Comparative Analysis of Three Single Trait Biometric Authentication Models

In traditional authentication model, user identification is based on Password, PIN or Signature, but sometimes these can be lost, stolen, or subjected to spoofing attack. In biometric authentication system, a user is identified through physical or behavioural features. These features include fingerprint, palmprint, face, iris, signature, speech, and so on. The main challenging task in biometric is about storage problem, and transmission of huge data. By replacing PINs, use of biometric traits can certainly deny unauthorized access or fraudulent use of ATMs, mobile phones, smart cards, data centres, personal computers, and computer networks. The performance can be measured by evaluating performance metrics i.e., FAR, FRR, FTC, and EER. This paper proposes three different uni-modal biometric authentication systems using fingerprint, face, and voice signal as sources. A new technique is designed and applied to each biometric authentication system.