Evaluation of Three Variants of LBP for Finger Creases Classification
摘要
Biometric technology improves security and authentication, especially in sensitive systems like attendance systems. The common traits for biometrics are typically from the iris, fingerprints, face, etc. Another trait that possibly comes from the finger creases and the research work evaluating the finger crease’s capability for biometric classification is proposed in this paper. Various local binary patterns (LBP) are employed to extract the features, and the classification performance is evaluated using Support Vector Machines (SVM) on two different kernels. From the evaluation, an accuracy of up to 94% with a percentage of FAR and FRR less than 3 is observed for all the proposed LBP methods.