Multifinger Penetration Rate and ROC Variability for Automatic Fingerprint Identification Systems

Abstract

AFIS system design and performance is heavily impacted by data correlationsand error rate variability both within and acrossindividual fingerprint records, but little is known about these issues.In this paper, we report on experiments done withthe best algorithms from six major AFIS vendors tested using a 4128× 4080 database of electronically collected flat prints.We obtain Receiver Operating Characteristic curves for thumb throughring finger on right and left hands, as well as experimentalpenetration and binning error rates for one-, two-, four-, and eight-printsystems. Impact of binning on the overall ROC curveis measured. “Zero-effort” impostor error rate variability (“lambs”and “wolves”) is observed across the data.