Face recognition using a training sample from each individual

Abstract

In this article, a new method for face recognition is presented with a picture of each person. Face detection methods are typically several images to learn all there is to consider each individual, while in many real-world applications only one training images are available for everyone.In this article, to improve facial recognition method using RSC and private displays data classes did Categories and to identify with only one picture of each person, to hold high the rate can be used to identify an auxiliary dictionary contains features such as eigen_face that are derived from other people's pictures. Dictionary size depends on the number of variables, not the number of classes (people), the volume calculation will be significantly reduced. This is done by adjusting only one of the original algorithm parameters. We have to set the parameters of the PSO algorithm used. The proposed face detection system, basing on images AR achieved 97.47 but on YALE data collection 68.96 carefully achieved. As well as with other methods in the field of education compared with an image face detection and the best results compared to other methods are provided