一种抗姿态与表情变化的三维人脸识别方法
|
摘要
为了提高人脸在姿态和表情变化下的识别率,结合局部平面距离(DLP)对曲面局部凹凸性优良的判断能力,提出了一种采用人脸的等距不变表示形式来匹配的人脸识别方法。首先,对深度摄像头采集到的深度图像进行距离约束、位置约束、转换等操作,得到干净完整的三维人脸,利用三维人脸上每一点DLP值确定鼻尖点,利用聚类的思想确定鼻根点;其次,采用改进的快速推进算法计算人脸的测地距矩阵,设置阈值并切割出有效的人脸区域;最后,计算有效的人脸区域的高阶矩特征,作为人脸的特征向量进行匹配。实验结果表明,对于不同的数据库,本文算法的识别率接近97%;将本文算法与基于轮廓线特征的人脸识别算法以及基于Gabor特征的人脸识别算法进行比较,其识别率分别提高了14.1%和8.3%,同时有着较高的运算效率。
In order to improve the facial recognition rate under the change of posture and expression,combined with the ability of local plane distance(DLP)to judge the convexity of local curved surface,a face recognition method based on face equidistant invariant representation was proposed.Firstly,several operations such as distance constraint,location constraint and transformation were conducted on the depth image captured by the deep camera to get the clean and complete 3D face;then the nose tip was determined by the DLP value of every point on the 3 D face,and the nasal root was determined by the clustering idea;secondly,the improved fast propulsion algorithm was used to calculate the geodesic distance matrix of face,then the threshold value was set and the effective face area was cut out;finally,the high-order moment feature of the effective face area was calculated as the feature vector of face for matching.The experimental results show that the recognition rate of this algorithm is close to 97%for different databases.Compared with the face recognition algorithms based on contour features and Gabor features,the recognition rate of this algorithm is increased by 14.1% and 8.3%,respectively,while having a high computing efficiency.
In order to improve the facial recognition rate under the change of posture and expression,combined with the ability of local plane distance(DLP)to judge the convexity of local curved surface,a face recognition method based on face equidistant invariant representation was proposed.Firstly,several operations such as distance constraint,location constraint and transformation were conducted on the depth image captured by the deep camera to get the clean and complete 3D face;then the nose tip was determined by the DLP value of every point on the 3 D face,and the nasal root was determined by the clustering idea;secondly,the improved fast propulsion algorithm was used to calculate the geodesic distance matrix of face,then the threshold value was set and the effective face area was cut out;finally,the high-order moment feature of the effective face area was calculated as the feature vector of face for matching.The experimental results show that the recognition rate of this algorithm is close to 97%for different databases.Compared with the face recognition algorithms based on contour features and Gabor features,the recognition rate of this algorithm is increased by 14.1% and 8.3%,respectively,while having a high computing efficiency.
引文
[1]ZOU Guofeng,FU Guixia,LI Haitao,et al.A survey of multi-pose face recognition[J].Pattern Recognition&Artificial Intelligence,2015,28(7):613-625.邹国锋,傅桂霞,李海涛,等.多姿态人脸识别综述[J].模式识别与人工智能,2015,28(7):613-625.
[2]HU Min,CHENG Tianmei,WANG Xiaohua.Face recognition based on fusion of global and local feature[J].Modern Computer,2013,27(9):817-822.胡敏,程天梅,王晓华.融合全局和局部特征的人脸识别[J].电子测量与仪器学报,2013,27(9):817-822.
[3]ZHU Binglian,WANG Dawei,LI Mingjun,et al.Study on 3Dface recognition system against expression change[J].Chinese Journal of Scientific Instrument,2014,35(2):299-304.朱冰莲,王大为,李明均,等.针对表情变化的三维人脸识别系统研究[J].仪器仪表学报,2014,35(2):299-304.
[4]LI Xiaoli,DA Feipeng.3Dface recognition based on profile and rigid regions[J].Journal of Image&Graphics,2010,15(2):266-273.李晓莉,达飞鹏.基于侧面轮廓线和刚性区域的3维人脸识别[J].中国图象图形学报,2010,15(2):266-273..
[5]WANG X,FANG Z,WU J,et al.Discrete geodesic graph(DGG)for computing geodesic distances on polyhedral surfaces[J].Computer Aided Geometric Design,2017,52(C):262-284.
[6]XU C,TAN T,WANG Y,et al.Combining local features for robust nose location in 3Dfacial data[J].Pattern Recognition Letters,2006,27(13):1487-1494.
[7]ZHOU H,YUAN Y,SADKA A H.Application of semantic features in face recognition[J].Pattern Recognition,2008,41(10):3251-3256.
[8]LEE K C,HO J,KRIEGMAN D J.Acquiring linear subspaces for face recognition under variable lighting.[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2005,27(5):684.
[9]LU J,YUAN X,YAHAGI T.A method of face recognition based on fuzzy c-means clustering and associated sub-NNs.[J].IEEE Transactions on Neural Networks,2007,18(1):150-60.
[10]ZHU Lei,ZHU Shan’an.Face recognition based on two-dimensional image principal component analysis[J].Journal of Zhejiang University,2007,41(2):264-267.祝磊,朱善安.基于二维广义主成分分析的人脸识别[J].浙江大学学报,2007,41(2):264-267.
[11]DONG Xiaoqing,CHEN Hongcai.Face recognition based on sub-pattern two-directional two-dimensional linear discriminan t analysis[J].Chinese Journal of Liquid Crystals and Displays,2015,30(6):1016-1023.董晓庆,陈洪财.基于子模式双向二维线性判别分析的人脸识别[J].液晶与显示,2015,30(6):1016-1023.
[12]CHAI Zhi,LIU Zhengguang.Face recognition using complex wavelet and independent component analysis[J].Journal of Computer Applications,2010,30(7):1863-1866.柴智,刘正光.应用复小波和独立成分分析的人脸识别[J].计算机应用,2010,30(7):1863-1866.
[13]PERAKIS P,THEOHARIS T,PASSALIS G,et al.Automatic 3Dfacial region retrieval from multipose facial datasets[C]//Eurographics Conference on 3dObject Retrieval.[S.l]:[s.n.],2009:37-44.
[14]LU Huiming.3Dface recognition based on differential geometric method[D].Nanjing:Nanjing University of Posts and Telecommunications,2009.陆慧铭.基于微分几何方法的3D人脸识别[D].南京:南京邮电大学,2009.
[15]WEBER O,DEVIR Y S,BRONSTEIN A M,et al.Parallel algorithms for approximation of distance maps on parametric surfaces[J].Acm Transactions on Graphics,2008,27(4):1-16..
[16]LI Xiaojuan,YANG Tangwen,RUAN Qiuqi,et al.Modified 3Dface recognition algorithm based on contour curve features[J].Journal of the China Railway Society,2012,34(1):34-38.李晓娟,杨唐文,阮秋琦,等.基于轮廓线特征的三维人脸识别改进算法[J].铁道学报,2012,34(1):34-38.
[17]XU C,LI S,TAN T,et al.Automatic 3Dface recognition from depth and intensity Gabor features[J].Pattern Recognition,2009,42(9):1895-1905.
[2]HU Min,CHENG Tianmei,WANG Xiaohua.Face recognition based on fusion of global and local feature[J].Modern Computer,2013,27(9):817-822.胡敏,程天梅,王晓华.融合全局和局部特征的人脸识别[J].电子测量与仪器学报,2013,27(9):817-822.
[3]ZHU Binglian,WANG Dawei,LI Mingjun,et al.Study on 3Dface recognition system against expression change[J].Chinese Journal of Scientific Instrument,2014,35(2):299-304.朱冰莲,王大为,李明均,等.针对表情变化的三维人脸识别系统研究[J].仪器仪表学报,2014,35(2):299-304.
[4]LI Xiaoli,DA Feipeng.3Dface recognition based on profile and rigid regions[J].Journal of Image&Graphics,2010,15(2):266-273.李晓莉,达飞鹏.基于侧面轮廓线和刚性区域的3维人脸识别[J].中国图象图形学报,2010,15(2):266-273..
[5]WANG X,FANG Z,WU J,et al.Discrete geodesic graph(DGG)for computing geodesic distances on polyhedral surfaces[J].Computer Aided Geometric Design,2017,52(C):262-284.
[6]XU C,TAN T,WANG Y,et al.Combining local features for robust nose location in 3Dfacial data[J].Pattern Recognition Letters,2006,27(13):1487-1494.
[7]ZHOU H,YUAN Y,SADKA A H.Application of semantic features in face recognition[J].Pattern Recognition,2008,41(10):3251-3256.
[8]LEE K C,HO J,KRIEGMAN D J.Acquiring linear subspaces for face recognition under variable lighting.[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2005,27(5):684.
[9]LU J,YUAN X,YAHAGI T.A method of face recognition based on fuzzy c-means clustering and associated sub-NNs.[J].IEEE Transactions on Neural Networks,2007,18(1):150-60.
[10]ZHU Lei,ZHU Shan’an.Face recognition based on two-dimensional image principal component analysis[J].Journal of Zhejiang University,2007,41(2):264-267.祝磊,朱善安.基于二维广义主成分分析的人脸识别[J].浙江大学学报,2007,41(2):264-267.
[11]DONG Xiaoqing,CHEN Hongcai.Face recognition based on sub-pattern two-directional two-dimensional linear discriminan t analysis[J].Chinese Journal of Liquid Crystals and Displays,2015,30(6):1016-1023.董晓庆,陈洪财.基于子模式双向二维线性判别分析的人脸识别[J].液晶与显示,2015,30(6):1016-1023.
[12]CHAI Zhi,LIU Zhengguang.Face recognition using complex wavelet and independent component analysis[J].Journal of Computer Applications,2010,30(7):1863-1866.柴智,刘正光.应用复小波和独立成分分析的人脸识别[J].计算机应用,2010,30(7):1863-1866.
[13]PERAKIS P,THEOHARIS T,PASSALIS G,et al.Automatic 3Dfacial region retrieval from multipose facial datasets[C]//Eurographics Conference on 3dObject Retrieval.[S.l]:[s.n.],2009:37-44.
[14]LU Huiming.3Dface recognition based on differential geometric method[D].Nanjing:Nanjing University of Posts and Telecommunications,2009.陆慧铭.基于微分几何方法的3D人脸识别[D].南京:南京邮电大学,2009.
[15]WEBER O,DEVIR Y S,BRONSTEIN A M,et al.Parallel algorithms for approximation of distance maps on parametric surfaces[J].Acm Transactions on Graphics,2008,27(4):1-16..
[16]LI Xiaojuan,YANG Tangwen,RUAN Qiuqi,et al.Modified 3Dface recognition algorithm based on contour curve features[J].Journal of the China Railway Society,2012,34(1):34-38.李晓娟,杨唐文,阮秋琦,等.基于轮廓线特征的三维人脸识别改进算法[J].铁道学报,2012,34(1):34-38.
[17]XU C,LI S,TAN T,et al.Automatic 3Dface recognition from depth and intensity Gabor features[J].Pattern Recognition,2009,42(9):1895-1905.