Analysis and Implementation Scale Invariant Feature Transform (SIFT) for Authentication System Using Palm Vein

Analysis and Implementation Scale Invariant Feature Transform (SIFT) for Authentication System Using Palm Vein

Author		: GHOFINKA PUTRI; TJOKORDA AGUNG BUDI WIRAYUDA; SITI SAADAH, ST.
Published on	: International Conference on Information and Communication Technology 2015

Abstract

Data increasing every day become trigger to theft of personal data by those who don’t responsible of it. To overcome that problem, biometric identification system using palm vein that are located in human hand now become a hot issue to improving security in accessing data. The input image are form of palm photos taken using near-infrared (NIR), so those blood vessel pattern will show as black line. The first stage is determine the interest area that will be analyzed, then the image will pass though preprocessing steps to obtain the pattern of vein vessels. Next, feature extraction process is performed to get the unique features of the image from preprocessing result. In this final task, feature extraction method that used is Scale Invariant Feature Transform (SIFT). In many aspects of object recognition, SIFT has been able to show the best performance. Hopefully, by applying this method can give a good performance for biometric authentication system using palm vein. Based on research, the system is able to identify 8 of 10 people. By applying the SIFT method, the resulting value of the Equal Error Rate (EER) of 45% of the use CASIA database.