Finger-Knuckle-Print Identification Based on Histogram of Oriented Gradients and SVM classifier

Recently, a wide variety of applications require reliable personal recognition systems to either confirm or determine the identity of an individual requesting their services. So, a reliable identity recognition system is a critical part in these applications that render their services only to genuine users. Thus, biometrics is an emerging technology that utilizes distinct behavioral or physiological traits in order to determine or verify the identity of an individual. In this context, the present paper attempts to design an effectively biometric system by using Finger-Knuckle-Print (FKP) traits. In this study, the feature vector of each segmented FKP is extracted using Histogram of Oriented Gradients (HOG). In addition, a multi-class Support Vector Machine (SVM) based learning algorithm is used to train the system using the extracted features vectors. From the test results, using PolyU FKP database with 165 persons, it is evident that our scheme has higher identification rate and very less classification error compared to several existing methods.