Multimodal Biometric System using Grasshopper Optimization

Biometric systems are need of the day because of their various advantages over traditional authentication systems. Multimodal Biometric systems combine information from multiple sources to reach a final decision. Score level fusion combines outcomes of individual classffiers to make a final decision. However, most of the biometric systems suffer from the issue of score confliction of individual classifiers. To resolve this issue, we have proposed a novel optimized score level fusion using Grasshopper optimization where the performance optimization of individual classffiers is performed and a concurrent solution is achieved by means of proportional conflict redistribution rules. The system does not require any classifier training and exhibits high performance. The proposed system is robust against the dynamic environment and exhibits high reliability.