The ability to uniquely associate a human fingerprint with an individual ha
s become an essential part of personal identification for security purposes
and law enforcement. This article presents an approach to fingerprint reco
gnition based on circularly sampling the fingerprint around an easily ident
ified center of the fingerprint. Three matching metrics are presented and c
ompared for matching fingerprints with no peculiarities, a missing area, an
d rotation with those in a database. The optimum metric, referred to as the
angular density, was seven times better for a self match than for a mismat
ch, even when the fingerprint was rotated or altered by erasing some of its
area.