A person identification system with biometrics using modified RBFN based multiple classifiers

Abstract

In this present paper, we have designed and developed a person identification system with biometrics using modified Radial Basis Function Network based multiple classifiers. Three different classifiers using the same Modified RBFN with Optimal Clustering Algorithm, separately identify fingerprint, iris and facial images and the individual conclusions are fused together with programming based boosting. The conclusions from individual classifiers as well as the super-classifier performing fusion of conclusions are fuzzy in nature. Holdout method with Fuzzy Confusion Matrix is used to compute different performance metrics like accuracy, precision, recall and F-score. The different performance metrics are quite satisfactory. Also the learning and performance evaluation time with Fuzzy Confusion Matrix is low and affordable.