Comparing the use of local and global features of handwritten signatures for user authentication

Abstract: The paper compares two methods of user authentication based on a handwritten signature. The first is based on the extraction of local signature features as functions that depend on the time of its application. The second extracts global signature features that characterize it entirely without reference to the moments of application. Authentication algorithms are presented for each of the methods; experiments are carried out on real handwritten signatures obtained during the study.

Keywords: handwritten signature, authentication, machine learning, pattern recognition

For citation:
Andreevskih D. A., Razboynikov S. D., Silich M. P., Sarin K. S. Comparing the use of local and global features of handwritten signatures for user authentication. Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki, 2020, vol. 23, no. 4, pp. 30–34. DOI: 10.21293/1818-0442-2020-23-4-30-34