摘要
为了提高人脸图像识别的速度和准确率,将多元统计分析中的降维方法引入到人脸图像识别系统中,分析并对比了主成分分析法(PCA)线性判别分析法(LDA)和局部保留投影(LPP)三种方法用于人脸图像识别的效果,采用Yale B作为识别的样本和测试库,在光照条件和姿态变化条件下对三种算法进行了比较实验。实验结果表明,三种方法对人脸图像识别均能够达到一定效果,在姿态变化条件下,LDA和LPP均优于PCA,但三种方法对于光照变化均不能保持很好的鲁棒性。
In order to improve the speed and accuracy of face image recognition,the dimension reduction method of muItivariate statisticaI anaIysis is introduced into the face image recognition system.The PrincipIe Component AnaIysis,Linear Discriminate AnaIysis and LocaIity Preserving Projection are anaIyzed and compared.In the condition of iI umination and pos-ture change,the comparing experiments for the three aIgorithms are carried out using YaIe B as the identified sampIes and test Iibrary in this paper.
出处
《工业控制计算机》
2015年第7期110-112,共3页
Industrial Control Computer
基金
国家自然科学基金项目(61379079)
