Multimodal Biometric Using Fusion of Fingerprint, Finger Knuckle Print and Palm Print
Shubhangi Neware · Bioscience Biotechnology Research Communications · 2020
Human recognition based on combination of various biometric traits is useful for considerably improving the performance accuracy.This paper is presenting different hand biometric traits used to identify a person.The objective of this paper is to propose multimodal biometric system using multiple biometric traits such as finger print, finger knuckle print (FKP) and palm print.The Convolution Neural Network and Softmax classifier is proposed to generate score of each modality separately and then fusion at decision level can be done.If final score is less than threshold value then person identity will be rejected.Fusion of different modalities at decision level will improve the performance of person identification system.Performance analysis is given for decision level fusion used with different modalities.