| 基于改进LDA算法的人脸识别 |
| |
| 引用本文： | 齐兴敏,尹朝庆,李智博.基于改进LDA算法的人脸识别[J].计算机与数字工程,2007,35(8):31-32,37. |
| |
| 作者姓名： | 齐兴敏 尹朝庆 李智博 |
| |
| 作者单位： | 武汉理工大学,武汉,430063 |
| |
| 摘 要： | 提出一种基于改进LDA的人脸识别算法,该算法克服传统LDA算法的缺点,重新定义样本类间离散度矩阵和Fisher准则,从而保留住最有辨别力的信息,增强算法的识别率.实验结果证明该算法是可行的,与传统的PCA LDA算法比较,具有较高的识别率.
|
| 关 键 词： | LDA PCA 识别率 样本类间离散度矩阵 样本类内离散度矩阵 改进 算法比较 人脸识别 Algorithm Improved Based Recognition 结果 实验 识别率 增强算法 信息 辨别力 准则 Fisher 样本类间离散度矩阵 重新定义 识别算法 |
| 修稿时间： | 2006-09-22 |
Face Recognition Based on Improved LDA Algorithm |
| |
| Qi Xingmin.Face Recognition Based on Improved LDA Algorithm[J].Computer and Digital Engineering,2007,35(8):31-32,37. |
| |
| Authors: | Qi Xingmin |
| |
| Abstract: | This article gives a new algorithm of face recognition based on LDA for face recognition.It overcomes the weakness of the traditional LDA algorithm by redefining the matrix between samples and the Fisher rules.Thus the most important information is reserved and the recognition rate is improved.The test result also shows this algorithm is feasible and it has a higher recognition rate compared to traditional way of PCA LDA. |
| |
| Keywords: | LDA PCA recognition rate matrix between samples matrix within samples |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
