摘要
结合Foley-Sammon最佳鉴别矢量集方法和统计不相关最佳鉴别矢量集方法,本文给出了正交线性无关鉴别矢量集方法。为了对已得到的鉴别矢量集方法进行分析比较,使用Matlab鳊程实现了有关算法,并采用UCI中典型数据集以及ORL人脸图像数据库进行测试。根据测试结果,对各种鉴别矢量集方法做出了比较。
In this paper, combining Foley-Sammon transformation(FST) and uncorrelated discriminant vector method(UDV), uncorrelated and orthogonal discriminant vector method(UODV) is derived. In order to analyze various discriminant vector methods, some relevant feature extraction algorithms are programmed using Matlab, and tested on the classical datasets from UCI as well as ORL face database. Based on the results of test, all discriminant vector methods concerned are compared.
出处
《闽江学院学报》
2004年第5期41-43,共3页
Journal of Minjiang University
