Final Project Report System Identification Approach in Signature verification Qiushi Fu

Abstract

This report will present an investigation into one of the common used biometrics, personal signatures. From the perspective of system identification, the signature dynamics will be studied, and two novel verification algorithms will be discussed in details. They are spectral analysis and correlation method. We verified our algorithms using both public dataset and private dataset. The result will be shown and compared. Furthermore, two unfinished algorithms will also be proposed.