面部显性特征,并给出了相应的人脸识别方法.这种新的人脸识别方法相对于基于统计学习的子空间方法具有一定的优势,同时也可作为后者的有益补充.实验表明,本文提出的人脸表示特征及识别方法对人脸表情变化和环境光照变化均有一定的鲁棒性.

In the current research on face recognition,facial geometric features have not been fully utilized.Thus,the importance of geometric features in face recognition is explicated,and a novel technique of facial geometric feature extraction is proposed.Then a facial explicit feature is constructed based on the fusion of geometric and texture information.The corresponding face recognition method using these features is also given.This novel face recognition method not only possesses some advantages over the popular subspace methods based on statistical learning,but can be a complement to the latter.Experiments demonstrate that the extracted features and the corresponding face recognition algorithm are robust to facial expression and environmental illumination variations.