基于局部二元模式直方图的红外人脸识别  

Infrared Face Recognition Based on Local Two Element Pattern Histogram

作　　者：于妍[1] 韩冬[2] 

机构地区：[1]许昌学院,河南许昌461000 [2]许继集团有限公司,河南许昌461000

出　　处：《激光杂志》2016年第5期40-43,共4页Laser Journal

基　　金：河南省教育厅人文社会科学研究项目(2013-GH-049)

摘　　要：红外人脸对光照、姿态等变化不敏感,较好的克服了可见光人脸的缺陷,为了提高红外人脸识别率,提出一种基于局部二元模式直方图的红外人脸识别方法。首先采集大量的红外人脸图像,并将它们归一化标准红外人脸图像,然后提取红外人脸图像的局部二元模式直方图作为识别特征,并采用核主成分分析选择对红外人脸图像有贡献的特征,最后采用相关向量机对进行红外人脸图像的分类与识别。实验结果表明,本文红外人脸识别方法的识别率要高于其它红外人脸识别方法,而且红外人脸的识别速度更快,可以应用于实时红外人脸识别中。Infrared face is not sensitive to illumination and poses variation which can overcome visible light face defects,in order to improve the recognition rate of infrared face,so this paper put forward a infrared face recognition method based on local binary pattern histogram.First of all,a lot of infrared face images are collected and are normalized as standard infrared face images,and then local binary pattern histograms of infrared face images are extracted as recognition features,and kernel principal component analysis is used to choose contribution features for infrared face,finally using relevance vector machine is used to establish the classifier for infrared face image.The experimental results show that the recognition rate of the proposed method is higher than other infrared face recognition methods,and the infrared face recognition speed is faster,it can be applied to real-time infrared face recognition.

关 键 词：红外技术 局部二元模式 人脸识别 直方图特征 

分 类 号：TN74[电子电信—电路与系统]