An automatic on-line signature verification system based on three models

Abstract

This paper proposes a new automatic on-line signature verification system using a combination of function and parameter models of a signature. The function model is based on the normalized arc-length of the signature curve whilst the parameter models are shape and dynamics related. The personalized features constituting the parameter models are selected based on a kind of discrimination power function (DPF). The parameter models are analyzed using fuzzy logic rather than the traditional Euclidean distance metric. Experimental results are presented to show the effectiveness of the proposed signature verification system.