一种新的多样本扩充协同表示分类人脸识别算法  

Study on Collaborative Representation Classification Method Based on Multiple Images for Face Recognition

作　　者：周先春[1,2,3] 许瑞 石兰芳[4] ZHOU Xian-chun;XU Rui;SHI Lan-fang(School of Electronic and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China;Ministry of Education Key Laboratory of Child Development and Learning Science,Nanjing 210044 ,China;Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing 210044, China;Department of College of Mathematics and Statistics, Nanjing University of Information Science ＆ Technology, Nanjing 210044, China)

机构地区：[1]南京信息工程大学电子与信息工程学院,南京210044 [2]儿童发展与学习科学教育部重点实验室,南京210096 [3]南京信息工程大学江苏省大气环境与装备技术协同创新中心,南京210044 [4]南京信息工程大学数学与统计学院,南京210044

出　　处：《小型微型计算机系统》2018年第7期1616-1620,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(11202106)资助;东南大学基本科研业务费项目(CDLS-2016-03)资助

摘　　要：由于人脸库中人脸样本的数目有限,远远不能够满足实际生活中人脸识别的需要.新算法提出一种新的多样本扩充的协同表示分类人脸识别算法,首先利用人脸的镜面性生成镜面图像,分别对同一类中任意的两个原始样本与镜像样本,取它们的平滑中值样本构造新的虚拟样本,然后用欧式距离选择出接近测试样本的训练样本.以往的多样本扩充的人脸算法,是将新生成的虚拟训练样本与原始训练样本结合在一起,作为总的训练样本进行人脸识别,新算法是将不同途径构成的训练样本分别进行参数加权融合,采用基于协同表示的分类算法进行人脸识别.实验结果表明,新算法能够在ORL和FERET人脸数据库上获得更好的人脸识别率,具有较好的人脸描述能力.As the number of face samples is limited,it can not meet the need of face recognition in real world application. We propose a new collaborative representation classification method based on multiple images for face recognition,the method firstly use the mirror feature of the face image to generate new samples,and then use the arbitrary two new samples and original samples respectively to synthesize the smoothing median virtual samples,we select useful training samples that are similar to the test samples from the set of all the original and synthesized virtual training samples based on the Euclidean distance. Previous methods such as multiple images for face recognition,based on the combination of original training samples and virtual training samples to perform face recognition. We design a parameter weighted fusion for training samples composed of different approaches respectively,finally,we use the collaborative representation to perform face recognition. Experimental results on ORL and FERET face databases show that the proposed algorithm can be applied to achieve high face recognition rates,it has strong ability of face description.

关 键 词：人脸识别 协同表示 平滑中值样本 加权融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]