Uses ridges on the skin, Ridges =
lines. Valleys = Spaces between
Major and Minor features
3 Major Features
Arch, Loop, Whorl
Major features used to classify fingerprints
so they are more easily searchable
Only look in or compare to Arch records etc.
Minor features = Minutiae
Minutiae are the
distinguishing features
Ridge endings and burifications
in fingerprint pattern
Most algorithms are based
on munutiae
Fingerprint processing
involves several steps
Image is captured by sensor
noise is reduced IE:dirt, dryness, scars.
definition against the valleys is enhanced
Image is binarised =
reduced to black or white
Thinning is performed = reduction of ridges
down to a single pixel to help detect ridge
endings and burifications in the pattern
Ridges are traced,endings are found at
termination points of the ridges and
burifications are found at the junction of lines
Extracted features consists of the sequence of bits
representing for each minutuae, type of minutiae
(ridge or burification) its location and direction
Matching is done by comparing extracted
features against the reference template
Like comparing the shape of 2 graphs, if the graphs are similar
enough according to the tolerance threshold the result is a match
Some FP recognition can be done without
the minutiae instead using pattern matching
Advantages
Mature technology, lots of
development, investment and research.
Takes up 2/3 of Biometric Market
Easy to use, non intrusive
High levels of accuracy, low
error rates, low FMR, low FNMR
Long term stability of
FP over a lifetime
Ability to enrol
multiple fingers
Disadvantages
Inability to Enrol some users IE elderly,
manual workers, thin fingerprints (Asian)
Accuracy affected
by skin condition
Association with
forensic applications
Privacy issues with users
Sensors
Optical
Oldest / most mature
Finger placed on glass and Image
taken using optical device (camera)
Not much further development
Ultrasound
Uses ultrasound for capturing images
Un affected by dirt grease moisture etc
Better than optical but can only operate
between 10c -32 c. Mainly used indoors
Chip Based
Capacitive
Most widely used
chip based sensor
E-field
reads sub surface of
skin, not affected by
skin surface issues
Direct Optical
Scanning
Robust fibre optic
sensor
Thermal
Uses the fingers natural heat,
no optic or light source required