中文摘要：

针对人脸识别中的非线性特征提取问题,提出一种基于核正交局部判别嵌入（KOLDE,kernel orthogonal local discriminant embedding）的人脸识别算法。首先通过引入基向量正交约束,得到OLDE算法,并给出算法的推导过程。然后为了更好地处理高度复杂非线性结构数据,将OLDE向高维空间扩展,在核空间提取图像的高阶非线性信息,得到核空间OLDE算法。在ORL和PIE库上的人脸识别实验验证了算法的有效性。

英文摘要：

In view of the problems of nonlinear feature extraction in face recognition,an algorithm based on kernel orthogonal local discriminant embedding（KOLDE） for face recognition is proposed in this paper.Firstly,the proposed algorithm gets orthogonal local discriminant embedding（OLDE） algorithm by introducing the orthogonal constraint of basis vectors.Then the OLDE algorithm is expanded to high dimensional feature space by nonlinear mapping in order to deal with dataset of highly nonlinear structure.Finally,the kernel orthogonal local discriminant embedding algorithm is obtained by introducing the orthogonal constraint of basis vectors and traditional LDE method in kernel space.The experimental results on ORL and PIE face database illustrate the effectiveness of the proposed method.