An approach for identification using ear and face biometrics employing score based image fusion and SIFT feature detection in surveillance system

Abstract: Biometric surveillance is any technology which measures and analyzes human physical and/or behavioural characteristics for authentication, identification, or screening purposes. Identification is an essential part of our lives. Identification of authentic candidate is essential in E-commerce, in keeping track of criminals, in airport and railway surveillance and many more aspects of the modern world. However identification of a person can be challenging especially when the person is not cooperating. This leads to classify the identification techniques broadly in to two categories: Passive and Active. In this current study an approach has been proposed combining the ear and face biometrics for the purpose of identification of a person using SIFT (Scale Invariant Feature Transform). The proposed multi biometric system achieves a recognition accuracy of 99.9958%.

Keywords: SIFT, biometric fusion, ear-face biometric.