Palm vein biometric identification system using local derivative pattern

This paper explains about the implementation of Local Derivative Pattern (LDP) as feature extraction algorithm and Histogram Intersection matching algorithm in a palm vein-based biometric identification system. LDP has been able to take on the image of the palm vein characteristics based on gray level differences of a pixel compared to adjacent pixels in the form of histogram. It is added by directional feature characteristic and it has order and radius parameters. Histogram Intersection method is used to calculate the similarity value of histogram characteristic between model and test data. The greater the matching value obtained, the more similar those histogram characteristics. This study showed the efficiency of LPD feature extraction algorithms which was able to get the best accuracy of 98.3%, with the configuration parameters: order value 2, radius 12, and image partition 4×4. Further testing was also done and obtains an optimal threshold value, which is about 37.2, generating FAR and FRR of 0.01 and 0.01 respectively.