计算机工程 ›› 2008, Vol. 34 ›› Issue (4): 198-200.doi: 10.3969/j.issn.1000-3428.2008.04.070

• 人工智能及识别技术 • 上一篇    下一篇

基于部件和支持向量机的人脸分类

罗 瑜1,李 涛2,何大可3,徐 图3   

  1. (1. 电子科技大学计算机科学与工程学院,成都 610054;2. 四川文理学院计算机科学系,达州 635000;3. 西南交通大学信息科学与技术学院,成都 610031)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2008-02-20 发布日期:2008-02-20

Face Classification Based on Component and Support Vector Machine

LUO Yu1, LI Tao2, HE Da-ke3, XU Tu3   

  1. (1. School of Computer Science & Engineering, University of Electronic Science and Technology of China, Chengdu 610054;2. Department of Computer Science, Sichuan University of Arts and Science, Dazhou 635000; 3. School of Information Science & Technology, Southwest Jiaotong University, Chengdu 610031)
  • Received:1900-01-01 Revised:1900-01-01 Online:2008-02-20 Published:2008-02-20

摘要: 提出一种基于部件的人脸分类方法,将人脸部件的离散余弦变换系数作为特征向量,通过支持向量机训练分类器。部件分类器确定了人脸图像中的部件区域,人脸分类器确定了人脸图像的所属类别。ORL人脸图像数据库仿真实验表明,该方法对表情、姿态变化具有很好的鲁棒性。

关键词: 人脸分类, 支持向量机, 人脸部件, 离散余弦变换

Abstract: This paper proposes a method for face classification based on face component. Discrete Cosine Transform(DCT) coefficients are extracted as feature vectors from the face component image, Support Vector Machine(SVM) is used to train component and face classification machine. Component classification machine distinguishes the component regions in the face image, and face classification machine determines which person the image should belong to. Based on ORL face image database, experimental results illustrate that the method is insensitive to expression and pose variations.

Key words: face classification, Support Vector Machine(SVM), face component, Discrete Cosine Transform(DCT)

中图分类号: