A Fast and Accurate Face Authentication Method Using Hamming-Trace Transform Combination

AbstractFace authentication, also called face verification, is used to confirm the identity of a person by classifying the person as either the person he or she claims to be or not. For face authentication, the user generally provides a user name (and in some cases a password as well) to an authentication system that allows access to an area or system. Stored models associated with the provided username are used to verify the person’s identity and then allow or deny access. This paper proposes a fast and robust method for face authentication with variant illumination, scaling, rotation, blur, and difference in emotions (smiling, angry, and screaming). Techniques introduced in this work are composed of two stages. First, the feature of the face is to be detected by a combination of Trace transform and Fourier transform. Then, in the second stage, the Hamming distance is employed to measure and determine the similarity between models and test images. Finally, our method is evaluated with experiments on the ...