Computer Engineering and Applications ›› 2008, Vol. 44 ›› Issue (10): 166-168.

• 图形、图像处理 • Previous Articles     Next Articles

Face recognition method handling for pose and illumination variations

LU Chun-mei,NIU Hai-jun,HAO Lin-bo   

  1. School of Computer Science and Technology,Xidian University,Xi’an 710071,China
  • Received:2007-07-20 Revised:2007-10-18 Online:2008-04-01 Published:2008-04-01
  • Contact: LU Chun-mei

消除光照和姿态变化影响的人脸识别

卢春梅,牛海军,郝琳波   

  1. 西安电子科技大学 计算机学院,西安 710071
  • 通讯作者: 卢春梅

Abstract: The influences of pose and illumination variations are two major bottlenecks toward full automatic face recognition.A novel method which handles both pose and lighting conditions is proposed here via this paper.First,the intensity of images in training set is normalized in order to reduce the susceptivity toward lighting intensity.Afterwards,the images’ pose is estimated,and eigenvector subspace of different pose is built up by means of eigenface.Ultimately,the conception called PWV(Pose’s Weight Value) is proposed for a new classification WMDC(Weighted Minimum Distance Classifier),which is designed according to normalized Euclidian distance.Due to this approach,the problem of variant pose is solved by assigning different pose’s weight value.Experimental results on FERET and Yale B database demonstrate that this technique can significantly improve the accuracy of face recognition under variant illumination and pose conditions.

Key words: face recognition, intensity normalization, eigenface, pose estimation, weighted minimum distance classifier

摘要: 光照和姿态变化带来的影响是自动人脸识别的两个主要瓶颈问题。提出了消除这两方面影响的处理方法:首先对训练集里的图像应用灰度归一化处理,降低对光照强度的敏感度;然后进行姿态估计,并用特征脸方法计算不同姿态的特征子空间,最后提出了“姿态权重PWV(Pose’s Weight Value)”这一概念,据此设计了加权的最小距离分类器WMDC(Weighted Minimum Distance Classifier),分配不同姿态权重消除姿态变化影响。在FERET和Yale B数据库上的实验结果表明,此方法能在很大程度上提高人脸光照和姿态改变时的识别率。

关键词: 人脸识别, 灰度归一化, 特征脸, 姿态估计, 加权最小距离分类器