Optimal feature level fusion for secured human authentication in multimodal biometric system
- PDF / 1,597,086 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 92 Downloads / 184 Views
ORIGINAL PAPER
Optimal feature level fusion for secured human authentication in multimodal biometric system Himanshu Purohit1 · Pawan K. Ajmera1 Received: 25 July 2019 / Revised: 14 August 2020 / Accepted: 21 October 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2021
Abstract The rising demand for high security and reliable authentication schemes, led to the development of the unimodal biometric system so that the multimodal biometric system has emerged. The multimodal biometric system will use more than one biometric trait of an individual for identification and security purpose. Fusion plays a major role in the multimodal biometric system. Several fusion techniques are used in biometric systems. Feature level fusion is a very much popular method as compared to the other fusion techniques. In this fusion, features are extracted from all biometric traits. After that extracted features are combined into a final feature vector of high dimension. In this paper, we introduce a new technique to perform fusion at the feature level by optimal feature level fusion; here the relevant features are selected using an optimization technique. Here, we proposed OGWO for selecting optimal features. Moreover, we suggested the recognition technique. For recognition, we use the multi-kernel support vector machine algorithm. Finally, the performance of our proposed method is evaluated by some evaluation measures. Our recommended method is implemented in the MATLAB platform. Keywords Biometric · Fingerprint · Feature level fusion · Feature extraction · Optimization · Recognition · Multi support vector machine
1 Introduction Biometric technology is the one among technologies in the scientific area. Nowadays, biometric technologies are applied in various areas, from the work entrance organization to the person identification among the payment transactions [1]. Biometrics are an active area of research in pattern recognition and machine learning community [2]. It is an integral component of identity science, and biometric modalities such as the face, fingerprint, iris, and voice are being applied to recognize an individual [3]. It offers a very convenient and secure mode of identification and verification solutions. It is used in several applications like computer network login, electronic data security, e-commerce, Internet access, ATM, credit card, physical access control, cellular phone, PDA, medical records management, distance learning [4]. In this technology, biometric systems rely on particular data about unique biological traits to unique work effectively. There are
B 1
Himanshu Purohit [email protected] Department of Electrical and Electronics Engineering, BITS PILANI, Pilani, India
two types of biometric systems such as the unimodal biometric system and the multimodal biometric system. In this, unimodal biometric systems that use only one biometric trait for recognition often affect issues such as biometric data variation, lack of distinctiveness, low recognition accuracy, and spoof attacks. To d
Data Loading...
