Information Technology Reference
In-Depth Information
requiring physical contact or proximity. However, they also have
elevated false acceptance rates compared to other solutions, partic-
ularly when attempting to discriminate between twins or siblings
with similar facial features or in the identification of facial geome-
tries from an oblique perspective angle.
Other biometric identification systems employ behavioral analysis,
including the following:
•
Gait identification.
Analysis of the movement patterns of a walk-
ing or running individual can be used for identification. Such sys-
tems are limited in use, because of the need for movement within
the identification area and an unobscured view of the subject.
•
Signature recognition.
Signature-based identification systems are
being used increasingly and include digital signature images along
with credit card point-of-purchase information. More robust versions
can identify the rate, stroke patterns, and level of pressure throughout
a signature in order to discriminate between authentic and counter-
feit signatures that may match visually but differ due to tiny varia-
tions of bone length and musculature in the hand. This solution can
be concealed under the target writing surface, but it is also subject to
an elevated false rejection rate compared to other biometric identifi-
cation methods because of variances in handwriting due to illness,
injury, level of attention, use of initials, or other variable factors.
•
Voice recog n it ion.
Voice-recognition analysis is one of the most
readily accepted forms of biometric identification. It involves the
identification of patterns in pitch, timbre, and tonal qualities in a
spoken pass phrase. Improvements in acoustic recording technolo-
gies may affect the viability of voice recognition for the purposes of
identification, but voice-recognition systems are being found increas-
ingly in automated telephone applications such as the 4-1-1 informa-
tion look-up service, available in most areas of the United States.
Background noise, changes in vocalization as a result of illness or
growth, and other factors can negatively affect the performance of
voice-recognition systems.
An additional form of “what you are” identification might be bet-
ter referred to as “where you are” and employs location-identifying
