Person Identification Using Footprint Minutiae

Abstract

Increasing demand of biometric security system is leading us to evaluate more novel approach which could recognize and identify humans using their unique traits. Like fingerprint, footprint of human also share the unique traits which could be used as additional biometric trait in absence of major biometric trait (fingerprint, face). This could be installed at the places where it is mandatory to remove shoe for security or religious reasons, for example, airports and places of worship like mosques, pilgrimage, etc. We propose a minutiae-based footprint recognition system for person identification. Furthermore, credibility of fingerprint matching matrices is used for the evaluation of our footprint approach. Minutiae features of foot fingers are used to evaluate the performance of person identification using footprint. We have collected dataset of 120 persons including 78 males and 42 females at different resolution (500, 600, and 1200 dpi). Collected dataset involves two samples of each person for every foot finger. A series of dataset preprocessing operations are applied for getting good quality images which also ensures the correct region of interest detection for getting the large number of minutiae in a small region. Concept of crossing number is used for ridge termination and bifurcation extraction. The method achieves False Non Match Rate (FNMR) of $$0.4{\%}$$ with the False Match Rate (FMR) of 0.2%.