Page 30 - Automated Fingerprint Identification Systems (AFIS)

INTRODUCTION 15

preprinted paper stock, can capture finger images on a glass platen of a
device called a livescan. The livescan takes a picture of the finger in a fashion
similar to rolling a finger onto the glass platen of a very compact, very well-
engineered copy machine. In this process, a picture of each finger of the right
hand is taken, then the left hand, then the four fingers for simultaneous
impressions of the right hand fingers, then right thumb. This process would
be repeated for the left hand and the palms, and a mug shot might also
be taken.
These livescan images can be sent to the state identification bureau elec-
tronically, so that within minutes of receipt, the images have been electronically
classified for pattern and minutiae characteristics. There may be more than 100
of these unique minutiae for each finger and over 1,000 for each palm. The
database can then be searched for similar pattern and minutiae configurations
for two or more fingers, usually the index fingers or thumbs. In a parallel
process, the subject’s name may also be checked against all the names in the
Master Name Index database. When the search of each of the index fingers
produces the same candidate that the name search produced, there is a very
high degree of probability that it is a match. The images are considered to
belong to the same person regardless of the sex, age, or other information
captured in earlier fingerprintable events. It is an ident, an IAFIS term for a
positive identification.

1.8 THE IMPACT OF AFIS SYSTEMS
AFIS completely changed the identification business model. Identifications are
now made on finger images based on minutiae and ridge characteristics. Com-
puters search millions of records in seconds. If the images match an existing
record but the sex does not, the record is updated to indicate that both male
and female genders have been reported for this person. Regardless of name
given, sex reported, height, weight, age, etc., it is rare that a suspect will not be
identified if his or her finger images are already on file.
While AFIS systems have migrated into a variety of uses, their primary
purpose remains to determine if a person has been previously printed
(enrolled) and has any history in the locale. Identification based on fingerprints
is among the most accurate form of identification in existence. Identification
is not affected by the name, sex, or year of birth entered in the database. What
affects the search is the clarity of the finger images and the clarity of the images
in the database.
Table 1.1 details how AFIS systems have changed the business model for
identifications.