A Multimodal Biometric Authentication System Using Ear and Face

The user confirmation systems witch using a biometric method are mostly encounter noisy data and infinite orders of error.To improve special matching in such situations, hybrid biometric systems are utilized.In this study we combined two biometric systems of ears and face to propose a multimodal biometric authentication system which is secure and reliable.To perform desired experiments three face databases including Cohen-Canad, Indian Institute of Technology and FEI new database are considered.We combined these databases separately with ear images database from PNU.Then, we tried to recognize people and evaluate the proposed system.Experimental results revealed that proposed hybrid biometric system is more reliable and more precise comparing to systems using a single biometric method.Owing to its fast recognition it is absolutely suitable for real-time systems.