Multimodal Biometric Authentication based on Score Level Fusion of Palmprint and Iris

Biometrics is a highly used technology which is earning research interest from past many years due to its applications in access control, identification, forensic investigation and surveillance. In this work, due to the inherent advantages of multimodal system over unimodal system, a multimodal system is proposed. Two or more biometric modalities when combined, offer more reliable identification and verification system. Here, palmprint and iris are used in identification due to its high robustness. Feature extraction methods used are average absolute deviation (AAD), histogram of gradients (HOG), and Gaussian membership feature (GMF) over both the modalities. K-nearest neighbor (KNN) classifier is used for matching with Euclidean distance. Finally, score level fusion is performed using frank T-norm, Schweizer-Sklar, Hamacher norm and Dubois-Prade fusion rules. The results show the robustness of the proposed fusion techniques over individual modality recognition rate.