| 基于小波变换和部分最小二乘的掌纹识别 |
| |
| 引用本文： | 郭金玉,苑玮琦.基于小波变换和部分最小二乘的掌纹识别[J].光电子．激光,2008,19(4):554-557. |
| |
| 作者姓名： | 郭金玉 苑玮琦 |
| |
| 作者单位： | 1. 沈阳工业大学视觉检测技术研究所,沈阳,110023;沈阳化工学院信息工程学院,沈阳,110142
2. 沈阳工业大学视觉检测技术研究所,沈阳,110023 |
| |
| 摘 要： | 子空间法作为一种传统的识别方法,识别时基于整幅图像,复杂性比较高,而且没有考虑类别信息.为了降低计算复杂性和在提取数据特征的同时融入类别信息,研究了一种基于小波变换和部分最小二乘(PLS)的掌纹识别算法.在建议的识别方法中,首先通过小波三级分解提取低频子图像,对低频子图像应用PLS提取掌纹特征,然后将样本投影到提取的特征上作为特征向量进行分类识别.应用PolyU掌纹图像库进行实验分析,实验结果表明:与主元分析(PCA)、二维主元分析(2DPCA)和独立主元分析(ICA)相比,该方法的识别率得到了很大的提高,大大减小了误识率和拒识率,验证了该方法的有效性.
|
| 关 键 词： | 掌纹识别 部分最小二乘(PLS) 主元分析 二维主元分析 独立主元分析 小波变换 |
| 文章编号： | 1005-0086(2008)04-0554-04 |
| 修稿时间： | 2007年6月15日 |
Palmprint recognition based on wavelet transformation and partial least square |
| |
| GUO Jin-yu,YUAN Wei-qi.Palmprint recognition based on wavelet transformation and partial least square[J].Journal of Optoelectronics·laser,2008,19(4):554-557. |
| |
| Authors: | GUO Jin-yu YUAN Wei-qi |
| |
| Affiliation: | GUO Jin-yu 1,2,YUAN Wei-qi1(1.Computer Vision Group,Shenyang University of Technology,shenyang,110023,2.School of Information Engineering,shenyang institute of chemical technology,110142,China) |
| |
| Abstract: | A palmprint recognition method based on wavelet transformation and partial least square(PLS)is proposed in order to reduce the computational complex and consider class information when the features of images are extracted.A palmprint image was decomposed to low frequency sub-images by the three-level wavelet transform so as to decrease the computation complexity.Then a low frequency sub-image was selected for feature extraction by PLS.The original palmprint image was mapped into feature space as feature vec... |
| |
| Keywords: | palmprint recognition partial least square(PLS) principal component analysis(PCA) two-dimensional principal component analysis(2DPCA) independent component analysis(ICA) wavelet transformation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《光电子．激光》浏览原始摘要信息 |
|
点击此处可从《光电子．激光》下载全文 |
|
