Stability, Speed and Accuracy for Online Signature Verification

We suggest a model of signature verification based upon handwriting generation studies and derive from it the characterization of the signing habits of a subject. Such characterization is given in terms of the signature’s stability regions, which are obtained by exploiting shape and temporal information conveyed by the genuine signatures captured by a writing pad. The effectiveness of the proposed method for characterizing the signing habits of a subject has been evaluated in a signature verification experiment on the Sabaci University Signatures (SUSIG) database. The experimental results, obtained by using very simple decision criteria in order to stress the role of stability regions on assessing the authorship of a signature, confirm that the proposed method captures to a large extent the behavior of the subject.