基于二维局部保持鉴别分析的特征提取算法  

Feature Extraction Based on Two-dimensional Locality Preserving Discriminant Analysis

作　　者：卢官明[1,2] 左加阔[3] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]台湾大学电机资讯学院,台湾台北10617 [3]东南大学水声信号处理教育部重点实验室,江苏南京210096

出　　处：《南京邮电大学学报（自然科学版）》2014年第5期1-8,共8页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家自然科学基金(61071167);江苏省高校优秀中青年教师和校长境外研修计划资助项目

摘　　要：提出了一种二维局部保持鉴别分析(Two-dimensional Locality Preserving Discriminant Analysis,2D-LPDA)特征提取算法。该算法直接对图像矩阵进行运算而不需要将矩阵转化为向量后进行运算,较好地保持了图像相邻像素之间的空间结构关系;在LPP算法的基础上,利用训练样本的类别信息计算二维类间散度矩阵和二维类内散度矩阵,并在2D-LPDA的目标函数中引入最大间距准则(Maximum Margin Criterion,MMC),从而求得具有良好鉴别能力的投影向量,同时还避免了小样本情况下矩阵的奇异性问题。通过在ORL人脸图像库上的人脸识别和新生儿面部图像库上的疼痛表情识别实验,验证了所提出的算法的有效性。A feature extraction algorithm, called the two-dimensional locality preserving discriminant analysis（2D-LPDA） is proposed. The 2D-LPDA algorithm works directly with image matrix which does not need to be converted into vector, and can keep the spatial structure information of pixels on image. Based on locality preserving projection,2D-LPDA can calculate the between-class scatter matrix and the within- class scatter matrix by using the class label information of training samples, and introduce the maximum margin criterion（MMC） into the objective function, so as to obtain discriminative projection vectors and to avoid the singularity of the matrices in small size samples. Experimental results on face databases demonstrate the effectiveness of the proposed algorithm.

关 键 词：人脸识别 表情识别 特征提取 局部保持投影 二维局部保持鉴别分析 最大间距准则 

分 类 号：TN911.73[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]