biometricSignature authentication has had a great attention in the field of biometrics since it is a traditional method to personal verification. Most of the work available in the literature use kinetic (e.g. axial and normal forces), kinematic (e.g., position, velocity, acceleration, inclination angle) and spatiotemporal features (e.g., stroke durations, stroke length, in-air time) for the
signature authentication and have ignored the writer’s biomechanics. Therefore, this project studies the grip patterns that people use when writing in order to use these features to identify a person’s identity. We are proposing to use the biomechanics of handwriting for signature authentication by using a special instrumented pen while writing the signature on a digitizing tablet. This pen measures how hard you press on the page. It also measures how you hold the pen. This information will help us to describe the uniqueness of each person’s writing pattern and be able to identify an authentic signature from a forged signature. This research which is funded by NSERC and Syngrafii Corp. could lead to the development of a new device that will recognize signatures. Many industries (e.g. government, banks, retail and hospitals) can benefit from this device as a new anti-fraud tool that can improve security and privacy.


This project will answer the following questions:

  • What are the intra-individual variability of grip forces while writing a signature between different days and different times within the same day?
  • What are the features that can be used to discriminate between forged and authentic signatures based on the collected grip force data?
  • How does the grip force parameters, velocity, fluency and in-air time change with age, gender and handedness?

Progress and Future Plans

biometricA massive amount of genuine and forged signatures is being collected using a digitizing tablet and an instrumented writing utensil that measures the individual dynamic grip forces while writing. Adults of 18 years or older with no history of musculoskeletal injuries or neurological impairments are participating to make this data base. 500 genuine signatures are being collected from 5 participants in the Prism Lab and 6000 forged signatures will be collected from 300 adults visiting the Ontario Science Center.

Once the data is available, pre-processing techniques will be needed to prepare the data for analysis. Few analysis techniques will be performed including topographical analysis in order to find the grip force features that best discriminate between the authentic and forged signatures. These features can then be combined with other kinematic and spatiotemporal features to improve the verification results.

Research Team

Graduate Students
Bassma Ghali

Clinical Collaborators
Ervin Sejdić, Ph.D.
Evdokia Anagnostou, M.D.

Industry Partner
Singrafii Corp.