The Architecture of Biometrics Systems

1
The Architecture of Biometrics Systems

• Bojan Cukic

2
Biometric Systems Segment Organization

• Introduction
• System architecture

3
Introduction

• Biometrics
• Engineering Definition and Approaches
• Definition, Criteria for Selection
• Survey of Current Biometrics and Relative
  Properties
• Introduction to socio-legal implications and
  issues

4
Recap Identification in the 21st Century

• Dispersion of people from their Natural ID
  Centers
• Social units have grown to tens of thousands or
  millions/billions.
• Need to assure associations of identity with
  end-to-end transactions without physical presence
• Project your presence (ID) instantly, accurately,
  and securely across any distance

5
Identification Methods

• We need to achieve this recognition automatically
  in order to authenticate our identity.
• Identity is not a passive thing, but associated
  with an act or intent involving the person with
  that identity
• Seek a manageable engineering definition.

6
Biometric Identification

• Pervasive use of biometric ID is enabled by
  automated systems
• Enabled by inexpensive embedded computing and
  sensing.
• Computer controlled acquisition, processing,
  storage, and matching using biometrics.
• Biometric systems are one solution to increasing
  demand for strong authentication of actions in a
  global environment.
• Biometrics tightly binds an event to an
  individual
• A biometric can not be lost or forgotten, however
  a biometric must be enrolled.

7
What is an Automated Biometric System?

• An automated biometric system uses biological,
  physiological or behavioral characteristics to
  automatically authenticate the identity of an
  individual based on a previous enrollment event.
• For the purposes of this course, human identity
  authentication is the focus. But in general,
  this need not necessarily be the case.

8
Characteristics of a Useful Biometric

• If a biological, physiological, or behavioral
  characteristic has the following properties
• Universality
• Uniqueness
• Permanence
• Collectability
• .then it can potentially serve as a
  biometric for a given application.

9
Useful Biometrics

• 1. Universality
• Universality Every person should possess this
  characteristic
• In practice, this may not be the case
• Otherwise, population of nonuniversality must be
  small lt 1

10
Useful Biometrics

• 2. Uniqueness
• Uniqueness No two individuals possess the same
  characteristic.
• Genotypical Genetically linked (e.g. identical
  twins will have same biometric)
• Phenotypical Non-genetically linked, different
  perhaps even on same individual
• Establishing uniqueness is difficult to prove
  analytically
• May be unique, but uniqueness must be
  distinguishable

11
Useful Biometrics

• 3. Permanence
• Permanence The characteristic does not change in
  time, that is, it is time invariant
• At best this is an approximation
• Degree of permanence has a major impact on the
  system design and long term operation of
  biometrics. (e.g. enrollment, adaptive matching
  design, etc.)
• Long vs. short-term stability

12
Useful Biometrics

• 4. Collectability
• Collectability The characteristic can be
  quantitatively measured.
• In practice, the biometric collection must be
• Non-intrusive
• Reliable and robust
• Cost effective for a given application

13
Current/Potential Biometrics

• Voice
• Infrared facial thermography
• Fingerprints
• Face
• Iris
• Ear
• EKG, EEG
• Odor

• Gait
• Keystroke dynamics
• DNA
• Signature
• Retinal scan
• Hand finger geometry
• Subcutaneous blood vessel imaging

• What is consensus evaluation of current
  biometrics based on these four criteria?

14
System-Level Criteria

• Our four criteria were for evaluation of the
  viability of a chosen characteristic for use as a
  biometric
• Once incorporated within a system the following
  criteria are key to assessment of a given
  biometric for a specific application
• Performance
• User Acceptance
• Resistance to Circumvention

15
Central Privacy, Sociological, and Legal
Issues/Concerns

• System Design and Implementation must adequately
  address these issues to the satisfaction of the
  user, the law, and society.
• Is the biometric data like personal information
  (e.g. such as medical information) ?
• Can medical information be derived from the
  biometric data?
• Does the biometric system store information
  enabling a persons identity to be
  reconstructed or stolen?
• Is permission received for any third party use of
  biometric information?

16
Central Privacy, Sociological, and Legal
Issues/Concerns (2)

• Continued
• What happens to the biometric data after the
  intended use is over?
• Is the security of the biometric data assured
  during transmission and storage?
• Contrast process of password loss or theft with
  that of a biometric.
• How is a theft detected and new biometric
  recognized?
• Notice of Biometric Use. Is the public aware a
  biometric system is being employed?

17
Biometric System Design

• Target Design/Selection of Systems for
• Acceptable overall performance for a given
  application
• Acceptable impact from a socio-legal perspective
• Examine the architecture of a biometric system,
  its subsystems, and their interaction
• Develop an understanding of design choices and
  tradeoffs in existing systems
• Build a framework to understand and quantify
  performance

18
Automated Biometric Identification A
Comprehensive View
Template Storage Database Search Match, Retrieval
MATCH?
Identification Process
Identity
Databases, Time series data Data
Mining Statistical Modeling
Arrhythmia, SIDS,
Biological Agents, Microbial pathogens...
Action Logical/Phys. Access (IA, medical, bio)
19
Biometric Systems Segment Organization

• Introduction
• System Architecture

20
System Architecture

• Application
• Authentication Vs. Identification
• Enrollment, Verification Modules
• Architecture Subsystems

21
Biometric Applications

• Four general classes
• Access (Cooperative, known subject)
• Logical Access (Access to computer networks,
  systems, or files)
• Physical Access (access to physical places or
  resources)
• Transaction Logging
• Surveillance (Non-cooperative, known subject)
• Forensics (Non-cooperative or unknown subject)

22
Biometric Applications (2)

• Transactions via e-commerce
• Search of digital libraries
• Computer logins
• Access to internet and local networks
• Document encryption
• Credit cards and ATM cards
• Access to office buildings and homes
• Protecting personal property
• Tracking and storing time and attendance
• Law enforcement and prison management
• Automated medical diagnostics
• Access to medical and official records.

23
System Architecture

• Architecture Dependent on Application
• Identification Who are you?
• One to Many (millions) match (1Many)
• One to few (less than 500) (1Few)
• Cooperative and Non-cooperative subjects
• Authentication Are you who you say you are?
• One to One Match (11)
• Typically assume cooperative subject
• Enrollment and Verification Stages common to both.

24
System Architecture (2)
Enrollment Capture and processing of user
biometric data for use by system in subsequent
authentication operations.
Database Template Repository
Acquire and Digitize Biometric Data
Extract High Quality Biometric Features/Represent
ation
Formulate Biometric Feature/Rep Template
Authentication/Verification Capture and
processing of user biometric data in order to
render an authentication decision based on the
outcome of a matching process of the stored to
current template.
Acquire and Digitize Biometric Data
Extract High Quality Biometric Features/Represent
ation
Formulate Biometric Feature/Rep Template
Template Matcher
Decision Output
25
System Architecture (3)

• Authentication Application
• Enrollment Mode/Stage Architecture

Additional image preprocessing, adaptive
extraction or representation
Require new acquisition of biometric
No
Quality Sufficient?
Biometric Data Collection
Signal Processing, Feature Extraction,
Representation
Transmission
Yes
Database
Generate Template
26
System Architecture (4)

• Authentication Application
• Verification/Authentication Mode/Stage
  Architecture

No
Additional image preprocessing, adaptive
extraction/representation
Require new acquisition of biometric
Quality Sufficient?
Signal Processing, Feature Extraction,
Representation
Biometric Data Collection
Transmission
Yes
Generate Template
Database
Template Match
Decision Confidence?
No
Yes
27
Architecture Subsystems

• Data Collection
• Transmission
• Signal Processing/Pattern Matching
• Database/Storage
• Decision
• What comprises these subsystems and how do they
  interact with other elements (what are their
  interface and performance specifications?)

28
Architecture Subsystems (2)

• Data Collection Module
• Biometric choice, presentation of biometric,
  biometric data collection by sensor and its
  digitization.

Recollect
Biometric Data Collection
Signal Processing Feature Extraction
Representation
Transmission
Biometric
Presentation
Sensor
29
Architecture Subsystems (3)

• Transmission Module
• Compress and encrypt sensor digital data, reverse
  process.

Recollect
Transmission
Biometric Data Collection
Signal Processing, Feature Extraction, Representat
ion
Biometric
Presentation
Sensor
Compression
Transmission
Decompress
Encryption
Decryption
30
Architecture Subsystems (4)

• Signal Processing/Matching Module
• Be aware of potential transmission prior to match

No
Reprocess
Recollect
Transmission
Quality Control
Signal Processing Feature Extraction, Representati
on
Yes
Compression
Transmission
Decompress
Encryption
Decryption
Generate Template
Database
Template Match
Decision Confidence?
No
Yes
31
Architecture Subsystems

• Database module
• In what form is biometric stored? Template or
  raw data?

No
Reprocess
Recollect
Transmission
Quality Control
Signal Processing Feature Extraction, Representati
on
Compression
Transmission
Expansion
Encryption
Decryption
Yes
Generate Template
Database Templates Images
Template Match
Biometric Template A file holding a
mathematical representation of the identifying
features extracted from the raw biometric data.
Decision Confidence?
No
Yes
32
Architecture Subsystems

• Decision module
• Is there enough similarity to the stored
  information to declare a match with a certain
  confidence ?

No
Reprocess
Recollect
Transmission
Quality Control
Signal Processing Feature Extraction, Representati
on
Compression
Transmission
Decompress
Encryption
Decryption
Yes
Generate Template
Database Templates Images
Template Match
Decision Confidence?
Decision Confidence?
No
Yes