Fuzzy vault template protection for multimodal biometric system

Abstract

Biometric security systems are gaining strength these days as compared to conventional authentication methods. But with the advancements in biometric systems, the attempts of attacks on these systems are also increasing day by day. Even though it is difficult for an intruder to steal the biometric trait of a person, he can still breach into the security of a biometric system. Biometric System can be implemented as unimodal and multimodal systems depending upon the application requirements. Multimodal biometrics systems however are more reliable and stable as compared to unimodal systems. The choice of fusion strategy while designing a multimodal biometric system has a direct impact on its performance. Out of the various fusion strategies, feature level fusion is considered more robust. The extracted feature set from the multiple biometric traits used for multimodal system design may be incompatible. Incompatible biometric feature sets have to be made compatible by using suitable conversion. The paper proposes a Multimodal Biometric systems using face and fingerprint traits with Fuzzy Vault template security.