| 基于PIDC和二叉决策树SVM的人耳识别 |
| |
| 引用本文： | 张伟伟,夏利民,欧阳军林.基于PIDC和二叉决策树SVM的人耳识别[J].计算机工程与应用,2006,42(27):181-183,193. |
| |
| 作者姓名： | 张伟伟 夏利民 欧阳军林 |
| |
| 作者单位： | 中南大学信息科学与工程学院,长沙,410075 |
| |
| 基金项目： | 国家自然科学基金;中国科学院基金 |
| |
| 摘 要： | 提出了一种基于PIDC和二叉决策树SVM的人耳识别方法。采用PIDC方法以类间概率信息距离为监督提取人耳特征,降低了提取特征的维数;将PIDC方法与二叉决策树SVM分类方法相结合,实现了利用多类间概率信息距离监督人耳特征提取和分类。利用该方法对400个人耳进行识别实验,并将识别结果同PCA方法进行了比较,实验表明,文中方法降低了分类难度,提高了人耳识别率。
|
| 关 键 词： | 特征提取 PIDC 二叉决策树SVM 人耳识别 |
| 文章编号： | 1002-8331-（2006）27-0181-03 |
| 收稿时间： | 2006-05-01 |
| 修稿时间： | 2006-05-01 |
Ear Recognition Based on PIDC and Binary Tree SVM Classification |
| |
| ZHANG Wei-wei,XIA Li-min,OUYANG Jun-lin.Ear Recognition Based on PIDC and Binary Tree SVM Classification[J].Computer Engineering and Applications,2006,42(27):181-183,193. |
| |
| Authors: | ZHANG Wei-wei XIA Li-min OUYANG Jun-lin |
| |
| Affiliation: | College of Information Science and Engineering,Central South University,Changsha 410075 |
| |
| Abstract: | In this paper,we present an ear recognition method using PIDC and binary tree SVM classification.On the one hand,we extract ear features using PIDC which is a kind of supervised algorithm regarded for the PID between two classes.And the PIDC algorithm decreases the feature dimension under not decreasing recognition effect.On the other hand,we build a binary tree SVM classification system so that we can use Multi-class PID to guide the feature extracting process.At last,we compare the experiment results between our method and PCA using 400 ear samples,the results show that our method can reduce the difficulties of calculation and get high recognition rate. |
| |
| Keywords: | PIDC |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
