Automated Authentication Using Information Fusion and Score Normalization in Multimodal Biometric Systems
Multimodal biometric system combines the evidence obtained from multiple modalities. By using an effective fusion scheme and normalization techniques the authors can significantly improve the over all accuracy and performance of biometric systems. In this paper, they have presented information fusion and score normalization approaches that performs better in identification process based on biological features. Unimodal biometric systems often affected by several problems like noisy sensor data, non-universality, lack of individuality, lack of invariant representations and spoofing traits. Multimodal biometric systems overcome these limitations by consolidating the evidence obtained from different sources.