A PalmprintAuthentication Based on Local Texture Features

Abstract

Palmprint era is one of the biometric strategies used toidentify an character. Recognition of palmprints is based on the features likepalm traces, texture, ridges and many others. Several line and texture extraction techniqueshave already been proposed.A biometrictemplate consists of N m-element function vectors, in which N is the overallrange of overlapping subimages, and m is the wide variety of neighborhood Haralick functions per subimage inside the ROI. A stay biometric template and templates fromdatabase are matched in N matching modules. Based on fusion at the matchingscore level, the overall similarity measures among a stay biometric template andtemplates from the database are calculated. By the usage of the most of generalsimilarity measure and the 1-NN category rule, the very last decision (personidentification) is made.