|
|
AFIS: Automated Fingerprint Identification System
Our Automated Fingerprint Identification System (AFIS)
is an application of large database indexing and retrieval. Designed for
biometric system integrators and law enforcement, our PositiveID™ AFIS system is
both fast and highly accurate for live-scan, 10-Print and latent matching with
the following features:
- Two-stage software matching process
- Possesses over 99.6% accuracy for two-finger matching
- Zero false accept rate
- Sub-millisecond timing per finger compare on regular off-the-shelf hardware
- Ability to detect multiple true IDents in the background database via absolute thresholds on match scores
- Independent of the use of cores/deltas
- Ability to incorporate the use of ridge counts
|
In its most basic form, PositiveID™ compares digitized fingerprint minutia maps of
an unknown fingerprint set against corresponding maps of reference fingerprint
sets already stored in memory. The system will identify and match unknown
fingerprints to those in the database by converting all of the fingerprints to
Attributed Relation Graphs (or ARGs); including nodes and branches, to which
attributes are appended.
Our system accuracy is based on tests performed on several independent databases, one
of which is NIST 9. At ARGTEC, we have applied the ARG representation and matching
technology to indexing, identification and retrieval in large fingerprint databases
with tremendous success in two major programs. For the U.S. FBI's Integrated
Automated Fingerprint Identification System (IAFIS) initiative, we developed the
fingerprint indexing, identification and retrieval subsystem. A patent was granted
for this system by the U.S. Patent Office on March 18, 1997. That system
significantly outperformed other competing approaches in a down-select contract
demonstration and was awarded the down select fielding contract.
PositiveID™ AFIS (Virtual Demo)
|
