| 融合类别信息的二维主成分分析人脸识别算法 |
| |
| 引用本文： | 徐倩,邓伟.融合类别信息的二维主成分分析人脸识别算法[J].计算机工程与设计,2008,29(22). |
| |
| 作者姓名： | 徐倩 邓伟 |
| |
| 作者单位： | 苏州大学,计算机科学与技术学院,江苏,苏州,215006 |
| |
| 摘 要： | 二维主成分分析(2DPCA)已被成功地应用在人脸识别领域,但是这种2DPCA是无监督方法,投影没有考虑到类别信息,在一定程度上影响了识别性能.因此提出一种新的2DPCA,它利用训练样本的类别标记来生成K-L变换的产生矩阵,融合了样本的类别信息,从而使2DPCA的识别性能更好.基于ORL和Yale人脸数据库的实验表明该方法比传统的2DPCA的识别性能更高.
|
| 关 键 词： | 二维主成分分析 人脸识别 无监督方法 类别信息 K-L变换 |
2DPCA algorithm integrating class information for face recognition |
| |
| XU Qian,DENG Wei.2DPCA algorithm integrating class information for face recognition[J].Computer Engineering and Design,2008,29(22). |
| |
| Authors: | XU Qian DENG Wei |
| |
| Affiliation: | XU Qian,DENG Wei(School of Computer Science , Technology,Soochow University,Suzhou 215006,China) |
| |
| Abstract: | Two-dimensional principal component analysis(2DPCA) is successfully developed in the area of face recognition.But 2DPCA is an unsupervised technique.The projection cannot take into consideration the class information and affects the performance of recognition at certain extent.A new 2DPCA is proposed in order to make good use of the class information.The produce matrix of K-L transform according to the training data's class labeling is created.So it can fuse the class information and make the performance of... |
| |
| Keywords: | two-dimensional principal component analysis face recognition unsupervised technique class information K-L transform |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
