摘要
提出了一种基于支持向量机的人脸分类方法,首先对人脸图像作二维离散余弦变换,取离散余弦变换系数作为特征,然后用支持向量机进行分类。用Essex人脸图像数据库进行性别分类,取得了很好的分类效果。
A method for face classification based on support vector machine(SVM) has been proposed in this paper. Discrete cosine transform (DCT) is used to the face image, resulting in its DCT coefficients as feature. The feature is fed to SVM for classification. The Essex face image database is selected to evaluate the method of sex classification. The results show that the method gives higher accuracy.
出处
《计算机工程》
CAS
CSCD
北大核心
2004年第11期110-112,共3页
Computer Engineering
关键词
人脸识别
支持向量机
离散余弦变换
Face recognition
Support vector machine(SVM)
Discrete cosine transform
