基于单演局部主方向能量模式的人脸识别方法  

FACE RECOGNITION ALGORITHM BASED ON MONOGENIC LOCAL DOMINANT ORIENTATION ENERGY PATTERN

作　　者：王畅平 李昆明[1] 闫海停[1] 

机构地区：[1]湖南大学电气与信息工程学院 [2]解放军95333部队

出　　处：《计算机应用与软件》2016年第1期163-166,225,共5页Computer Applications and Software

摘　　要：针对人脸识别中不能有效利用图像局部能量信息的缺点,提出一种基于单演局部主方向能量二元模式的人脸识别方法。首先应用主元成分分析方法求取单演局部主方向能量,然后对单演局部主方向能量进行局部二元编码。在此基础上进一步分块,求取直方图,最后运用基于分块的线性判别方法进行降维,增强特征判别能力。实验结果表明,在AR和ORL人脸库上,单演局部主方向能量二元模式方法比单演幅值二元模式方法有更好的识别性能,这说明该方法能够有效提取图像的判别特征,提高系统的识别性能。Aiming at the shortcoming in face recognition that the local energy information of image cannot be effectively used, we proposed a face recognition algorithm which is based on the monogenic local dominant orientation energy binary pattern. First, it applies the principal component analysis （PCA） to extract the monogenic local dominant orientation energy, and then encodes the energy using binary strategy. Based on this, it further conducts blocking for obtaining the histogram, and finally uses the blocking-based linear discrimination method to reduce the dimensions and enhance the features discrimination ability. Experimental result showed that, on AR and ORL face databases, the recognition performance of monogenic local dominant orientation energy binary pattern method was better than that of monogenic amplitude binary pattern method, this indicated that the energy binary pattern could effectively extract the discriminative features and improve the performance of recognition.

关 键 词：单演信号 二元模式 PCA 主方向能量 人脸识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]