Performance Evaluation of Face Recognition Technique using Hartley and DCT with Fractional Feature Vector

Abstract

ABSTRACT This paper contributes to the performance evaluation of face recognition system that uses Hartley transform. Feature extraction is an important step in face recognition system. In transform based face recognition system, full and partial feature vectors are considered. Hartley transform is applied on resized image of size 128x128. Partial feature vector is selected by cropping the equal sized squares from four corners of transformed image. These cropped feature vectors are concatenated to form the resultant feature vector. Accuracy is calculated by varying the size of cropped feature vector at the corner of transformed image. Accuracy obtained by Hartley transform is compared with accuracy obtained by 2D-DCT. Both transforms give same accuracy and even computational complexity of both is same. General Terms Face recognition, Biometrics. Keywords Hartley Transform, Feature vector. 1. INTRODUCTION In Biometrics field, different techniques are rapidly developing to identify the human being. Face recognition has been actively researched in recent years. In the literatures, face recognition problem can be formulated as: given static (still) or video images of a scene, identify or verify one or more persons in the scene by comparing with faces stored in a database [1]. In general, a biometric identification system makes use of either physiological characteristics such as a fingerprint, iris pattern, or face or behavior patterns such as hand-writing, voice, or key-stroke pattern to identify a person. Face recognition is preferred over other methods because it does not require explicit cooperation from user and special costly equipments. A human face is complex object with features that can vary over the time due to aging. Other factors that influence face recognition include shape, pose, poor lighting, occlusion, sunglasses, long hair, or other objects partially covering the subject’s face and low resolution images [2]. A face recognition system is a computer application for automatically identifying or verifying a person from a digital image or a video frame from a video source. A face recognition system can operate in following two modes: