一种基于分块小波的人脸识别算法

来源期刊:中南大学学报(自然科学版)2013年第5期

论文作者:杨淑平 易国栋 袁修贵 刘再明

文章页码:1902 - 1909

关键词:分块小波变换;PCA+LDA;KNN;支持向量机;分类器组合

Key words:blocking wavelet transform; PCA+LDA; KNN; SVM; classifications combination

摘    要:提出一种基于分块小波的人脸识别新算法。在充分考虑提取局部特征,又克服小样本问题的基础上,提出分块小波的概念。首先,对小波分解后的低频子图进行分块,提取局部特征,从而降低图像维数并除去冗余噪声;将其先后进行PCA和LDA变换,得到组合特征向量;最后,根据KNN的快速分类能力及SVM在少数类别分类上的优势,提出KNN+SVM融合分类器对组合特征向量进行分类识别。研究结果表明:该方法识别率高,识别速度快,具有一定的实用价值。

Abstract: A new face recognition based on the blocking wavelet transforms was proposed. To fully extract local features, and overcome the small sample size problem, the concept of block wavelet was presented. First the local characteristics of the face were extracted by the blocking methods after wavelet decomposition of low frequency sub-band. The dimensions were reduced and the redundancy noise was eliminated. After that, the PCA and LDA transformations were taken to obtain the combine eigenvectors. Finally, taking advantage of the fast classification ability of the KNN and the high recognition rate of SVM on a few category classifications, a new blend classifier named KNN+SVM was constructed to recognize the combined eigenvectors. The results show that the new method has high recognition rate, fast identifying speed, and practical value.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号