Biometric fusion: Combining multimodal and multi algorithmic approach
V A Thasiyabi, Rani Koshy, Smitha Satheesh · 2016
Biometrics, due to their uniqueness and accuracy are being widely used for authentication and identification. In multimodal biometrics more than one trait of an individual is fused. Multimodal biometrics is of different types, multi algorithmic, multi instance and multi sensorial. This paper combines two biometric traits namely face and fingerprint. Face is being processed using two algorithms, PCA and modular kernel PCA, to reduce the recognition error. After being processed by PCA, output of PCA is provided to modular kernel PCA. Then fingerprint is processed using crossing number based method. Finally output from modular kernel PCA andfingerprint are combined based on sum rule.