Biometric modality: Signatures – key considerations
Signatures are one of the most socially acceptable biometric applications as they have always been associated with document authentication, business transactions and letter writing in many cultures around the world. As such, they do not carry the same stigma as fingerprints or DNA that are often perceived to be only associated with criminals and crime investigation.
Static signatures can be susceptible to forgery by human or computer replication, but the dynamic signature recognition process is very accurate and extremely difficult for another person or machine to reproduce. However, the major challenge for both static and dynamic signature recognition systems is the changes that occur in the same person’s signature both in the short term and over time. These may be ad hoc or represent a gradual shift in writing style that might occur for a variety of reasons. Some of the latest dynamic systems have incorporated machine learning functions that monitor these changes and adapt the software accordingly.
