一种基于核主成分特征组合的人脸识别方法

A Method for Face Recognition Based on Combination of KPCA Features

提出了一种基于核主成分特征组合的人脸识别方法。首先利用主成分分析,获得原始输入图像的二阶特征脸图像;然后运用核主成分分析分别抽取原始图像和二阶特征脸图像的核主成分特征,最后将它们组合成一个组合特征向量,进行人脸识别。在ORL人脸库上的实验表明,两种图像的核主成分特征分别有着良好的特点,取得了较好的识别效果,优于核主成分分析和二阶特征脸的结果。

A new face recognition method based on combination of KPCA features is proposed in this paper.Firstly K-L transform method is used to transform initial images,then we get the second-order face image by rebuilding images, then KPCA is used to get two kinds of feature vectors for the initial image and its second-order face image.Lastly,we combine the two kinds of vectors of everyone into a longer vector.To verify the efficient of the method,we test this method on ORL face database and experiment result shows that this face recognition method is more available than kernel principal component analysis and the second-order principal component analysis.