Distinguishing a Person by Face and Iris Using Fusion Approach
Md. Zahidur Rahman, Md. Hasan Hafizur Rahman, Md. Mujibur Rahman Majumdar · 2019
Biometric system deals with recognizing a person according to his/her specific physiological and behavioral characteristics. Several physiological and behavioral characteristics have been used for biometric systems for example, face, fingerprints, iris, retina, etc. These characteristics can be utilized in a unimodal or a multimodal system for solid and secure identification of individuals. Unimodal biometric systems suffer from numerous difficulties, for example, noisy sensor data, non-universality, spoof attacks, etc. Multimodal biometric systems evolved to overcome these difficulties effectively by utilizing and fusing two or more individual modalities. Therefore, this paper presents a strategy to develop a multimodal biometric system by combining face and iris biometric system with the fusion technique to improve the identification efficiency. This paper also shows the comparison of performance between single biometric system (face, iris) and proposed multimodal biometric system. In this work, PCA method is utilized for face recognition process and Daugman method is utilized for iris recognition process; finally fusion method is utilized to fuse these two modalities. Compared with some state-of-the-art fusion methods, the proposed method combines two fusion techniques (matching score level fusion and decision level fusion) and shows a significant performance advantage.