Multimodal fusion of the finger vein, fingerprint and the finger-knuckle-print using Kernel Fisher analysis

Unimodal biometric have improved the possibility to establish systems capable of identifying and managing the flow of individuals according to the available intrinsic characteristics that we have. However, a reliable recognition system requires multiple resources. This is the main objective of the multimodal systems that consists of using different resources. Although multimodality improves the accuracy of the systems, it occupies a large memory space and consumes more execution time considering the collected information from different resources. Therefore we have considered the feature selection, that is, the selection of the best attributes that enhances the accuracy and reduce the memory space as a solution. As a result, acceptable recognition performances with less forge and steal can be guaranteed. In this paper we propose an identification system using multimodal fusion of finger-knuckle-print, fingerprint and finger's venous network by adopting several techniques in different levels for multimodal fusion. A feature level fusion and decision level is proposed for the fusion of these three biological traits. An optimization method for this multimodal fusion system by enhancing the feature level fusion is introduced. The optimization consists of the space reduction by using different methods.