首页 | 本学科首页   官方微博 | 高级检索  
     

基于PCA/ICA的人脸特征提取新方法
引用本文:范群贞,刘金清.基于PCA/ICA的人脸特征提取新方法[J].电子测量技术,2010,33(8):31-34.
作者姓名:范群贞  刘金清
作者单位:福建师范大学,物理与光电信息科技学院,福州,350007
基金项目:福建省教育厅科技项目 
摘    要:人脸自动识别方法已成为当前模式识别和人工智能领域的一个研究热点,人脸特征提取是人脸识别技术的关键。首先介绍了人脸图像的预处理过程,PCA和ICA算法的原理,针对PCA和ICA算法本身优缺点,提出一种新的基于PCA/ICA的人脸特征提取方法,最后采用最近邻分类器,对ORL人脸库进行分类识别。实验证明,改进的方法优于PCA算法和ICA算法。

关 键 词:人脸识别  特征提取  人脸图像  主成分分析法(PCA)  独立成分分析法(ICA)

Novel face feature extraction method based on PCA/ICA
Fan Qunzhen,Liu Jinqing.Novel face feature extraction method based on PCA/ICA[J].Electronic Measurement Technology,2010,33(8):31-34.
Authors:Fan Qunzhen  Liu Jinqing
Affiliation:Fan Qunzhen Liu Jinqing(School of Physics and Opto-Electronics Technology,Fujian Normal University,Fuzhou 350007)
Abstract:Face recognition is attractive in pattern recognition and artifical intelligence field,face feature extraction is important in face recognition.This paper first introduced preprocessing of face images,PCA and ICA algorithm.Considering PCA and ICA algorithm each have its advantages and disadvantages,then proposes a novel face feature extraction method based on PCA/ICA.The NN classifier,is select to face classification and recognition on the ORL face database.Experiments indicated that the modified method is supe rior to PCA algorithm and ICA algorithm.
Keywords:face recognition  feature extraction  face image  principle component nalysis (PCA)  independent component analysis (ICA)
本文献已被 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号