|
Fingerprints
are composed of ridges and furrows arranged in different
patterns making them distinct between each individual.
Fingerprints contain special features called
“minutiae” that are extracted upon fingerprint
enrollment and serves as personal identification badge.
The
minutiae are
the points of interest or details in a fingerprint, such
as bifurcations (a ridge splitting into two) and ridge
endings.
Fingerprint
Identification requires three components to complete the
process: your fingerprint, the fingerprint recorder, and
the database. At
the initial phase called “enrollment”, fingerprints
are scanned, extracting the minutiae and transforms it
into encrypted numerical data to be stored as
“templates” in the database or written within a
card. Normally,
a fingerprint recorder will require users to input their
fingerprint thrice. This step allows the system to
select the best quality among three attempts based on
the predefined fingerprint classes of the system.
The fingerprint identification system categorizes
the fingerprints into different classes depending upon
the image quality.
The higher the class is, the better chance of
getting the fingerprint accepted by the system.
After the enrollment of the first fingerprint, a
substitute fingerprint is requested in case the first
finger is injured or unable to recognize.
Upon
fingerprint authentication, the system matches the given
minutiae of the fingerprint with the one stored within
the database. A
matching score, similarity degree among 2 fingerprints,
is computed. The
higher the score is, the higher is the degree of
similarity between fingerprints.
The deciding factor that determines whether the
fingerprint is accepted to be valid or not depends upon
the threshold score set within the system.
Threshold is a score or a level that the matching
score needs to reach or exceed in order for the
fingerprint to be deemed as valid identity.
If the threshold score is high, the barrier for
authenticating individuals is also elevated.
The
verification errors “False Acceptance Rate”(FAR) as
well as “False Rejection Rate”(FRR) arises in
relation with the threshold.
False Acceptance Rate is the rate of impostors
accepted by the system while False Rejection Rate is the
rate of valid users rejected by the system.
|
