Biometrics: Personal Identification. BIOM 426: Biometrics Systems. Instructor: Natalia Schmid. Outline. Introduction Applications Identification methods Requirements to biometrics Biometrics technology Automatic Identification: design representation feature extraction
BIOM 426: Biometrics Systems
Instructor: Natalia Schmid
Person’s identity is everything what person represents and believes.
Engineering approach: reduce the problem to:
(i)some possession ("something what he has") or
(ii) some knowledge ("something what he knows")
Another approach: reduce it to a problem of authentication based on physical characteristics (physiological or behavioral).
Definition: Biometrics are person's identification based on his/her physiological or behavioral characteristics.
"something that you are"
Biometrics are automated methods of recognizing a person
based on a physiological or behavioral characteristic.
1. universality: everyone should have it
2. uniqueness: small probability that two persons are the same in terms of this characteristic
3. permanence: invariance with the time
4. collectability: can be measured quantitatively
5. performance: high identification accuracy
6. acceptability: acceptance by people
7. circumvention: how easy to fool the system by fraudulent technique
Accepted and studied biometrics: voice, hand geometry, gait, fingerprint, ear, face, iris, retina, fingerprint, infrared facial and hand vein thermograms, key stroke, signature, DNA
DNA, signature, and fingerprint are recognized in court of law
Fingerprints:are graphical flow-like ridges.
Their formation depends on embrionic development. Factors:
(i) genetic, (ii) environmental
(i) scanning inked impression, (ii) life-scan
image, ridges, minutia (features derived from ridges), or pores
Basic approaches to identification: (i) correlation based; (ii) global ridge patterns (classes); (iii) ridge patterns; (iv) fingerprint minutiae (ridge endings and bifurcations)
Face:one of the most acceptable biometrics
Two identification approaches:
(i) transform (eigenvalues, analysis of covariance matrix, orthonormal basis vectors)
(ii)attribute-based approach (geometirc features)
Factors that influence recognition:
(i) facial disguise
(ii) facial expressions
(iii) lighting conditions
(iv) pose variation
Iris:is one of the most reliable
Frontal images are obtained using near infrared camera (320 x 480 pixels) at distance < 1 meter.
Iris images are (i) segmented and
Twins have different iris patterns.
Voice:is a behavioral characteristic and is not sufficiently unique (large database).
Processing: signal subdivided into a few frequency bands. The most commonly used feature is cepstral feature (log of FT in each band).
Matching strategies: hidden Markoff, vector quantization, etc.
Types of verification:
Voice print is highly accepted
Used for identification over the
Easy to fool the system.
Infrared Facial and Hand Vein Thermogram:
Human bodies radiate heat.
Infrared sensors acquire an image of heat distribution along the body. Images = thermograms.
Imaging methods: similar to visible spectrum photographs.
Processing: raw images are normalized with respect to heat radiating from landmark features.
In uncontrolled environment, other sources of heat could be disturbance.
Gait:is the specific way one walks.Complex spatio-temporal behavioral characteristic.
Gait is not unique and does not stay invariant over time.
It is influences by: distribution of body weight, injuries involving joints or brain, aging.
Gait features are derived from a video sequence and consists of charactertization of several movements (computer vision problem).
Retinal vasculature isrich in structure.
Unique characteristic of each individual and each eye.
Not easy to change or replicate.
Image capture: requires person to peep into an eye-piece and focus on a specific spot. A predetermined part of retinal vasculature is imaged.
Not accepted by public.
Can reveal some medical conditions as hypertension.
Signature:the way person signs his/her name.
Highly acceptable behavioral biometrics.
Evolves over time and depends on physical and mental conditions.
Modeling the invariance and automating signature recognition process is
Two approaches to signature verification:
(i) static (geometric features = strokes)
(ii) dynamic (strokes and acceleration, velocity, trajectory)
Hand and finger geometry:is used for
access control (50% of market).
System captures frontal and side views of
Measurements: length and width of fingers,
The representation requirements are only
Hand geometry is not unique but highly
From “Biometrics: Personal Identification in Networked Society,” p. 16
Prehistoric Chinese used thumb-
print for identification
Alphonse Bertillon’s System of
Anthropometric Identification (1882)
is based on bodily measuments,
physical description, and photographs.
Henry’s fingeprint classification
system (1880) classifies in > 100 classes.
are developed for:
(i) matching of biometrics
(ii) searching databases
is due to: inexpensive computer resources, advances in
computer vision, pattern recognition, and image understanding.
Identification system operates in two modes:
(i) enrollment mode and
- biometric measurement is captured
- information from raw data extracted;
- (feature, person) information is stored;
- ID is issued (for verification).
- biometric is sensed (live-scan);
- features are extracted from the raw data;
- match is performed (search of the database).
In verification mode, person presents ID. Then system performs
match only against one template in the database.
Architecture of a typical pattern recognition system (see A. K. Jain, et al., p. 22).
Given the speed, accuracy, and cost specifications
1. How to collect the input data? (3D, 2D, multiple views, high or low resolution)
2. Internal representation (features) for automatic feature extraction
3. How to extract features? (Algorithms, etc.)
4. How to select the "matching" metric? (Measurements are made in specific space)
5. How to implement it?
6. Organization of database
7. Effective methods for searching a template in the database (binning, etc.)
Quality of collected data determines
performance of the entire system.
(i) quality assessment
(ii) segmentation (separation of the data into foreground and background).
(i) richer data (3D, color, etc.)
(ii) metrics for assessment quality of measurements.
(iii) realistic models
Similarity metric should be robust against:
- structural and statistical variations,
- aging, and artifacts of feature extraction module.
Example: signature (hard to define the ground truth)
Performance is determined by: (i) representation and (ii) similarity metric.
Trade-off: better engineering design vs. more complex matcher.
Example: fingerprint (variations in features and rigid matcher vs.
Sources of distortion and noise:
(i) inconsistent contact (3D-to-2D)
(ii) non-uniform contact (due to dryness of skin, sweat, dirt,
humidity in the air, etc.)
(iii) irreproducible contact (injuries to the finger)
(iv) feature extraction artifacts (measurement error)
(v) sensing itself adds noise
An end-user questions:
(i) Does the system makes an accurate identification?
(ii) Is the system sufficiently fast?
(iii) What is the cost of the system?
Because of noise, distortions, and limited information no metric is adequate for reliable identification.
- genuine individual,
Four types of outcomes:
(1) Genuine individual is accepted (true)
(2) Genuine individual is rejected (error)
(3) Imposter is rejected (true)
(4) Imposter is accepted (error)
FAR - false acceptance rate
FRR - false rejection rate
EER - equal error rate
Given a database, performance is a RV and only can be estimated.
Measure of performance:
ROC - receiveroperating curve