Friday, January 8, 2010

Biometrics

BIOMETRICS

History of Biometrics

Biometrics dates back to the ancient Egyptians who measured people to identifythem. Such rudimentary means of identification based on measurements ofparts of bodies or aspects of behavior have continued to be used ever sincethroughout the centuries. Fingerprint identification dates back to ancient China.Identification based on fingerprints has been in effect in the United States andWestern Europe for over 100 years.Commercial advancements for biometric devices began in earnest in the 1970s when a system called Identimat which measured the shape of the hand and length of the fingers was used as part of a time clock at Shearson Hamill, a WallStreet investment firm. Subsequently, hundreds of Identimat devices were used to establish identity for physical access at secure facilities run by Western Electric, U.S. Naval Intelligence, the Department of Energy, U.S. Naval Intelligence and like organizations.2 Identimat went out of business in the 1980s, but it set the stage for future biometric identification systems based on hand measurement. Progress was made on fingerprint biometric devices during the 1960s and 1970s when a number of companies developed products to automate the identification of fingerprints for use in law enforcement. In the late 1960s, the FBI began to automatically check fingerprints, and by the mid 1970s, it had installed a number of automatic fingerprint systems across the U.S. Automated Fingerprint Identification Systems (AFIS) are now used by police forces all around the globe. This widespread use of fingerprint data for law enforcement lends a ‘Big Brother’ feel to the use of fingerprint biometrics for identification, making it important for current fingerprint identification system providers to reassure consumers that their identity is ‘safe,’ their privacy maintained, and that their fingerprint will not be entered into a law enforcement database. Consumers must understand that current fingerprint recognition systems used for digital transactions differ widely from traditional AFIS systems. Automated systems for measuring other biometrics developed similarly to those used with fingerprints. The first systems to measure the retina were introduced in the 1980s. The work of Dr. John Daughman at Cambridge University led to the first iris measurement technology. Identification based on signature and face biometrics is relatively new. Biometrics has been widely researched inside certain universities for the past two to three decades, and most commercial products emerging today have strong roots inside institutions of advanced education. Caltech and MIT are twoleaders in the study of biometrics and the related fields of pattern recognition, learning theory and artificial intelligence. Because of its inherent complexity and because of their longer history with biometrics, individuals inside universities are closely involved with the most important product innovations involving biometrics.

Definition

The term "Biometrics" has also been used to refer to the emerging field of technology devoted to identification of individuals using biological traits, such as those based on retinal or iris scanning, fingerprints, or face recognition. Biometric technology is used for automatic personal recognition based on biological traits—fingerprint, iris, face, palm print, hand geometry, vascular pattern, voice—or behavioral characteristics—gait, signature, typing pattern. Fingerprinting is the oldest of these methods. This paper mainly proposes the use of security systems using the biometrics. There are many types of security systems that enhance with the biometrics technique,here our paper mainly focus on the iris recognition, the iris recognition which supports with the iris scanning. In these pilots the customer’s iris data became the verification tool for access to the bank account, thereby eliminating the need for the customer to enter a PIN number or password. When the customer presented their eyeball to the ATM machine and the identity verification was positive, access was allowed to the bank account. These applications were very successful and eliminated the concern over forgotten or stolen passwords and received tremendously high customer approval ratings. Iris recognition technology combines computer vision, pattern recognition, statistical inference, and optics. Its purpose is real-time, high confidence recognition of a person's identity by mathematical analysis of the random patterns that are visible within the iris of an eye from some distance. Because the iris is a protected internal organ whose random texture is complex, unique, and stable throughout life, it can serve as a kind of living passport or password that one need not remember but can always present. Because the randomness of iris patterns has very high dimensionality, recognition decisions are made with confidence levels high enough to support rapid and reliable exhaustive searches through national-sized databases. Here there are some advantages that combine at SecuriMetrics, we know that the most effective method of obtaining an accurate identification is the proper use of biometric technology – identification of an individual.


Introduction

Biometrics are used to identify the input sample when compared to a template, used in cases to identify specific people by certain characteristics. Biometrics are used to identify the input sample when compared to a template, used in cases to identify specific people by certain characteristics. Standard validation systems often use multiple inputs of samples for sufficient validation, such as particular characteristics of the sample. This intends to enhance security as multiple different samples are required such as security tags and codes and sample dimensions.

Human Biometrics Characteristics

Physiological are related to the shape of the body. The oldest traits, which have been used for more than 100 years, are fingerprints. Other examples are face recognition, hand geometry and iris recognition.

Recently, a new trend has been developed that merges human perception to computer database in a brain-machine interface. This approach has been referred to as Cognitive biometrics. Cognitive biometrics is based on specific responses of the brain to stimuli which could be used to trigger a computer database search and Behavioral are related to the behavior of a person.

Other biometric strategies are being developed such as those based on gait (way of walking), retina, hand veins, ear canal, facial thermogram, DNA, odor and scent and palm prints.

Functions:
A biometric system can provide the following two functions

ü Verification

ü Identification

Biometric system

The National Science and Technology Council provides the following overview of biometric system components: “A typical biometric system is comprised of five integrated components: A sensor is used to collect the data and convert the information to a digital format. Signal processing algorithms perform quality control activities and develop the biometric template. A data storage component keeps information that new biometric templates will be compared to. A matching algorithm compares the new biometric template to one or more templates kept in data storage. Finally, a decision process (either automated or human-assisted) uses the results from the matching.

Comparison of various biometrics technologies

Biometrics:

Universality

Uniqueness

Permanence

Collectability

Performance

Acceptability

Face

H

L

M

H

L

H

Fingerprint

M

H

H

M

H

M

Hand geometry

M

M

M

H

M

M

Keystrokes

L

L

L

M

L

M

Hand veins

M

M

M

M

M

M

Iris

H

H

H

M

H

L

Retinal scan

H

H

M

L

H

L

Signature

L

L

L

H

L

H

Voice

M

L

L

M

L

H

Facial thermograph

H

H

L

H

M

H

Odor

H

H

H

L

L

M

DNA

H

H

H

L

H

L

Gait

M

L

L

H

L

H

Ear Canal

M

M

H

M

M

H

Future enhancements

ü 3D Face Recognition

ü Face Recognition in Video

ü Extended Feature Set for Fingerprint Matching

ü Multispectral Fingerprint Matching

ü Fingerprint Individuality

ü Dental Biometrics

ü Multibiometrics

ü Image Quality

ü Biometric System Security

ü Sample Size Requirement for Performance Evaluation

Introduction to Iris Recognition

Iris recognition technology combines computer vision, pattern recognition, statistical inference, and optics. Its purpose is real-time, high confidence recognition of a person's identity by mathematical analysis of the random patterns that are visible within the iris of an eye from some distance. Because the randomness of iris patterns has very high dimensionality, recognition decisions are made with confidence levels high enough to support rapid and reliable exhaustive searches through national-sized databases.

Finger Prints

Finger prints is one of the identification method related to the Biometrics. All fingerprints are unique. The critical issue is whether we can get to the information that is unique and express it in a way that meets the objective of positive identification. Getting to and expressing the unique information in the fingerprint biometric is the Mission of Digital Personal and its Fingerprint Recognition Engine. Although the field of biometrics is still in its infancy, it’s inevitable that biometric systems will play a critical role in the future of security. Strong or two-factor authentication —identifying One self by two of the three methods of something you know (for example, a password), have (for example, a swipe card), or is (for example, a fingerprint) —is becoming more of a de facto standard in secure computing environments.

Introduction

Biometrics refers to authentication techniques that rely on measurable physiological and individual characteristics that can be automatically verified. In other words, we all have unique personal attributes that can be used for distinctive identification purposes, including a fingerprint, the pattern of a retina, and voice characteristics. It uses some algorithm and some objectives for identify the personality they are given below

  1. Face
  2. Fingerprint
  3. Handprint
  4. Iris
  5. Retina
  6. Signature
  7. Voice
  8. Watermarking

The Basics of Fingerprint identification

Ridges

The skin on the inside surfaces of our hands, fingers, feet, and toes is “ridged” or Covered with concentric raised patterns. These ridges are called friction ridges and they serve the useful function of making it easier to grasp and hold onto objects and surfaces without slippage. It is the many differences in the way friction ridges are patterned, broken, and forked which make ridged skin areas, including fingerprints, unique.

Fingerprint Identification Terminology

Fingerprints are extremely complex. In order to “read” and classify them, certain

Defining characteristics are used, many of which have been established by law Enforcement agencies as they have created and maintained larger and larger Databases of prints. Even though biometrics companies like DigitalPersona do Not save images of fingerprints and do not use the same manual process to Analyze them, many of the methodologies that have been established over the years in law enforcement are useful for digital algorithms as well.

Global Versus Local Features

We make use of two types of fingerprint characteristics for use in identification of individuals: Global Features and Local Features. Global Features are those characteristics that you can see with the naked eye. Global Features include:

  • Basic Ridge Patterns
  • Pattern Area
  • Core Area
  • Delta
  • Type Lines
  • Ridge Count

The Local Features are also known as Minutia Points. They are the tiny, unique characteristics of fingerprint ridges that are used for positive identification. It is possible for two or more individuals to have identical global features but still have different and unique fingerprints because they have local features – minutia points - that are different from those of others.

Global Features

§ Pattern Area The Pattern Area is the part of the fingerprint that contains all

the global features. Fingerprints can be read and classified based on the

information in the Pattern Area. Certain minutia points that are used for final

identification might be outside the Pattern Area. One significant difference

between DigitalPersona’s fingerprint recognition algorithm and those of

competing companies is that Digital Persona uses the entire fingerprint for

analysis and identification, not just the Pattern Area. While other companies’

devices require users to line up their fingerprints on the sensor,

Digital Persona acquires a greater amount of information over the entire

fingerprint, and can obtain enough information to "read" a print even if only

part of the print is placed on the sensor.

§ Core Point -- The Core Point, located at the approximate center of the finger

impression, is used as a reference point for reading and classifying the print.

§ Type Lines – Type Lines are the two innermost ridges that start parallel,

diverge, and surround or tend to surround the pattern area. When there is a

definite break in a type line, the ridge immediately outside that line is

considered to be its continuation.

§ Delta – The Delta is the point on the first bifurcation, abrupt ending ridge,

meeting of two ridges, dot, fragmentary ridge, or any point upon a ridge at or

nearest the center of divergence of two type lines, located at or directly in

front of their point of divergence. It is a definite fixed point used to facilitate

ridge counting and tracing.

Basic Ridge Patterns

Basic ridge patterns is classifies in to three things they are given below

1. loop

2. Arch

3. Whorls.

LOOP

The loop is the most common type of fingerprint pattern and accounts for about

65% of all prints.

ARCH

The Arch pattern is a more open curve than the Loop. There are two types of

arch patterns – the Plain Arch and the Tented Arch.

WHORL

Whorl patterns occur in about 30% of all fingerprints and are defined by at least

one ridge that makes a complete circle.

Constructing the Ridge Feature Map

The 240 fi 240 input Fingerprint image, I, is convolved with the 8 Gabor filters, {Ge}. Since the input image may be noisy, it is first enhanced before applying the filters. Enhancement improves the clarity of the ridge and furrow structure in the fingerprint image . We use the technique described in to enhance the fingerprint image . A segmentation algorithm is also applied on the input image to identify the foreground and background regions. The foreground corresponds to those regions in the image that have ridges and furrows, while the background represents those regions that do not have this information . Segmentation is useful during the matching phase, when the distance

between two feature maps is computed. Let H indicate the 240fi240 enhanced image. Convolving H with the 8 Gabor filters in the spatial domain would be a computationally intensive operation. In order to speed-up this operation, the convolution is performed in the frequency domain. Let F(H) denote the discrete Fourier transform of H, and let F(Gfi)

indicate the discrete Fourier transform of the Gabor filter having the spatial

orientation fi as described by Equation (1). Then the Gabor filtered image, Vfi,

may be obtained as,

Vfi = F*1[F(H)F(G*)];

where F fi1is the inverse Fourier transform. 8 filtered images are obtained in

this way . Each Vfi is used to construct a standard deviation image, Sfi, where Sfi(x; y) represents the standard deviation of the pixel intensities in a 16 * 16 neighborhood of (x; y) in Vfi. The standard deviation map, S = fS*g, comprises of 8 images corresponding to the 8 filtered images. Thus, the standard deviation map, S, captures the variation in the ridge strength at various orientations . Each standard deviation image, Sfi, is then sampled at regular intervals (every 16 th pixel) in both the horizontal and vertical directions to obtain the ridge feature image, Rfi (Figure 5). The ridge feature map, R = fRfig, is composed of these 8 images. The size of Rfi (15 fi15) is lesser than that of Sfi (240 fi 240). We, therefore, have a compact fixed-length (15 fi 15 fi 8 = 1; 800-valued) representation for the fingerprint.

Fingerprint Matching Using Ridge Feature Maps

The process of fingerprint matching involves comparing a query print with a set of one or more template prints. Prior to the matching process, ridge feature maps are extracted from all template images present in the database. When a query print, Q, is presented to the system, it is matched against a template ridge map, R T = fR T fi g as follows:

1. The query image is enhanced and the set of 8 Gabor filters is applied to the enhanced image, resulting in 8 filtered images.

2. The standard deviation map, S Q = fSQ for the query image is constructed

using these filtered images. Appeared in Proc. of Post-ECCV Workshop on Biometric Authentication, LNCS 2359, pp.48-57, Denmark, June 1, 2002.

2.1 Constructing the Ridge Feature Map

The 240 fi 240 input fingerprint image, I, is convolved with the 8 Gabor filters, fGfig. Since the input image may be noisy, it is first enhanced before applying the filters. Enhancement improves the clarity of the ridge and furrow structure in the fingerprint image We use the technique described to enhance the fingerprint image (Figure 2(b)). A segmentation algorithm is also applied on the input image to identify the foreground and background regions. The foreground corresponds to those regions in the image that have ridges and furrows, while the background represents those regions that do not have this information. Segmentation is useful during the matching phase, when the distance between two feature maps is computed. Let H indicate the 240fi240 enhanced image. Convolving H with the 8 Gabor filters in the spatial domain would be a computationally intensive operation. In order to speed-up this operation, the convolution is performed in the frequency domain. Let F(H) denote the discrete Fourier transform of H, and let F(Gfi) indicate the discrete Fourier transform of the Gabor filter having the spatial orientation fi as described by Equation (1). Then the Gabor filtered image, Vfi,

may be obtained as,

Vfi = Ffi1[F(H)F(Gfi)]; (2)

where Ffi1is the inverse Fourier transform. 8 filtered images are obtained in this way (Figure 3). Each Vfi is used to construct a standard deviation image,

Sfi, where Sfi(x; y)

represents the standard deviation of the pixel intensities in a 16 fi 16 neighborhood of (x; y) in Vfi. The standard deviation map, S = fSfig, comprises of 8 images corresponding to the 8 filtered images. Thus, the standard deviation map, S, captures the variation in the ridge strength at various orientations. Each standard deviation image, Sfi, is then sampled at regular intervals (every 16 th pixel) in both the horizontal and vertical directions to obtain the ridge feature image, Rfi. The ridge feature map, R = fRfig, is composed of these 8 images. The size of Rfi (15 fi15) is lesser than that of Sfi (240 fi 240). We, therefore, have a compact fixed-length (15 fi 15 fi 8= 1; 800-valued) representation for the fingerprint.

Fingerprint Matching Using Ridge Feature Maps

The process of fingerprint matching involves comparing a query print with a set of one or more template prints. Prior to the matching process, ridge feature maps are extracted from all template images present in the database. When a query print, Q, is presented to the system, it is matched against a template ridge map, R T = fR T fi g as follows:

1. The query image is enhanced and the set of 8 Gabor filters is applied to the enhanced image, resulting in 8 filtered images.

2. The standard deviation map, S Q = fSQ fi g, for the query image is constructed

3. Each of the 8 template ridge feature images, R Tfi , is `expanded' to the sizeof SQ

fi by interpolating with 0's. Let the ridge feature map consisting of theinterpolated images be indicated by S T = fS T

4. To determine the alignment between S Q and S T , a 2D correlation of the two

maps is performed. Correlation involves multiplying corresponding entries in

the two maps at all possible translation offsets, and determining the sum. The

offset that results in the maximum sum is chosen to be the optimal alignment

between the two maps. Correlation is done in the frequency domain, and

every offset is appropriately weighted. The weighting is necessary to account

for the amount of overlap between the two maps. Let UTQ represent the

unweighted correlation matrix, and CTQ represent the weighted correlation

matrix. Let N fi N be the size of a standard deviation image

Experiments and Results

Our database consists of fingerprint impressions (300 * 300) obtained from 160 users using the Veridicom sensor. Each user provided 4 different impressions (over 2 time sessions) of each of 4 different fingers - the left index finger, the left middle finger, the right index finger and the right middle finger. A set of 2; 560 (160 * 4 * 4) images were collected. The 300 * 300 images were resized to 240 fi240 (inter-ridge spacing changed from 10 pixels to 8 pixels) in order to speed-up the Fourier operations. This database is a difficult one for a fingerprint matcher due to the following reasons: (a) There is temporal variance imposed on the fingerprint impressions as they were acquired over two different time sessions. (b) The users, though cooperative, were non-habituated users of the system. (c) Some users were observed to have dry fingers that resulted in partial or faint fingerprint images. Initial experiments on this database indicate that the proposed technique provides a very good alignment of fingerprint image pairs. We compare the proposed technique with a minutiae-based matcher by plotting the Genuine Accept Rate against the False Accept Rate at various thresholds of the matching score. As expected, the minutiae-based matcher demonstrates better performance than the correlation-based matcher. However, fusing the two matchers (by normalizing and adding the matching scores) results in an improved performance of the fingerprint verification system

Fingerprint Recognition Systems

Fingerprints have been used for centuries for establishing identity. Digital Persona regards the fingerprint as the most viable biometric for mass-market identification schemas. In a general sense, fingerprint recognition technologies analyze global pattern schemas along with small unique marks on the fingerprint known as minutiae, which are the ridge endings and bifurcations

or branches in the fingerprint ridges. The data that is extracted from fingerprints is extremely dense which explains why fingerprints are an extremely reliable means of personal identification. There are on average seventy unique, measurable minutiae points in each fingerprint, and each point has seven unique characteristics - more than enough to establish identity. Should we desire an Biometric Solutions to Personal Identification even more foolproof identification schema, because each of our fingerprints is unique, we can use as many as ten fingerprints which yield at least 4900 independent measurable characteristics. Here we give an overview of fingerprint capture, extraction and verification. Optical and Capacitive Technologies for Fingerprint Capture

The two main technologies used to capture fingerprints are optical and capacitive. Optical technologies require a light source that is refracted through a prism. The fingerprint is placed on a surface known as a platen. Light shines on the print and an impression is captured. Digital Persona has created a proprietary approach to this technology, which eliminates the need for a large, costly prism, replacing it with a thin proprietary component that acts as a prism would. With capacitive-based semiconductor technologies, the finger is placed on the sensor chip and the ridges and valleys create capacitance variations between the skin and the chip. The chip captures the fingerprint by measuring the spatial variations in the field voltages.

Weaknesses of Capacitive Solutions

Capacitive sensors call for a chip as large as the fingerprint and this chip is expensive. Several companies attempt to get around this problem by offering smaller chips, which measure a part of the fingerprint. With these sensors the user must place a precise portion of his fingerprint on the small sensor in order to get a proper reading. This requirement for precise positioning of the

fingerprint makes the sensor harder to use in unattended installations. A further problem with smaller chips is that the amount of data used for matching part of a fingerprint is less than for a full print, making the security that the system provides less robust overall. Capacitive sensors are also susceptible to noise, including noise from the 60Hz power line noise picked up by the user and electrical noise from within the sensor. And a final problem with capacitive sensors is reliability. Electrostatic discharge, salt from sweat and other contaminants, and physical wear are hard on a semiconductor-sensing chip. Capacitive sensors have difficulty reading dry fingerprints. DigitalPersona believes that the optical sensor currently provides a more reliable solution. DigitalPersona’s U.are.U product provides the superior performance of an optical solution with the small size and low cost of the chipbased or capacitive solution. Biometric Solutions to Personal Identification

Fingerprint Identification Strengths

ü Fingerprints are unique, and they are complex enough to provide a robust template for identification verification.

ü Should we want increased levels of security we can easily register and require verification of one, two, or more fingerprints, up to ten prints. Each of our fingerprints is unique.

ü Scanning a fingerprint is quick and extremely convenient for consumers.

ü Users must interact with the fingerprint scanner, placing their finger on the scanner to obtain a biometric reading. Direct interaction between the user and the sensor is the most accurate way to obtain a biometric reading, and it is one of the main reasons why fingerprint scanning can be applied to mass markets.

ü Fingerprint scanners can be easily miniaturized and they can be mass- produced at low cost.

Fingerprint Identification Weaknesses

ü Some people and groups have relatively poor quality fingerprints that are difficult to image. DigitalPersona’s software has been optimized to deal with poor quality fingerprints.

ü The use of fingerprints in law enforcement leads some consumers to fear “having their fingerprint on file.” DigitalPersona reassures consumers that it never stores an image of their fingerprint. Rather the information from the print is stored as a small, encrypted numeric file.

ü Latent prints are afterimages that are left on the sensor after each use. Potentially such latent fingerprints could be used to recreate a fingerprint for identification. DigitalPersona has proprietary technology that removes latent fingerprints.

The Importance of Software and Firmware in Fingerprint Identification

Although the cost, size and design of the hardware sensor is important in fingerprint identification, it is only one determiner of whether a biometric system is completely robust or not. The firmware and the recognition algorithms are at least as important as hardware in creating a total solution, especially one that is ready for mass-market employment. The firmware resides in the hardware sensor and coordinates the capture of the image and its connection to the PC. In most fingerprint systems the firmware is simple and relatively ineffective. It dumps a continuous stream of image data to the host computer. However, in doing so it creates a number of problems. One major problem with this overly simple firmware is that the image data stream, either digital or analog, can be recorded and played back later. This opens the entire identification system to a "replay attack," and compromises security. further problem is that the sensor's power must always be on full, and the computer must be continuously capturing and processing the image stream to determine whether there is a fingerprint present. If there is, it must capture a single fingerprint image at the optimal time. Lastly, simple firmware cannot dynamically adjust the system to deal with a dirty sensor surface or "latent" fingerprint images resulting from the last use. DigitalPersona has addressed these issues in developing the firmware for its U.are.U Fingerprint Sensor, and this makes a difference in the overall effectiveness of the system. The DigitalPersona proprietary firmware handles the USB connection, which provides the power, bandwidth, and plug and play ease of use. It continuously processes within the Sensor to determine when the environment has changed, such as when there is dirt on the Sensor surface, when a fingerprint is present on the Sensor, and when to grab the optimal image. The firmware cannot be too quick to grab the image or the entire fingerprint area will not have been placed on the Sensor, nor can it be too late and risk the user pressing too hard. Once the image is captured the firmware sets up a 128-bit challenge-response encryption link with the host PC or server, which it uses to transmit the image in a secure manner. In doing so it also alerts the PC or server with an interrupt that a fingerprint has been captured. Afterward, the Sensor reverts back to a lowpower mode to await the next finger tap -- an important feature for laptops. Once the host computer has securely obtained the fingerprint image from the Sensor, the recognition algorithm must take over to perform the verification or identification. Fingerprints are a very robust biometric. They contain so much unique information that only a small portion of the total print is needed for accurate identification. This fact is not apparent with most fingerprint recognition systems. Most systems require that users place their full finger on the sensor, and that they do so carefully so that the fingerprint is aligned with crosshairs that are visible on the screen. If the finger placement is wrong, or if the print quality is not good, the user must retry until correct. Such a requirement makes these sensors less effective for mass-market adoption. From the beginning, DigitalPersona's commitment to ease of use and reliability has grown from the robustness of the company’s core recognition algorithm. The user does not need to worry about finger placement when using the U.are.U system. The algorithm is completely rotation and distortion invariant. The user taps the finger on the Sensor, without having to worry about precisely placing it on the platen. Finger rotation and pressure, fingerprint quality, and the presence or absence of dirt and moisture has no adverse effect on the system.

Attended and Non-Attended Fingerprint Capture - One-to-One and One-to-Many Verification

With fingerprint identification it is important to understand the difference between technologies that are based on attended or supervised capture of the print versus unattended capture, as well as one-to-one identification versus one-tomany verification schemas. In traditional law enforcement a fingerprint is given under supervision and a match is established by comparing that print to a database of many prints. For fingerprint recognition to be applied to a mass-market, the capture of the print must be an unattended process in which an individual gives a print quickly without an attendant lining up the print and ensuring a perfect capture of the data. Further, in the case of identification for entry or electronic transaction, the data that is captured is compared to a template that has previously been registered as the valid print. In most instances, the matching is a one-to-one match rather than a one-to-many match. Unattended fingerprint capture and one-to-one or one-to-many matching place different constraints on the DigitalPersona fingerprint recognition system than exist for the traditional law enforcement fingerprint capture and matching technologies. The unattended model requires that the sensor be much more tolerant of poor quality prints and that the software algorithms be much more robust than those used with attended models.

Understanding Iris Recognition

Iris scan biometrics employs the unique characteristics and features of the human iris in order to verify the identity of an individual. The iris is the area of the eye where the pigmented or coloured circle, usually brown or blue, rings the dark pupil of the eye.

Eyeglasses and contact lenses present no problems to the quality of the image and the iris-scan systems test for a live eye by checking for the normal continuous fluctuation in pupil size.

The inner edge of the iris is located by an iris-scan algorithm which maps the iris’ distinct patterns and characteristics. An algorithm is a series of directives that tell a biometric system how to interpret a specific problem. Algorithms have a number of steps and are used by the biometric system to determine if a biometric sample and record is a match. The Daugman algorithms for iris recognition earned the British Computer Society's 1997 IT Award and Medal.

The false acceptance rate for iris recognition systems is 1 in 1.2 million, statistically better than the average fingerprint recognition system. The real benefit is in the false-rejection rate, a measure of authenticated users who are rejected. Fingerprint scanners have a 3 percent false-rejection rate, whereas iris scanning systems boast ratees at the 0 percent level.

Applications

Iris-scan technology has been piloted in ATM environments, the customer’s iris data became the verification tool for access to the bank account, thereby eliminating the need for the customer to enter a PIN number or password. When the customer presented their eyeball to the ATM machine and the identity verification was positive, access was allowed to the bank account. Airports have begun to use iris-scanning for such diverse functions as employee identification/verification for movement through secure areas and allowing registered frequent airline passengers. A system that enables fast and easy identity verification in order to expedite their path through passport control.

The major applications of this technology so far have been: substituting for passports (automated international border crossing); aviation security and controlling access to restricted areas at airports; database access and computer login.

Premises access control; hospital settings including mother-infant pairing in maternity wards; "watch list" screening at border crossings; and it is under consideration for biometrically enabled National Identity Cards.

Iris recognition is forecast to play a role in a wide range of other applications in which a person's identity must be established or confirmed. These include electronic commerce, information security, entitlements authorisation, building entry, automobile ignition, forensic and police applications, network access and computer applications.

Other applications include monitoring prison transfers and releases, as well as projects designed to authenticate on-line purchasing, on-line banking, on-line voting and on-line stock trading to name just a few. Iris-scan offers a high level of user security, privacy and general peace of mind for the consumer.

General issues and concerns

The biggest concern is the fact that once a fingerprint or other biometric source has been compromised it is compromised for life, because users can never change their fingerprints.

A theoretical example is a debit card with a personal Identification Number (PIN) or a biometric. Some argue that if a person's biometric data is stolen it might allow someone else to access personal information or financial accounts, in which case the damage could be irreversible

With careful matching of tested biometric technologies to the particular use that is intended, biometrics provide a strong form of authentication that effectively serves a wide range of commercial and government applications.

Current projects

· Efficient iris recognition by characterizing key local variations

. This paper describes an efficient algorithm for iris recognition by characterizing key local variations. The basic idea is that local sharp variation points, denoting the appearing or vanishing of an important image structure, are utilized to represent the characteristics of the iris. The basic idea is that local sharp variation points, denoting the appearing or vanishing of an important image structure, are utilized to represent the characteristics of the iris. The whole procedure of feature extraction includes two steps: 1) a set of one-dimensional intensity signals is constructed to effectively characterize the most important information of the original two-dimensional image; 2) using a particular class of wavelets, a position sequence of local sharp variation points in such signals is recorded as features. We also present a fast matching scheme based on exclusive OR operation to compute the similarity between a pair of position sequences. Experimental results on 2 255 iris images show that the performance of the proposed method is encouraging and comparable to the best iris recognition algorithm found in the current literature.

· An effective and fast iris recognition system based on a combined multiscale feature extraction technique

The randomness of iris pattern makes it one of the most reliable biometric traits. On the other hand, the complex iris image structure and the various sources of intra-class variations result in the difficulty of iris representation. Although, a number of iris recognition methods have been proposed, it has been found that several accurate iris recognition algorithms use multiscale techniques, which provide a well-suited representation for iris recognition. In this paper and after a thorough analysis and summarization, a multiscale edge detection approach has been employed as a pre-processing step to efficiently localize the iris followed by a new feature extraction technique which is based on a combination of some multiscale feature extraction techniques. This combination uses special Gabor filters and wavelet maxima components. Finally, a promising feature vector representation using moment invariants is proposed. This has resulted in a compact and efficient feature vector. In addition, a fast matching scheme based on exclusive OR operation to compute bits similarity is proposed where the result experimentation was carryout out using CASIA database. The experimental results have shown that the proposed system yields attractive performances and could be used for personal identification in an efficient and effective manner and comparable to the best iris recognition algorithm found in the current literature.

Conclusion

Although Biometric characteristics lack the distinctiveness and permanence to identify an individual uniquely and reliably, they provide some evidence about the user identity that could be exploited to our advantage. Experiments show that the recognition performance of a biometric system can be improved significantly by making use of additional soft biometric user information like gender, ethnicity, and height. . In this paper, we study techniques to automatically select and update templates by considering system performance along with storage and computational overheads associated with multiple templates.