Automated PolyU Palmprint sample Registration and Coarse Classification

Abstract: Biometric based authentication for secured access to resources has gained importance, due to their reliable, invariant and discriminating features. Palmprint is one such biometric entity. Prior to classification and identification registering a sample palmprint is an important activity. In this paper we propose a computationally effective method for automated registration of samples from PlolyU palmprint database. In our approach we preprocess the sample and trace the border to find the nearest point from center of sample. Angle between vector representing the nearest point and vector passing through the center is used for automated palm sample registration. The angle of inclination between start and end point of heart line and life line is used for basic classification of palmprint samples in left class and right class.