Face Recognition Abstract

Abstract— Face recognition is a biometric system used to verify or identify a person from a digital image. This involve extracting features of face and then recognize it, regardless of expression, lighting, illumination, transformations (translate, rotate and scale image), ageing and pose. The existing approaches are deblurring based, joint deblurring and recognition, deriving blur invariant features and direct approach, which uses convolution model for performing face recognition in the presence of blur. So these methods cannot handle non uniform blurring situations that frequently arise from rotations and tilts in hand held cameras. In this paper, face recognition is done, in the presence of space varying motion blur. We have taken the concept

I. INTRODUCTION

Face recognition is a biometric approach that employs automated methods to recognize or verify the identity of a living person based on his/her physiological characteristics. In general, a biometric identification system makes use of either behavior patterns (such as voice, hand-writing, or key-stroke pattern) or physiological characteristics (such as an iris pattern, fingerprint, or face) to identify a person. For all other biometric system, we need the active participation of subject. But in the case of face recognition system, we require neither the knowledge nor the cooperation of the subject.

Face recognition is a type of pattern recognition of a person's face which is recorded in a database for future attempts to recognize a person's identity. Face recognition are used in wide range of applications like security control and video retrieving. Automatic face recognition involves three key steps. Face detection, feature extraction and face

Currently existing face recognition systems are analyzed in Section II together with their design challenges. The proposed solution is presented in Section III. Simulation results are presented and discussed in Section IV. Finally, the conclusions of the work are summarized in Section V.

II. LITERATURE REVIEW

The images are classified as unconstrained domain and constrained domain images. There are still some challenging problems to address in face recognition under uncontrolled conditions. The main factors that affect face recognition systems are degradations of image due to blur, variations in appearance and noise due to illumination and pose.

In existing systems, blurring due to camera shake has been modeled as a convolution model. In convolution model a single blur kernel is present. And the blur is assumed to be uniform across the image [2]. However, in hand-held cameras space-variant blur is encountered frequently [11], which cannot be explained using convolution