电子学报2009，Vol.37Issue(10)：2211-2214,4.
基于DCT和线性判别分析的人脸识别
Face Recognition Based on DCT and LDA
尹洪涛 1付平 1沙学军2
作者信息
- 1. 哈尔滨工业大学自动化测试与控制研究所,哈工大科学园3033信箱,黑龙江哈尔滨,150080
- 2. 哈尔滨工业大学通信技术研究所,黑龙江哈尔滨,150080
- 折叠
摘要
Abstract
A face recognition method based on the discrete cosine transform and linear discriminant analysis is presented. Discrete cosine transform cannot compress the data and it only maps the resource data to another data field. How to select the DCT coefficients that are most effective to classify is an important problem. In order to obtain the effective features,a feature selection algorithm according to the separability discriminant is use to select the DCT coefficients. The experiments on ORL and Yale face databases show that the improved method is effective.关键词
人脸识别/离散余弦变换/线性判别分析Key words
face recognition/ DCT(discrete cosine transform)/ LDA( linear discriminant analysis)分类
信息技术与安全科学引用本文复制引用
尹洪涛,付平,沙学军..基于DCT和线性判别分析的人脸识别[J].电子学报,2009,37(10):2211-2214,4.
