摘要
提出了一种基于"分集"的人脸识别方法,该方法在人脸识别前加入脸型预分类环节,将一个大规模的人脸数据库分解为依脸型分类的若干个子库。这样做一方面可通过分集降低后续识别处理的数据量,提高人脸识别的速度,另一方面可利用脸型特征对候选人脸集合进行粗筛选,降低系统的错误接受率。为了实现脸型分类,进一步提出了一种基于人体测量学的分类方法,即首先借助AAM技术提取脸部特征点,然后在此基础上计算面型指数,并由此实现对脸型的分类。对较大规模的人脸数据库所进行的实验结果表明,所提出的方法可有效提高人脸识别系统的识别率和识别速度。
A pre-classification based face recognition method is presented in this paper. According to the face shape, large-scale database is divided into several smaller sub-databases. It will decrease the amount of data in subsequent face recognition so as to speed up the recognition. On the other hand, it will get rid of the potential non-candidates by using the features of face shape so as to decrease the false acceptance rate. An anthropometry based method is proposed in face shape classification. First AAM technique is used to extract feature points in face images. Then facial indexes are computed according to the feature points. Finally the face images are classified based on the facial indexes of them. Experiments show that the presented method can remarkably improve the recognition speed and rate of a face recognition system with large-scale database.
出处
《电子技术（上海）》
2009年第11期77-79,68,共4页
Electronic Technology
