Face recognition algorithm in hyperspectral imagery by employing the K-means method and the Mahalanobis distance

M. I. Elbakary; M. S. Alam; M. S. Aslan

doi:10.1117/12.737191

21 September 2007 Face recognition algorithm in hyperspectral imagery by employing the K-means method and the Mahalanobis distance

M. I. Elbakary, M. S. Alam, M. S. Aslan

Author Affiliations +

Proceedings Volume 6697, Advanced Signal Processing Algorithms, Architectures, and Implementations XVII; 669705 (2007) https://doi.org/10.1117/12.737191
Event: Optical Engineering + Applications, 2007, San Diego, California, United States

Abstract

Recently, spectral information is introduced into face recognition applications to improve the detection performance for different conditions. Besides the changes in scale, orientation, and rotation of facial images, expression, occlusion and lighting conditions change the overall appearance of faces and recognition results. To eliminate these difficulties, we introduced a new face recognition technique by using the spectral signature of facial tissues. Unlike alternate algorithms, the proposed algorithm classifies the hyperspectral imagery corresponding to each face into clusters to automatically recognize the desired face and to eliminate the user intervention in the data set. The K-means clustering algorithm is employed to accomplish the clustering and then Mahalanobis distance is computed between the clusters to identify the closest cluster in the data with respect to the reference cluster. By identifying a cluster in the data, the face that contains that cluster is identified by the proposed algorithm. Test results using real life hyperspectral imagery shows the effectiveness of the proposed algorithm.

Citation Download Citation

M. I. Elbakary, M. S. Alam, and M. S. Aslan "Face recognition algorithm in hyperspectral imagery by employing the K-means method and the Mahalanobis distance", Proc. SPIE 6697, Advanced Signal Processing Algorithms, Architectures, and Implementations XVII, 669705 (21 September 2007); https://doi.org/10.1117/12.737191

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available