Uniform Classifier for Biometric Ear and Retina Authentication Using Smartphone Application

Abstract

As biometric systems provide more reliable, efficient and secure means of identity verification, it gains its place as a most important system. Implementing ear biometrics using smartphone application development will justify the best suitable measure for authentication. The retina images are robust non-intrusive functional biometric. This paper proposes an authentication system which uses uniform feature extraction, Histogram of Gradient (HoG) and uniform classifier, Support Vector Machine (SVM) for two physiological biometrics. The best fitting kernel for SVM classification is achieved for radial basis function. The accuracy with uniform classier is 94% for ear biometrics and 93% considering retina images.