An overview of hand-based multimodal biometrie system using multi-classifier score fusion with score normalization

摘要

In the emerging trends of biometrie authentication, multimodal biometries getting more attention from researchers due to its universality, uniqueness, no intra-class variations, no inter-class similarities, and anti-spoofing attacks than unimodal biometrics. Hand-based multibiometric system is the most successful and used in many real time systems especially law enforcement and forensics. Moreover, it is very user friendly and ease of use among all other biometric traits. Multi classifier fusion is the use of more classifiers for each modality involved rather than single at the score fusion of multibiometric system. Furthermore, hand-based multimodal biometrics can use either or all traits of fingerprint, palm print, finger vein, palm vein, dorsal vein, hand geometry, finger knuckle print and many more. In this paper, we reviewed various score normalization techniques and multi-classifiers used in the hand-based multimodal biometrics for each modality involved at the matching score fusion for enhancing the system performance further.