By Topic

Fast face recognition approach using a graphical processing unit “GPU”

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Yousri Ouerhani ; Département d'optoélectronique, Laboratoire L@bISEN, 20 rue Cuirassé Bretagne CS 42807, 29228 Brest Cedex 2, France ; Maher Jridi ; Ayman AlFalou

In this manuscript, we present an implementation of a correlation method for face recognition application on GPU. Our correlator is based on the famous “4f” setup and the use of a Phase Only Filter (POF). Traditionally, the correlation method is implemented using optical components for real-time application. Unfortunately, optical implementation is complex and has exorbitant price. To cope with these drawbacks and in order to benefit from the accuracy of the correlation method, we propose in this work to implement the correlation using GPU. To this end, we will take an interest in the mathematical aspect of the correlation method to identify the processing to be implemented on GPU. Simulations results about the implementation of the face recognition application on GPU showed the efficiency of our proposed design. Moreover, comparison between GPU and CPU in terms of execution time have been made and shows that, to identify one face among 4, GPU Nvidia Geforce 8400 GS is 3 times faster than the Intel Core 2 CPU 2.00 GHZ (using Matlab).

Published in:

2010 IEEE International Conference on Imaging Systems and Techniques

Date of Conference:

1-2 July 2010