| 基于核的图像欧氏距离人脸识别 |
| |
| 引用本文： | 郝建东,张伟伟.基于核的图像欧氏距离人脸识别[J].计算机工程与设计,2011,32(11):3844-3847. |
| |
| 作者姓名： | 郝建东 张伟伟 |
| |
| 作者单位： | 1. 解放军理工大学理学院基础电子学系,江苏南京,211101
2. 南京春海科技有限公司软件研发部,江苏南京,210012 |
| |
| 摘 要： | 引进了两幅图像之间的一种新的距离度量方法——图像欧氏距离,该距离是利用核函数对传统的欧氏距离进行改进而得到的。在此基础上,设计了一种新的分类识别方法——基于核的图像欧氏距离人脸识别方法,并应用于人脸识别中。为验证该算法的可行性,对人脸图像进行DCT变换得到预处理样本,并在ORL和Yale人脸库上进行多角度的比较实验。分析实验结果表明,该方法优于其它距离分类器算法。
|
| 关 键 词： | 人脸识别 图像距离 欧式距离 核 分类器 |
Image Euclidean distance based on kernel for face recognition |
| |
| HAO Jian-dong,ZHANG Wei-wei.Image Euclidean distance based on kernel for face recognition[J].Computer Engineering and Design,2011,32(11):3844-3847. |
| |
| Authors: | HAO Jian-dong ZHANG Wei-wei |
| |
| Affiliation: | HAO Jian-dong1,ZHANG Wei-wei2(1.Department of Foundation Electronics,College of Science,PLA University of Science and Technology,Nanjing 211101,China,2.Department of Software Research and Development,Nanjing Chunhai Science and Technology Limited Company,Nanjing 210012,China) |
| |
| Abstract: | A new distance named Euclidean distance in the image is proposed,which consists of the idea of kernel function.On this basis,a novel recognition method called image Euclidean distance based on kernel for face recognition is designed.In order to verify the feasibility of the proposed method,firstly,it takes pretreatment to facial images to get pretreatment samples.The experiments show that this new algorithm outperforms other distance classifier algorithms. |
| |
| Keywords: | face recognition image distance Euclidean distance kernel distance classifier |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
