Is There Still Confidence In Hand-Crafted Feature Extraction Techniques To Use Them In Biometric Systems?

Abstract

At present, the Information security against attacks is considered as a key point in the success of most institutions, thus the security of persons' identity is therefore primary to these successes. For that, biometrics techniques were adopted for reinforcing the security of information systems by considering their effectiveness compared to traditional techniques. On this basis, this paper proposed an effective multimodal biometric system utilizing two hand biometrics (palmprint and palm-vein). In this work, two LBP based texture analysis variants which are Three-Patch based LBPs (TP-LBP) and Four-Patch based LBPs (FP-LBP) are used. The presented work of the security analysis can aid in grasping what a biometric system is, therefore getting an effective solution to safeguard information systems. The experimental results determined that the proposed system utilizing a multibiometric dataset of 300 persons showed a high identification rate in addition to a considerable low classification error compared to the existing techniques.