Generating Cryptosystem for Multimodal Biometrics Based on Feature Level Fusion
Multi-biometrics are used worldwide because there will be low error rate and large population coverage. Multi-biometric data are stored in template and these template needs to be protected. There are many attacks possible in these templates. So here, the authors transform the different biometric representation (iris, fingerprint and face) into common representation and then security is made by inserting key with these templates. In this paper, they propose a feature level fusion framework to protect multiple templates as a single secure sketch. Their implementations include fuzzy vault and fuzzy commitment and security measures of multimodal biometrics. Their experimental result show that multi-biometrics have advantage than uni-biometric system.