A Case Study of Data Fusion for Biometric Applications

Abstract

The paper presents a case study for using the data fusion in biometric security applications. A hierarchical data fusion is evaluated on the available data. The fusion rules do not depend on the feature extraction algorithm as in other application cases. The experimental results show the potential of the pre-classification data fusion for improving the accuracy in applications belonging to biometric and bio-medical domains. Further research efforts are still required in order to design proper fusion models that should be adapted to the particular application requirements and constraints.