Proposed Cancelable Face Recognition System Based on Histogram of Oriented Gradients

Abstract

nowadays have seen exponential growth in the usage of various biometric technologies in authenticated automatic recognition of humans. With the fast adaptation of biometric systems, there is a vital concern that biometric technologies may compromise the privacy and anonymity of individuals. To restrain the theft of biometric templates, it is desirable to alter them through noninvertible and revocable transformations to produce a cancelable biometric pattern. In this paper, we propose cancelable face pattern generation technique based on component Histogram of Oriented Gradients (HOG) and The Optical Double Random Phase Encoding (DRPE) algorithm. The proposed is evaluated by the security degree and receiver operating characteristic (ROC) Experiments carried out on diverse face databases assure the efficiency of the proposed approach. The proposed template succeeds to hide the details of the images and enhances ROC compared to the traditional standard.