Face recognition systems are highly required for variety of applications like user authentication, advanced video surveillance, biometrics etc. Majority of existing systems worked on higher dimensional spaces whereas a human face image (somewhat similar shapes and placement of objects) can be projected on a lower dimensional subspace. This dimensionality reduction is possible by using “Principle Component Analysis” method. PCA approach gives eigenvectors having eigenvalues for a given set of images which represents variations amongst images, creates aneigenspace. Input image can be identified as a “face” image or a “non-face” image by projecting it on eigenspace and measuring the distance of mean adjusted test image from the face space. If the test image is identified as face image then it can be checked as “known” or “unknown” face image by measuring the distance of its projection on face space with the nearest neighbor face class. So to recognize a input face, two thresholds are required, one for identifying a image as “face” image or a “non-face” image and another for classifying a face image as “known” or “unknown”. This paper also suggests a method to calculate both the thresholds from the set of image database of known images itself. This method can be used to recognize faces in spite of considerable variations in pose, expression, scale and disguise since the recognition is not based on features but it is based on variations amongst face images. The most important advantage of this method is its lower computational complexity because of working on a lower dimensional workspace.

1. W.W.Bledsoe, 1996 “The model method in facial recognition,” Panaromic Research Inc., Palo Alto, CA August,Rep. PR1-15.
2. M.Turk and A. Pentland, 1991 “Eigen facesfor Recognition,” J.Cogn.Neurosci, vol. 3, no. 1, pp. 71-86
3. L. Sirovich and M. Kirby, 1987 “Low Dimensional Procedure for the Characterization of Human faces,” Journal of Optical Society of America, 4(3), 519-524.
4. L. D. Harmon, 1971 “Some Aspects of Recognition of Human Faces,” O. J. Grusser & R. Klinke (Eds.), Pattern Recognition in Biological and Technical Systems, Berlin:Springer-Verlag.
5. Y.Kaya andK.Kobayashi, 1972 “A Basic study on Human Face Recognition,” S. Watanabe (Ed.), Frontiers of Pattern Recognition, New York.
6. M. Kirby and L. Sirovich, 1990 “Application of Karhunen-Loeve Procedure for the Characterization of Human faces,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(1).
7. M. Turk and A. Pentland, 1991 “Face recognition using eigenfaces”, Proc. IEEE Conf. on Computer Vision and Pattern Recognition, pages 586-591.
8. Tal Hassner Lihi Zelnik-Manor, “Principal Component Analysis, Eigenfaces and Fisher discriminant”.

Eigenspace Eigenfaces Eigenvectors Facespace PrincipalComponentAnalysis (PCA) Pattern Recognition