| 探地雷达回波信号的特征提取和分类 |
| |
| 引用本文： | 周辉林,田茂,熊俊志.探地雷达回波信号的特征提取和分类[J].电波科学学报,2006,21(4):586-589. |
| |
| 作者姓名： | 周辉林 田茂 熊俊志 |
| |
| 作者单位： | 武汉大学电子信息学院,湖北,武汉,430079 |
| |
| 摘 要： | 探地雷达的回波信号具有非平稳特征.为实现非平稳信号的检测和分类,采用了二进小波变换、主分量分析与Fisher线性判别分析和前馈多层感知器分类器分别对探地雷达回波信号进行小波分解、去相关与特征选择和分类.实验结果表明,该算法具有很好的分类效果.
|
| 关 键 词： | 探地雷达 非平稳信号 特征提取与选择 分类 |
| 文章编号： | 1005-0388（2006）04-0586-04 |
| 收稿时间： | 2004-12-30 |
| 修稿时间： | 2004年12月30 |
Feature extraction and classification of ground penetrating radar echo signals |
| |
| ZHOU Hui-lin,TIAN Mao,XIONG Jun-zhi.Feature extraction and classification of ground penetrating radar echo signals[J].Chinese Journal of Radio Science,2006,21(4):586-589. |
| |
| Authors: | ZHOU Hui-lin TIAN Mao XIONG Jun-zhi |
| |
| Affiliation: | School of Electronic Information , Wuhan University, Wuhan Hubei 430079, China |
| |
| Abstract: | Ground penetrating radar (GRP) has been widely used in engineering to obtain the information of subsurface structure and object. To achieve feature extraction and classification from GPR echo signal of inherently non-stationary characteristic, this paper employs the dyadic wavelet transform,principal components analysis,Fisher linear discriminant analysis and feedforward multi-layer perceptron classifier, respectively, to deal with the wavelet decomposition,decorrelation,feature selection and classification. The measured GPR data obtained from the Mei-shan railway show that the presented schmes have good performance of classification. |
| |
| Keywords: | ground penetrating radar non-stationary signal decorrelation feature extraction and selection classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
