计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (26): 230-232.

• 工程与应用 • 上一篇    下一篇

基于核独立成分分析和BP网络的人脸识别

陈玉山,席 斌   

  1. 厦门大学 信息科学与技术学院 自动化系 模式识别与智能系统,福建 厦门 361005
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-09-11 发布日期:2007-09-11
  • 通讯作者: 陈玉山

Face recognition based on KICA and BP neural network

CHEN Yu-shan,XI Bin   

  1. Department of Automation,School of Information Science and Technology,Xiamen University,Xiamen,Fujian 361005,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-09-11 Published:2007-09-11
  • Contact: CHEN Yu-shan

摘要: 用基于非线性子空间的核独立成分分析方法(KICA)对人脸图像进行特征提取,用三层的BP网络作为分类器,对人脸进行识别。在简单介绍基本的独立成分分析(ICA)的基本原理的基础上,对KICA的原理和算法作了详细的描述,并详细介绍了三层BP网络的设计。最后为了验证KICA+BP网络的效果,进行对比实验和分析。实验和分析的结果表明,在人脸识别中,该方法的效果明显好于其它方法。

关键词: 人脸识别, 独立成分分析, 核独立成分分析, BP网络

Abstract: The Kernel Independent Component Analysis(KICA) based on an entire function space of nonlinear subspace is used for feature extraction,and BP neural network is used as the classifier.First explains the basic concept of ICA in a concise way,then the KICA’s basic principle and algorithm and how to construct the BP neural network is discussed,finally the experimental and analysis results show that in the face recognition KICA+BP algorithm outperforms other algorithm.

Key words: face recognition, ICA, KICA, BP neural network