On the Use of Finger Knuckle Patterns For Person Identification

An accurate hand biometric modality for person authentication is studied within the scope of this research. Patterns are being extracted from minor and major finger knuckle patterns in this work utilizing texture-based feature extraction methods. These patterns are being used for biometric systems. The use of images of the finger knuckles in biometric systems is still relatively new, but it has lately garnered a significant amount of interest in the research literature. Finger knuckle biometrics has gained an increasing amount of interest in the biometrics literature, and a variety of matching algorithms have been researched in order to increase the accuracy of the matching. Finger knuckle patterns have received a lot of attention in the biometrics world during the past decade's worth of research. The finger dorsal skin patterns created between the metacarpal and the proximal phalanx bones of fingers have been researched in the scientific literature for their potential use as a biometric feature. These patterns are formed as the metacarpal bone moves closer to the proximal phalanx bone. On a few finger knuckle databases, research is done to determine whether or not these patterns are unique. This thesis focuses on finger knuckle patterns, namely minor and major variations in finger knuckle patterns, as well as texture-based feature extraction approaches. In order to come up with a reliable way for identifying individuals, many different texture-based approaches for finger knuckle biometrics are investigated. Experiments are carried out on finger knuckle databases that are accessible to the public, such as the PolyU Finger Knuckle Print Database and the IIT Delhi Finger Knuckle Database. A number of texture-based methods, namely Local Binary Patterns (LBP), Binarized Statistical Image Features (BSIF), Local Phase Quantization (LPQ) and Weber Local Descriptor (WLD), are utilized in the process of finger knuckle identification in this thesis. Experiments are conducted to show the influence of various texture-based approaches on the identification of finger knuckles.