A Robust Multimodal Biometric System for Person Identification

Abstract

A more secured multimodal biometric Identification system can be developed using multiple biometric traits. We have used fingerprint and ear as biometric traits to provide a robust Identification System with reduced computational complexity and increased recognition rate. Here in this paper we have proposed and implemented a new enhanced edge intersection point detection method which is used to extract the required ear features .The fingerprint features are extracted using Gabor feature extraction. Extracted features are then fused by concatenation and stored in the database. For recognition, the fused test feature vectors are compared with stored data in the database using Euclidean distance. By this new enhanced method of feature extraction we were able to achieve a matching rate of 100% .Cumulative Match Characteristic (CMC) and Receiver Operating Characteristic (ROC) were plotted and it was observed that this enhanced method provides good results compared to other methods. We have also tested the classification of matching rate using confusion matrix which shows that our method produces perfect classification.