A Novel Multimodal Biometric Person Authentication System Based on ECG and Iris Data.

Area of Science:

• Biometrics and Security Engineering
• Signal Processing and Pattern Recognition
• Human-Computer Interaction

Background:

• Traditional security methods (keys, passwords) are prone to loss or forgetting.
• Biometric authentication offers enhanced security through unique physiological or behavioral traits.
• Multimodal biometric systems, combining multiple traits, improve accuracy and spoof resistance.

Purpose of the Study:

• To develop and evaluate a novel multimodal biometric person authentication system.
• To leverage the liveness detection capabilities of Electrocardiogram (ECG) and the distinctiveness of iris patterns.
• To enhance system robustness against spoofing attempts compared to unimodal systems.

Main Methods:

• A multimodal system integrating Electrocardiogram (ECG) and iris biometrics was designed.
• Key processing steps included data preprocessing, segmentation, feature extraction, and feature fusion.
• An ensemble classifier with majority voting was employed for final authentication decisions.

Main Results:

• The proposed multimodal system demonstrated strong performance across key metrics.
• Precision reached 96.55%, F1-score 96.2%, sensitivity 96.2%, specificity 96.5%, and overall accuracy 95.65%.
• The combined approach significantly improved authentication reliability and spoof resistance.

Conclusions:

• The multimodal biometric system integrating ECG and iris recognition offers a robust and accurate solution for person authentication.
• This approach effectively addresses the limitations of conventional security measures.
• The findings highlight the potential of combining physiological biometrics for advanced security applications.