Performance of Multimodal Biometric System Based on Level and Method of Fusion

Abstract

User authentication is essential to provide security that restricts access to system and data resources. Automated recognition of individuals based on their biological and behavioral characteristics is referred as biometric system. Recognition of legitimate user depends upon a feature vector(s) extracted from either their distinguishing behavioral or both distinguishing behavioral and physiological traits such as face, finger, speech, iris and gait. Research on biometrics has distinctly increased for solving identification and authentication issues in forensics, physical and computer security, custom and immigration. However, unimodal biometric system is not able to fulfill reliability constraints, speed and acceptability of authentication in real applications due to noise in sensed data, spoof attacks, data quality, lack of distinctiveness, restricted degree of freedom, non-universality and other factors. Therefore, multimodal biometric systems are used to increase security as well as better performance. To establish the identity of individuals, multimodal biometric systems unite the information presented by multiple biometric sensor, samples, algorithms, units and traits. Multimodal biometric systems are not only used for enhancing matching performance, but these systems also provide improved people coverage, discourage deceiving, make possible continuous monitoring and impart fault tolerance to biometric applications. This paper presents an overview of different multimodal biometric (multibiometric) systems and their fusion techniques with respect to their performance.