Fingerprint Individuality Model Based on Pattern Type and Singular Point Attributes

Özet

This paper presents a singular point and pattern type model for the investigation of fingerprint individuality. The extraction of the singular point is based on the modified Poincare method while the determination of the pattern type is based on plane geometry and the attributes of the singular point on the quadrants. The experimental study involved Matlab version R2018a as the frontend while Microsoft Access Relational Database Management System served the backend. Benchmarked FVC2002 fingerprint database which comprises four datasets from different sources and of varied types were used for the experimental study. The experimental study established the viability and the functionality of the model while results for average matching time, false non match rate and false match rate confirmed that the model is practically feasible as well as its suitability for use in Automated Fingerprint Identification System (AFIS)..