分块式双向压缩的二维主成分分析  

Two-dimensional principal component analysis with block bidirectional compression

作　　者：于彬 沈永良[1] YU Bin;SHEN Yong-Liang(School of Electronic Engineering,Heilongjiang University,Harbin 150080,China)

机构地区：[1]黑龙江大学电子工程学院,哈尔滨150080

出　　处：《黑龙江大学工程学报》2020年第4期68-74,共7页Journal of Engineering of Heilongjiang University

基　　金：国家自然科学基金项目(61503127)。

摘　　要：人脸识别是生物识别研究热点问题之一,目前,大多数传统的人脸识别算法运算速度慢,人脸识别精度较低。对此,提出了一种分块式双向压缩的二维主成分分析与径向基核函数支持向量机相结合的算法。将人脸图像分割为大小相等并且互不重合的子块,每个子块都包含重要的特征信息。使用双向压缩的2DPCA与PCA相结合的算法来进行人脸特征的提取,可以有效地减少特征数量和PCA的计算时间,充分地保留重要的信息,再与支持向量机相结合,其运算时间和训练时间都充分降低,并且提高了识别率。在ORL、Yale和自建的人脸库上的实验表明,该方法的运算速度和识别率明显高于传统的识别方法。Face recognition is one of the current hot issues in biometrics research.At present,most traditional face recognition algorithms have slow calculation speeds and low face recognition accuracy.In this regard,an algorithm combining two-dimensional principal component analysis with block-wise double compression and radial basis kernel function support vector machine is proposed.First,the face image is divided into sub-blocks of equal size and not overlapping each other,and each sub-block contains important feature information.Then use the two-way compressed 2DPCA and PCA algorithm to extract facial features,which can effectively reduce the number of features and the calculation time of PCA,and fully retain important information;and then combine with the support vector machine,its calculation time and training time are fully reduced,and the recognition rate is improved.Experiments on ORL,Yale and self-built face database show that the calculation speed and recognition rate of this method are significantly higher than traditional recognition methods.

关 键 词：人脸识别 图像处理 二维主成分分析 支持向量机 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]