基于肤色增强和分块PCA的人脸表情识别方法  

A Facial Expression Recognition Method Based on Skin Color Enhancement and Block PCA

作　　者：谢非[1,2] 龚俊 王元祥[1,2] 吴茜[1,2] 杨建飞[1,2] 

机构地区：[1]南京师范大学江苏省三维打印装备与制造重点实验室,江苏南京210042 [2]南京师范大学电气与自动化工程学院,江苏南京210042

出　　处：《南京师范大学学报（工程技术版）》2017年第2期49-56,共8页Journal of Nanjing Normal University(Engineering and Technology Edition)

基　　金：国家自然科学基金(61601228;51407095);江苏省自然科学基金(BK20161021;BK20151548);江苏省高校自然科学基金(15KJB510016);江苏省三维打印装备与制造重点实验室项目(BM2013006)资助开放课题(3DL201607)

摘　　要：人脸的表情识别在智能人机交互应用中具有重要意义.本文提出了一种基于肤色增强和分块PCA的人脸检测及表情识别方法.首先,使用同态滤波增强肤色图像的亮度范围及对比度,利用YCb Cr色彩空间分量分离肤色背景区域,再通过轮廓分析确定人脸目标,最后对分割出的人脸进行均衡化处理,并引入分块主成分分析(PCA)算法进行表情识别.结果表明,该方法在光线较弱以及背景较复杂的情况下均能有效地进行人脸检测与表情识别,相对于传统的LBP方法可提高识别率约为2.3%.The facial expression recognition is of great significance in the application of intelligent man-machine interaction.This paper has proposed the face detection and expression recognition method based on the skin color enhancement and block PCA. Firstly, the skin color image luminance range is broadened and the contrast ratio is strengthened by the homo-morphic filtering method. Secondly,the skin color background region is separated through YCbCrThirdly, the face target is determined by the contour analysis. Finally, the equalization processing is made of the segmented face and the princijDal component analysis（ PCA） is imported to accomplisli the facial expressmental results show that in case of tlie weaker light and more complicated background, the face detection and facial expres-sion recognition botli can be achieved efectively through the metliod proposed in this paper,has improved 2.3% compared with the traditional LBP method.

关 键 词：人脸检测 表情识别 主成分分析 肤色模型 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]