摘要
为了获得具有较高识别率的算法,提出了一种将Fisher线性鉴别分析(Fisher Linear Discriminant Analysis)、复主分量分析(Principal Analysis in the Complex Space)与隐马尔可夫模型(Hidden Markov Models)相结合进行人脸识别的方法。对于输入的不同光照、人脸表情和姿势的图像先进行归一化处理,然后将归一化后的图像转化成一维向量,再用FLDA方法提取每幅图像的特征,形成新的复向量空间;通过运用复主分量分析,来抽取人脸图像的有效鉴别特征;最后通过HMM对这些特征进行训练,得到一个优化的HMM并应用于识别。在ORL人脸数据库中进行实验,实验结果表明,该方法具有较高的识别率。
In order to obtain a better recognition rate,a new algorithm which focuses on the use of FLDA,CPCA and Hidden Markov Models for face recognition is presented.First,the differernt images are translated into one-dimension vector with the same mean and variance.Second,FLDA is used to get the features of the pictures and complex vector space,then ＇CPCA is applied to get the new features and these new features are trained by HMMs.Finally,an optimized HMMs is obtained.Compared with other face recognition algorithms on the ORL face database,this method can get better recognition rate.
出处
《计算机工程与应用》
CSCD
北大核心
2008年第16期191-193,共3页
Computer Engineering and Applications
基金
教育部新世纪优秀人才支持计划(the Program for New Century Excellent Talents in University No.NCET-06-0298)
辽宁省高等学校优秀人才支持计划(the Program for Liaoning Excellent Talents in University No.RC-05-07,No.2006R06)
辽宁省教育厅科学研究计划(the Program for Study of Science of the Educational Department of Liaoning Province No.05L020)
大连市科学技术计划(the Programfor Dalian Science and Technology No.2005A10GX106)
大连大学辽宁省智能信息处理重点实验室开放课题(the Open Fund of LiaoningKey Lab of Intelligent Information Processing,Dalian University No.2005-9)
关键词
人脸识别
FISHER线性鉴别分析
隐马尔可夫模型
复主分量分析
face recognition
Fisher Linear Discriminant Analysis （FLDA）
Hidden Markov Models （HMM）
Principal Analysis inthe Complex Space（CPCA）
