l1范数最近邻凸包分类器在人脸识别中的应用

l1范数作为重要的距离测度，在模式识别中有着较为广泛的应用。在不同的范数定义下，相同分类机理的分类算法一般会有不同的分类效果。本文提出l1范数下的最近邻凸包人脸识别算法。该算法将最近邻凸包分类算法的范数定义由l2范数推广到l1范数，以测试点到各训练类凸包的l2范数距离作为最近邻分类的相似性度量。在ORL标准人脸数据库上的验证实验中，该方法取得了良好的识别效果。

l1 Norm Nearest Neighbor Convex Hull Classifier for Face Recognition

As an important distance measure, l1 norm is widely used in pattern classification. In general, under different distance measure definition, the classifiers with same mechanism will have different performances. In this paper, a face recognition algorithm based on l1 norm nearest neighbor convex hull classifier (l1 NNCH)is presented. The classifier replaces the l2 norm of Nearest Neighbor Convex Hull algorithm with l1 norm. The l1 distance from a test point to the convex hull of a class training set is taken as the similarity measurement of nearest neighbor rule. The experiments on the ORL face database show the good performances of l1 NNCH.