• Share
  • Email
  • Embed
  • Like
  • Save
  • Private Content
Final Report Biometrics

Final Report Biometrics






Total Views
Views on SlideShare
Embed Views



0 Embeds 0

No embeds



Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.


11 of 1 previous next

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
Post Comment
Edit your comment

    Final Report Biometrics Final Report Biometrics Document Transcript

    • ACKNOWLEDGEMENT First and foremost I concede the surviving presence and the flourishingrefinement of ALMIGHTY GOD for His concealed hand yet substantialsupervision all through the seminar. I am extremely indebted to our respected principal Prof. Rajesh Mohan forrendering us all the facilities for the successful completion of our project. I express my heartiest thanks to our Head of the Department, seminarcoordinator and my seminar guide Mrs. Jaya V.L for the facilities and valuableguidance she provided in moulding this seminar to a successful one. I would also like to express my thanks to all other faculty members atCollege Of Engineering, Kottarakkara for the valuable help provided by them. Above all I would like to express my sincere gratitude and thanks to all myfriends for their valuable comments and suggestions for making this work a success.
    • Biometrics Seminar Report ’09 1. INTRODUCTION  Definition of Biometrics: Any automatically measurable, robust and distinctive physical characteristic orpersonal trait that can be used to identify an individual or verify the claimed identity of anindividual is called Biometrical Identification or simply Biometrics. It’s a combination of twoGreek words: Bios means Life and Metrics means To Measure.  History: Motivation behind invention of Biometrics is the basic need of a person beingauthenticated automatically and accurately. Authentication is the process of verifying that auser requesting a network resource is who he/she claims to be or not. Conventionalauthentication methods are based on two types:  Something that you “have” – e.g., Key, Magnetic Card or Smart card.  Something that you “know” – e.g., PIN or Password Biometric authentication method uses personal features i.e.  Something you “are”. Humans have used body characteristics such as face, voice, gait, etc. for thousandsof years to recognize each other. Alphonse Bertillon, chief of the criminal identificationdivision of the police department in Paris, developed and then practiced the idea of using anumber of body measurements to identify criminals in the mid 19th century. Just as his ideawas gaining popularity, it was obscured by a far more significant and practical discoveryof the distinctiveness of the human fingerprints in the late 19th century. Soon after thisdiscovery, many major law enforcement departments embraced the idea of first “booking”the fingerprints of criminals and storing it in a database (actually, a card file). Later, theleftover (typically, fragmentary) fingerprints (commonly referred to as latents) at the sceneof crime could be “lifted” and matched with fingerprints in the database to determine theidentity of the criminals. Although biometrics emerged from its extensive use in lawDept. of Electronics & Communication Engg. 1 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09enforcement to identify criminals (e.g., illegal aliens, security clearance for employees forsensitive jobs, fatherhood determination, forensics, positive identification of convicts andprisoners), it is being increasingly used today to establish person recognition in a largenumber of civilian applications.  Requirements of Biological Systems: Any human physiological and/or behavioral characteristic can be used as a biometriccharacteristic as long as it satisfies the following requirements:  Universality: each person should have the characteristic;  Distinctiveness: any two persons should be sufficiently different in terms of the characteristic;  Permanence: the characteristic should be sufficiently invariant (with respect to the matching criterion) over a period of time;  Collectability: the characteristic can be measured quantitatively.However, in a practical biometric system (i.e., a system that employs biometrics forpersonal recognition), there are a number of other issues that should be considered, including:  Performance, which refers to the achievable recognition accuracy and speed, the resources required to achieve the desired recognition accuracy and speed, as well as the operational and environmental factors that affect the accuracy and speed;  Acceptability, which indicates the extent to which people are willing to accept the use of a particular biometric identifier (characteristic) in their daily lives;  Circumvention, which reflects how easily the system can be fooled using fraudulent methods.Dept. of Electronics & Communication Engg. 2 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 2. BIOMETRIC SYSTEMS A biometric system is essentially a pattern recognition system that operates byacquiring biometric data from an individual, extracting a feature set from the acquired data,and comparing this feature set against the template set in the database. Depending on theapplication context, a biometric system may operate either in verification mode oridentification mode:  Verification Mode: In the verification mode, the system validates a person’s identity by comparing the captured biometric data with her own biometric template(s) stored system database. In such a system, an individual who desires to be recognized claims an identity, usually via a PIN (Personal Identification Number), a user name, a smart card, etc., and the system conducts a one-to- one comparison to determine whether the claim is true or not (e.g., “Does this biometric data belong to Bob?”). Identity verification is typically used for positive recognition, where the aim is to prevent multiple people from using the same identity.  Identification Mode: In the identification mode, the system recognizes an individual by searching the templates of all the users in the database for a match. Therefore, the system conducts a one-to-many comparison to establish an individual’s identity (or fails if the subject is not enrolled in the system database) without the subject having to claim an identity (e.g., “Whose biometric data is this?”). Identification is a critical component in negative recognition applications where the system establishes whether the person is who she (implicitly or explicitly) denies to be. The purpose of negative recognition is to prevent a single person from using multiple identities. Identification may also be used in positive recognition for convenience (the user is not required to claim an identity). While traditional methods of personal recognition such as passwords, PINs, keys, and tokens may work for positive recognition, negative recognition can only be established through biometrics.Dept. of Electronics & Communication Engg. 3 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 The block diagrams of a verification system and an identification system aredepicted in Figure 1; user enrollment, which is common to both the tasks is also graphicallyillustrated. Figure 1: Block diagrams of enrollment, verification and identification tasks are shown using the four main modules of a biometric system, i.e., sensor, feature extraction, matcher, and system database.
    • Biometrics Seminar Report ’09 The verification problem may be formally posed as follows: given an input featurevector XQ (extracted from the biometric data) and a claimed identity I, determine if (I, XQ)belongs to class w1 or w2, where w1 indicates that the claim is true (a genuine user) and w2indicates that the claim is false (an impostor). Typically, XQ is matched against XI, thebiometric template corresponding to user I, to determine its category. Thus, , , ≥ ( , )∈ , ℎwhere S is the function that measures the similarity between feature vectors XQ and XI, and t isa predefined threshold. The value S(XQ, XI) is termed as a similarity or matching scorebetween the biometric measurements of the user and the claimed identity. Therefore, everyclaimed identity is classified into w1 or w2 based on the variables XQ, I, XI and t, and thefunction S. Note that biometric measurements (e.g., fingerprints) of the same individual takenat different times are almost never identical. This is the reason for introducing the threshold t. The identification problem, on the other hand, may be stated as follows: given aninput feature vector XQ, determine the identity Ik, k ∈ {1, 2, ... , N, N +1}. Here I1 , I2 ,..., IN are theidentities enrolled in the system and IN+1 indicates the reject case where no suitable identitycan be determined for the user. Hence, max , ≥ , = 1, 2, … , ∈ ℎwhere , the biometric template corresponding to identity Ik, and t is a predefinedthreshold. Modules in Biometric Systems:A biometric system is designed using the following four main modules: i. Sensor module: It captures the biometric data of an individual. An example is a fingerprint sensor that images the ridge and valley structure of a user’s finger. ii. Feature extraction module: Here the acquired biometric data is processed to extract a set of salient or discriminatory features. For example, the position and orientation of minutiae points (local ridge and valley singularities) in a fingerprintDept. of Electronics & Communication Engg. 5 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 image are extracted in the feature extraction module of a fingerprint-based biometric system. iii. Matcher module: Here the features during recognition are compared against the stored templates to generate matching scores. For example, in the matching module of a fingerprint-based biometric system, the number of matching minutiae between the input and the template fingerprint images is determined and a matching score is reported. The matcher module also encapsulates a decision making module, in which a users claimed identity is confirmed (verification) or a user’s identity is established (identification) based on the matching score. iv. System database module: It is used by the biometric system to store the biometric templates of the enrolled users. The enrollment module is responsible for enrolling individuals into the biometric system database. During the enrollment phase, the biometric characteristic of an individual is first scanned by a biometric reader to produce a digital representation (feature values) of the characteristic. The data capture during the enrollment process may or may not be supervised by a human depending on the application. A quality check is generally performed to ensure that the acquired sample can be reliably processed by successive stages. In order to facilitate matching, the input digital representation is further processed by a feature extractor to generate a compact but expressive representation, called a template. Depending on the application, the template may be stored in the central database of the biometric system or be recorded on a smart card issued to the individual. Usually, multiple templates of an individual are stored to account for variations observed in the biometric trait and the templates in the database may be updated over time.Dept. of Electronics & Communication Engg. 6 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 3. BIOMETRIC SYSTEM ERRORS Two samples of the same biometric characteristic from the same person (e.g., twoimpressions of a user’s right index finger) are not exactly the same due to imperfect imagingconditions (e.g., sensor noise and dry fingers), changes in the user’s physiological orbehavioral characteristics (e.g., cuts and bruises on the finger), ambient conditions (e.g.,temperature and humidity) and user’s interaction with the sensor (e.g., finger placement). Therefore, the response of a biometric matching system is the matching score,S(XQ, XI) (typically a single number), that quantifies the similarity between the input andthe database template representations (XQ and XI, respectively). The higher the score, themore certain is the system that the two biometric measurements come from the sameperson. The system decision is regulated by the threshold, t: pairs of biometric samplesgenerating scores higher than or equal to t are inferred as mate pairs (i.e., belonging to thesame person); pairs of biometric samples generating scores lower than t are inferred asnon-mate pairs (i.e., belonging to different persons). The distribution of scores generatedfrom pairs of samples from the same person is called the genuine distribution and fromdifferent persons is called the impostor distribution (see Figure 2a). Figure 2(a): FMR and FNMR for a given threshold t are displayed over the genuine and impostor score distributions
    • Biometrics Seminar Report ’09 Figure 2(b): Choosing different operating points (t) results in different FMR and FNMR. The curve relating FMR to FNMR at different thresholds is referred to as Receiver Operating Characteristics (ROC). A biometric verification system makes two types of errors: i. False Match: mistaking biometric measurements from two different persons to be from the same person. ii. False Non-Match: mistaking two biometric measurements from the same person to be from two different persons. There is a trade-off between false match rate (FMR) and false non-match rate(FNMR) in every biometric system. In fact, both FMR and FNMR are functions of thesystem threshold t; if t is decreased to make the system more tolerant to input variationsand noise, then FMR increases. On the other hand, if t is raised to make the system moresecure, then FNMR increases accordingly. The system performance at all the operatingpoints (thresholds, t) can be depicted in the form of a Receiver Operating Characteristic(ROC) curve. A ROC curve is a plot of FMR against (1-FNMR) or FNMR for variousthreshold values, t (see Figure 2b). Mathematically, the errors in a verification system can be formulated as follows. Ifthe stored biometric template of the user I is represented by XI and the acquired input orrecognition is represented by XQ, then the null and alternate hypotheses are: H0: input XQ does not come from the same person as the template XI,
    • Biometrics Seminar Report ’09 H1: input XQ comes from the same person as the template XI. The associated decisions are as follows: D0: person is not who she claims to be; D1: person is who she claims to be. The decision rule is as follows: if the matching score S(XQ, XI) is less than thesystem threshold t, then decide D0, else decide D1. The above terminology is borrowedfrom communication theory, where the goal is to detect a message in the presence of noise.H0 is the hypothesis that the received signal is noise alone, and H1 is the hypothesis that thereceived signal is message plus the noise. Such a hypothesis testing formulation inherentlycontains two types of errors: Type I: false match (D1 is decided when H0 is true); Type II: false non-match (D0 is decided when H1 is true). FMR is the probability of type I error (also called significance level in hypothesis testing) and FNMR is the probability of type II error: FMR = P (D1| H0); FNMR = P (D0| H1). The expression (1-FNMR) is also called the power of the hypothesis test. Toevaluate the accuracy of a fingerprint biometric system, one must collect scores generatedfrom multiple images of the same finger (the distribution p(S (XQ, XI)|H1)), and scoresgenerated from a number of images from different fingers (the distribution p(S(XQ,XI)|H0)).Figure 2(a) graphically illustrates the computation of FMR and FNMR over genuine andimpostor distributions: = ∫ , | , = ∫ , | . The failure to capture (FTC) rate and the failure to enroll (FTE) rate are also used tosummarize the accuracy of a biometric system. The FTC rate is only applicable when thebiometric device has an automatic capture functionality implemented in it and denotes thepercentage of times the biometric device fails to capture a sample when the biometriccharacteristic is presented to it. This type of error typically occurs when the device is notable to locate a biometric signal of sufficient quality. The FTE rate denotes the percentage oftimes users are not able to enroll in the recognition system. There is a tradeoff between theDept. of Electronics & Communication Engg. 9 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09FTE rate and the perceived system accuracy (FMR and FNMR). FTE errors typically occurwhen the system rejects poor quality inputs during enrollment. Consequently, the databasecontains only good quality templates and the perceived system accuracy improves. Becauseof the interdependence among the failure rates and error rates, all these rates (i.e., FTE,FTC, FNMR, and FMR) constitute important specifications in a biometric system, andshould be reported during performance evaluation. The accuracy of a biometric system in theidentification mode can be inferred using the system accuracy in the verification modeunder simplifying assumptions.Dept. of Electronics & Communication Engg. 10 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 4. BIOMETRIC DEVICES A number of biometric characteristics exist and are in use in various applications.Each biometric has its strengths and weaknesses, and the choice depends on theapplication. No single biometric is expected to effectively meet the requirements of all theapplications. In other words, no biometric is “optimal”. The match between a specificbiometric and an application is determined depending upon the operational mode of theapplication and the properties of the biometric characteristic. Various Biometric Technologies that were designed till now are as given below:  DNA Fingerprinting  Vein Thermogram  Face Recognition  Fingerprint  Hand & Finger Geometry Recognition  Iris Scanning  Retina Scanning  Signature Verification  Voice Recognition  Ear Scanning  Gait Sequence Analyzing  Keystroke Analyzing  Odor Sensing  Palmprint Recognition Of these the last five viz: Ear Scanning, Gait Sequence Analyzing, KeystrokeAnalyzing, Odor Sensing, and Palmprint Recognition are not very effective.Dept. of Electronics & Communication Engg. 11 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 I. DNA FINGERPRINTING Deoxyribo Nucleic Acid (DNA) is the one-dimensional ultimate unique code forone’s individuality. Studies say that 99.9% of DNA patterns in every individual is the same.Only 0.1% of DNA pattern differs from individual to individual - except for the fact thatidentical twins have identical DNA patterns. But further studies say that the percentage ofoccurrence of identical twins during birth is just 5-10%. This fact opens the gate to theeffective use of DNA fingerprinting. Because when a large population is considered thechances of identical twins is very much negligible.  Advantages: a) There are strong similarities shown between genetic fingerprints of parents and children. This is a benefit because a childs genetic fingerprint is made up of half the fathers genetic information and half of the mothers information. This means that the bands of a childs genetic fingerprint will match the bands on both of their parents, making it possible to establish paternity and maternity tests. b) It is very much accurate. Since no two individuals have same DNA pattern the accuracy is very large.  Demerits: Three issues limit the utility of this biometrics for applications other than forensicapplications: i. Contamination and Sensitivity: It is easy to steal a piece of DNA from an unsuspecting subject that can be subsequently abused for an ulterior purpose; ii. Automatic Real-Time Recognition Issues: The present technology for DNA matching requires cumbersome chemical methods (wet processes) involving an expert’s skills and is not geared for on-line non-invasive recognition; iii. Privacy Issues: Information about susceptibilities of a person to certain diseases could be gained from the DNA pattern and there is a concern that the unintended abuse of genetic code information may result in discrimination, e.g., in hiring practices.Dept. of Electronics & Communication Engg. 12 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 II. VEIN THERMOGRAM Vein measurement generally focuses on blood vessels on the back of the hand. Aninfrared light combined with a special camera captures an image of the blood vessels in theform of tree patterns. Deoxidized Hemoglobin present in the veins absorbs the IR rays, andthese areas appear black in the image. This image is then converted into data and stored in atemplate. Vein patterns are large robust internal patterns. It is unique to every individual,even twins. Second, the procedure does not implicate the criminal connotations associatedwith the taking of fingerprints. The procedure has not yet won full mainstream acceptance.  Advantages:  This type of recognition can be used only if deoxidized hemoglobin is continuously flowing through the veins. So the person should be alive.  It has lack of proven reliability.  Weakness:  The person whose scan is being performed should stand still for a longer time, than usual fingerprint method or hand geometry recognition method.  Since this depends on the amount of IR rays reflected back to the camera, so the surrounding environments do have a large influence on this methodology. III. FACE RECOGNITION People recognize one another by looking at each others faces so it is no surprise thatbiometric technology can do the same. Face recognition is a non-invasive process where aportion of the subjects face is photographed and the resulting image is reduced to digitalcode. It is based on the location & shape of Facial Attributes like eyes, eye brows, nose,lips, etc. Even thermogram can be used. PC Week Lab, in its recent review of facerecognition biometric product, noted that it "affords nearly foolproof authentication;eliminates user-related security issues such as forgotten passwords or token theft; is fastand easy to use; and provides an audit trail of all access attempts."  Advantages:  The photograph of a person can be used from a distance. Even the subject mayDept. of Electronics & Communication Engg. 13 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 not be even known that his picture is being taken.  When thermogram is used, even the use of a mask will be detected.  Disadvantages:  The main disadvantage of face recognition is similar to problems of photographs: people who look alike can fool the scanners. There are many ways in which people can significantly alter their appearance and slight changes in facial hair are one way to fool the device.  Some systems have difficulty in maintaining high levels of performance as the database grows in size. So mostly used in identification purpose.  Different poses do affect the entire process. Because pictures of the same person taken from different positions will be different. IV. FINGERPRINT Humans have used fingerprints for personal identification for many centuries andthe matching accuracy using fingerprints has been shown to be very high. A fingerprint isthe pattern of ridges and valleys on the surface of a fingertip, the formation of which isdetermined during the first seven months of fetal development. Fingerprints of identicaltwins are different and so are the prints on each finger of the same person. Today, afingerprint scanner costs about US $20 when ordered in large quantities and the marginalcost of embedding a fingerprint-based biometric in a system (e.g., laptop computer) hasbecome affordable in a large number of applications. The accuracy of the currently availablefingerprint recognition systems is adequate for verification systems and small- to medium-scale identification systems involving a few hundred users. Multiple fingerprints of aperson provide additional information to allow for large-scale recognition involvingmillions of identities.  Advantages:  It is very easy to use.  The device is very cheap and portable.  The device used to recognize Fingerprints consumes less power.  Disadvantages:  The current fingerprint recognition systems require a large amount of computational resources, especially when operating in the identificationDept. of Electronics & Communication Engg. 14 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 mode.  Fingerprints of a small fraction of the population may be unsuitable for automatic identification because of genetic factors, aging, or environmental reasons. V. HAND AND FINGER GEOMETRY RECOGNITION Hand geometry recognition systems are based on a number of measurements takenfrom the human hand, including its shape, size of palm, and lengths and widths of thefingers. Commercial hand geometry-based verification systems have been installed inhundreds of locations around the world.  Advantages:  It is very simple and relatively easier to use.  The device is cheap.  Environmental factors such as dry weather or individual anomalies such as dry skin do not appear to have any negative effects on the verification accuracy of hand geometry-based systems.  Disadvantages:  The geometry of the hand is not known to be very distinctive and hand geometry-based recognition systems cannot be scaled up for systems requiring identification of an individual from a large population.  Hand geometry information may not be invariant during the growth period of children. In addition, an individuals jewelry (e.g., rings) or limitations in dexterity (e.g., from arthritis), may pose further challenges in extracting the correct hand geometry information.  The physical size of a hand geometry-based system is large, and it cannot be embedded in certain devices like laptops. VI. IRIS SCANNING Iris recognition uses the iris-the colored circle that surrounds the pupil-as thephysical characteristic to be measured. The iris contains many randomly distributedDept. of Electronics & Communication Engg. 15 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09immutable structures, which means that, like snowflakes, no two irises are ever same.Moreover, the iris does not change over time. The visual texture of the iris is formed duringfetal development and stabilizes during the first two years of life. An iris scanner willanalyze over 200 points of the iris, such as rings, furrows, corona, etc. Using standard videotechnology, its features can be quickly recorded from about nine inches away, thuseliminating the need for invasive physical contact. Software captures the identifyinginformation from the iris and stores it in a 256-byte code.  Advantages:  It is very easier to detect Artificial Irises.  The pattern of iris is not changed by glasses, contact lens, and even surgery.  Even twins have different patterns of Iris.  This is the most secure form of Biometric Methodology after Retinal Scan.  Disadvantages:  The devices used to scan iris are very expensive.  The user should remain still till the whole process is over. Due to all these reasons iris recognition technology is used only in high security regions VII. RETINA SCANNING Retinal scanning involves an electronic scan of the retina-the innermost layer of thewall of the eyeball. These blind vessels at the back of the Eye have unique pattern. Evenour two eyes have different retinal pattern. This pattern is used to recognize a person. Byemitting a beam of incandescent light that bounces off the persons retina and returns tothe scanner, a retinal scanning system quickly maps the eyes blood vessel pattern andrecords it into an easily retrievable digitized database. The eyes natural reflective andabsorption properties are used to map a specific portion of the retinal vascular structure. Adevice accomplishes this mapping, which uses a scan wheel/lens apparatus rotating at therate of six complete rotations per second to collect a total of 700 data points in the retinaduring each rotation. Once the data is collected, it is digitized and stored as a 96-bytetemplate.Dept. of Electronics & Communication Engg. 16 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09  Advantages:  Most accurate and secure form of Biometric Methodology.  Retina generally remains fairly stable through life.  Disadvantages:  Retinal scanning includes the need for fairly close physical contact with the scanning device. Entire scanning process is slow.  The fact that trauma to the eye and certain diseases can change the retinal vascular structure.  Retinal Scanning can reveal some medical conditions of the subject, e.g., Hypertension, Hypotension, etc. VIII. SIGNATURE VERIFICATION Signature recognition, or signature dynamics, uses computer technology to recordcomponents of an individuals signature such as pen/stylus speed, pressure, direction insignature, and other characteristics. The way a person signs her name is known to be acharacteristic of that individual. The key is to differentiate between the parts of thesignature that are habitual and those that vary with almost every signing. Althoughsignatures require contact with the writing instrument and an effort on the part of the user,they have been accepted in government, legal, and commercial transactions as a method ofverification and have examined signature recognition systems as a way to reduce fraud. Theidea is that to the extent a persons handwriting is unique; an on-line writing device at thecheckout line could instantly compare the customers signature to one in a database.  Advantages:  It is very easier and cheaper.  It does not consume too much time for entire process; just the time required for signing is taken for verification.  Disadvantages:  Signatures are a behavioral biometric that change over a period of time and are influenced by physical and emotional conditions of the signatories. Successive impressions of signatures of some people are significantlyDept. of Electronics & Communication Engg. 17 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 different.  Professional forgers may be able to reproduce signatures that fool the system. Due to these disadvantages, this type of Biometric Methodology is considered as themost insecure recognition method. IX. VOICE RECOGNITION Voice is a combination of physiological and behavioral biometrics. The features ofan individual’s voice are based on the shape and size of the appendages (e.g., vocal tracts,mouth, nasal cavities, and lips) that are used in the synthesis of the sound. Thesephysiological characteristics of human speech are invariant for an individual, but thebehavioral part of the speech of a person changes over time due to age, medical conditions(such as common cold), emotional state, etc. Voice is also not very distinctive and may not beappropriate for large-scale identification. There are two types of voice recognition systems.They are text dependent voice recognition systems and text independent voice recognitionsystems. A text-dependent voice recognition system is based on the utterance of a fixedpredetermined phrase. A text-independent voice recognition system recognizes the speakerindependent of what she speaks. A text-independent system is more difficult to design than atext-dependent. Voice recognition involves taking the acoustic signal of a persons voice andconverting it to a unique digital code, which can then be stored in a template.  Advantages:  The text-independent system offers more protection against fraud.  Disadvantages:  The quality of voice signals can be degraded by the microphone and the communication channel through which it is transmitted.  It is very much sensitive to noise. So the background noises as such reflect in the digital code of the acoustic signal.  Voice recognition does not have many points of dissimilarities. So this type of verification is not used in verification purpose. But it is used in the identification.Dept. of Electronics & Communication Engg. 18 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09  A fairly large byte code is required to store and data.  Peoples voice can change, due to biological factors like common cold, etc.  Recently a device has been invented, which when connected to the telephone or mobile can change the frequency of our voice by tuning the frequency. So using this device we can convert a male voice to female voice. So this affects telephone based applications. A brief comparison of the above biometric techniques based on seven factors isprovided in Table below. The applicability of a specific biometric technique depends heavilyon the requirements of the application domain. No single technique can out-perform all theothers in all operational environments. In this sense, each biometric technique is admissibleand there is no optimal biometric characteristic. For example, it is well known that both thefingerprint-based and iris- based techniques are more accurate than the voice-basedtechnique. However, in a tele-banking application, the voice-based technique may bepreferred since it can be integrated seamlessly into the existing telephone system. Circumvention Distinctiveness Collectability Acceptability Performance Universality Permanence Biometric identifier DNA Fingerprinting H H H L H L L Vein Thermogram M M M M M M L Face Recognition H L M H L H H Fingerprint M H H M H M M Hand geometry M M M H M M M Iris H H H M H L L Retina H H M L H L L Signature L L L H L H H Voice M L L M L H HDept. of Electronics & Communication Engg. 19 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 5. LIMITATIONS OF (UNIMODAL) BIOMETRIC SYSTEMS The successful installation of biometric systems in various civilian applications does not imply that biometrics is a fully solved problem. It is clear that there is plenty of scope for improvement in biometrics. Researchers are not only addressing issues related to reducing error rates, but they are also looking at ways to enhance the usability of biometric systems. Biometric systems that operate using any single biometric characteristic have thefollowing limitations: i. Noise in sensed data: The sensed data might be noisy or distorted. Afingerprint with a scar, or a voice altered by cold are examples of noisy data. Noisy datacould also be the result of defective or improperly maintained sensors (e.g., accumulationof dirt on a fingerprint sensor) or unfavorable ambient conditions (e.g., poor illuminationof a users face in a face recognition system). Noisy biometric data may be incorrectlymatched with templates in the database (see Figure 5) resulting in a user being incorrectlyrejected. ii. Intra-class variations: The biometric data acquired from an individualduring authentication may be very different from the data that was used to generate thetemplate during enrollment, thereby affecting the matching process. This variation istypically caused by a user who is incorrectly interacting with the sensor or when sensorcharacteristics are modified (e.g., by changing sensors - the sensor interoperability problem)during the verification phase. As another example, the varying psychological makeup of anindividual might result in vastly different behavioral traits at various time instances iii. Distinctiveness: While a biometric trait is expected to vary significantlyacross individuals, there may be large inter-class similarities in the feature sets used torepresent these traits. This limitation restricts the discriminability provided by the biometrictrait. Golfarelli et al have shown that the information content (number of distinguishablepatterns) in two of the most commonly used representations of hand geometry and face areonly of the order of 105 and 103, respectively. Thus, every biometric trait has sometheoretical upper bound in terms of its discrimination capability.Dept. of Electronics & Communication Engg. 20 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 iv. Non-universality: While every user is expected to possess the biometric traitbeing acquired, in reality it is possible for a subset of the users to not possess a particularbiometric. A fingerprint biometric system, for example, may be unable to extractfeatures from the fingerprints of certain individuals, due to the poor quality of the ridges.Thus, there is a failure to enroll (FTE) rate associated with using a single biometric trait.It has been empirically estimated that as much as 4% of the population may have poorquality fingerprint ridges that are difficult to image with the currently available fingerprintsensors and result in FTE errors. v. Spoof attacks: An impostor may attempt to spoof the biometric trait of alegitimate enrolled user in order to circumvent the system. This type of attack is especiallyrelevant when behavioral traits such as signature and voice are used. However, physical traitsare also susceptible to spoof attacks. For example, it has been demonstrated that it is possible(although difficult and cumbersome and requires the help of a legitimate user) to constructartificial fingers/fingerprints in a reasonable amount of time to circumvent a fingerprintverification system.Dept. of Electronics & Communication Engg. 21 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 6. MULTIMODAL BIOMETRIC SYSTEMS Some of the limitations imposed by unimodal biometric systems can be overcome byusing multiple biometric modalities. Such systems, known as multimodal biometricsystems, are expected to be more reliable due to the presence of multiple, independentpieces of evidence. These systems are also able to meet the stringent performancerequirements imposed by various applications. Multimodal biometric systems address theproblem of non-universality, since multiple traits ensure sufficient population coverage.Further, multimodal biometric systems provide anti-spoofing measures by making it difficultfor an intruder to simultaneously spoof the multiple biometric traits of a legitimate user. Byasking the user to present a random subset of biometric traits, the system ensures that a“live” user is indeed present at the point of data acquisition. Thus, a challenge-response typeof authentication can be facilitated using multimodal biometric systems.  Modes of Operation: A multimodal biometric system can operate in one of three different modes: serialmode, parallel mode, or hierarchical mode. In the serial mode of operation, the output ofone biometric trait is typically used to narrow down the number of possible identities beforethe next trait is used. This serves as an indexing scheme in an identification system. Forexample, a multimodal biometric system using face and fingerprints could first employface information to retrieve the top few matches, and then use fingerprint information toconverge onto a single identity. This is in contrast to a parallel mode of operation whereinformation from multiple traits is used simultaneously to perform recognition. Thisdifference is crucial. In the cascade operational mode, the various biometric characteristicsdo not have to be acquired simultaneously. Further, a decision could be arrived at withoutacquiring all the traits. This reduces the overall recognition time. In the hierarchical scheme,individual classifiers are combined in a treelike structure.Dept. of Electronics & Communication Engg. 22 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09  Levels of Fusion: Multimodal biometric systems integrate information presented by multiplebiometric indicators. The information can be consolidated at various levels. Figure 3illustrates the three levels of fusion when combining two (or more) biometric systems inparallel. These are: a) Fusion at the Feature Extraction Level: The data obtained from each biometricmodality is used to compute a feature vector. If the features extracted from one biometricindicator are (somewhat) independent of those extracted from the other, it is reasonable toconcatenate the two vectors into a single new vector, provided the features from differentbiometric indicators are in the same type of measurement scale. The new feature vector hasa higher dimensionality and represents a persons identity in a different (and hopefully, morediscriminating) feature space. Feature reduction techniques may be employed to extract asmall number of salient features from the larger set of features. System Feature database Biometric extraction snapshot Features module Fusion Matching Decision Class Module module module Label Feature Biometric extraction Features snapshot module Figure 3(a): Fusion at the Feature Extraction Level c) Fusion at the Matching Score (Confidence or Rank) Level: Each biometricmatcher provides a similarity score indicating the proximity of the input feature vector withthe template feature vector. These scores can be combined to assert the veracity of theclaimed identity. Techniques such as weighted averaging may be used to combine thematching scores reported by the multiple matchers.Dept. of Electronics & Communication Engg. 23 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 Feature Matching Biometric extraction module Confidence Level snapshot module System Fusion Decision Class database Module module Label Feature Biometric extraction snapshot Matching Confidence Level module module Figure 3(b): Fusion at the Matching Score (Confidence or Rank) Level d) Fusion at the Decision (Abstract Label) Level: Each biometric system makes itsown recognition decision based on its own feature vector. A majority vote scheme can beused to make the final recognition decision. Feature Matching Decision Biometric extraction module module Decision Level snapshot module System Fusion Class database Module Label Feature Biometric extraction snapshot Matching Decision Decision Level module module module Figure 3(c): Fusion at the Decision (Abstract Label) Level The integration at the feature extraction level assumes a strong interaction among theinput measurements and such schemes are referred to as tightly coupled integrations. Theloosely coupled integration, on the other hand, assumes very little or no interaction among theinputs and integration occurs at the output of relatively autonomous agents, each agentindependently assessing the input from its own perspective. It is generally believed that a combination scheme applied as early as possible in therecognition system is more effective. For example, an integration at the feature level typicallyresults in a better improvement than at the matching score level. This is because the featureDept. of Electronics & Communication Engg. 24 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09representation conveys the richest information compared to the matching score of a matcher,while the abstract labels contain the least amount of information about the decision beingmade. However, it is more difficult to perform a combination at the feature level because therelationship between the feature spaces of different biometric systems may not be known andthe feature representations may not be compatible. Further, the multimodal system may nothave access to the feature values of individual modalities because of their proprietary nature.In such cases, integrations at the matching score or decision levels are the only options.  What to Integrate??? Multimodal biometric systems can be designed to operate in one of the following fivescenarios: 1) Multiple sensors: In this type of scenario the information obtained fromdifferent sensors for the same biometric are combined. For example, optical, solid-state,and ultrasound based sensors are available to capture fingerprints. 2) Multiple Biometrics: Multiple biometric characteristics such as fingerprint andface are combined. These systems will necessarily contain more than one sensor witheach sensor sensing a different biometric characteristic. In a verification system, the multiplebiometrics are typically used to improve system accuracy, while in an identification systemthe matching speed can also be improved with a proper combination scheme (e.g., facematching which is typically fast but not very accurate can be used for retrieving the top Mmatches and then fingerprint matching which is slower but more accurate can be used formaking the final identification decision). 3) Multiple units of the same biometric: Fingerprints from two or more fingers of aperson may be combined, or one image each from the two irises of a person may becombined. 4) Multiple snapshots of the same biometric: More than one instance of the samebiometric is used for the enrollment and/or recognition. For example, multiple impressionsof the same finger, or multiple samples of the voice, or multiple images of the face may becombined. 5) Multiple representations and matching algorithms for the same biometric:This involves combining different approaches to feature extraction and matching of theDept. of Electronics & Communication Engg. 25 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09biometric characteristic. This could be used in two cases. Firstly, a verification or anidentification system can use such a combination scheme to make a recognition decision.Secondly, an identification system may use such a combination scheme for indexing. In scenario 1, multiple sensors are used to sense the same biometric identifier whilescenario 2 uses multiple sensors to sense different biometric identifiers. An example ofscenario 1 may be the use of multiple cameras mounted to capture different views of aperson’s face. An example of scenario 2 is the use of a camera for capturing face and anoptical sensor to capture a fingerprint. While scenario 1 combines moderately independentinformation, scenarios 2 and 3 combine independent (or weakly dependent) information andare expected to result in a much larger improvement in recognition accuracy. In scenario 4,only a single input may be acquired during recognition and matched with several storedtemplates acquired during the one-time enrollment process; alternatively, more dataacquisitions may be made at the time of recognition and used to consolidate the matchingagainst a single/multiple template. Scenario 5 combines different representation and matchingalgorithms to improve the recognition accuracy. In our opinion, scenarios 4 and 5 combinestrongly correlated measurements and are expected to result in a smaller improvement inrecognition accuracy than scenarios 2 and 3, but they are more cost effective than scenario2 and more convenient than scenario 3. Scenarios 4 and 5 do require more computationaland storage resources than a unimodal biometric system but in principle, different featureextractors and matchers can work in parallel. As a result, the overall response time of thesystem is limited by the slowest individual feature extractor and/or matcher. Finally, acombination of more than one of these scenarios may also be used.Dept. of Electronics & Communication Engg. 26 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 7. SOCIAL ACCEPTANCE & PRIVACY ISSUES Human factors dictate the success of a biometric-based identification system to alarge extent. The ease and comfort in interaction with a biometric system contribute to itsacceptance. For example, if a biometric system is able to measure the characteristic of anindividual without touching, such as those using face, voice, or iris, it may be perceived to bemore user-friendly and hygienic. Additionally, biometric technologies requiring very littlecooperation or participation from the users (e.g., face and face thermograms) may beperceived as being more convenient to users. On the other hand, biometric characteristicsthat do not require user participation can be captured without the knowledge of the user,and this is perceived as a threat to privacy by many individuals. The very process ofrecognition leaves behind trails of private information. The issue of privacy becomes more serious with biometric-based recognitionsystems because biometric characteristics may provide additional information about thebackground of an individual. For example, retinal patterns may provide medical informationabout diabetes or high blood pressure in an individual. A health insurance company mayuse this information in an unethical way for economic gains by denying benefits to aperson determined to be of high risk. More importantly, people fear that biometricidentifiers could be used for linking personal information across different systems ordatabases. On the positive side, biometrics can be used as one of the most effective means forprotecting individual privacy. In fact, biometrics ensures privacy by safeguarding identityand integrity. For example, if a person loses a credit card and an adversary finds it, then thecredit history of this person is compromised. But, if the credit card could be used onlywhen the user supplies h i s / her biometric characteristics, then the user is protected.Biometrics can also be used to limit access to personal information. For instance, abiometric-based patient information system can reliably ensure that access to medicalrecords is available only to the patient and authorized medical personnel. Nevertheless, manypeople are uneasy about the use of their personal biological characteristics in corporate orDept. of Electronics & Communication Engg. 27 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09government recognition systems. To alleviate these fears, companies and agencies thatoperate biometric systems have to assure the users of these systems that their biometricinformation remains private and is used only for the expressed purpose for which it wascollected. Legislation is necessary to ensure that such information remains private and thatits misuse is appropriately punished. Most of the commercial biometric systems available today do not store the sensedphysical characteristics in their original form but, they store a digital representation (atemplate) in an encrypted format. This serves two purposes. First, the actual physicalcharacteristic cannot be recovered from the digital template thus ensuring privacy andsecondly, the encryption ensures that only the designated application can use this template.Dept. of Electronics & Communication Engg. 28 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 8. APPLICATIONS There are many concerning potential biometric applications, some popular examplesbeing: i. Automated Teller Machine (ATM): Most of the leading banks have beenexperimenting with biometrics of ATM machines use and as general means of combiningcard fraud. Surprisingly, these experiments have rarely consisted of carefully integrateddevices into a common process, as could be achieved with certain biometric devices.Previous comments in this paper concerning user psychology come to mind here onewonder why we have not seen a more professional and carefully considered implementationfrom this sector. The banks will of course have a view concerning the level of fraud and costof combating it via technology solutions such as biometrics. They will also express concernabout potentially alienating customers with such as approach. However, it still surprisesmany in the biometric industry that the banks and financial institutions have so far failed toembrace this technology with any enthusiasm. ii. Workstation and Network Access: For a long time this was an area oftendiscussed but rarely implemented until recent developments aw the unit price of biometricdevices fall dramatically as well as several designs aimed squarely at this application. Inaddition, with household names such as Sony, Compaq, Samsung and others entering themarket, these devices appear almost as a standard computer peripheral. Many are viewingthis as the application, which will provide critical mass for biometric industry and create thetransition between sci-fi device to regular systems component, thus raising public awarenessand lowering resistance to the use of biometrics in general. iii. Travels and Tourism: There are many in this industry who have the vision ofa multi application card for travelers which, incorporating a biometric, would enable themto participate in various frequent flyer and border controls systems as well as paying fortheir air ticket, hotel rooms, hire care etc, all with one convenient token. Technically this iseminently possible, but from a political and commercial point of view there are many issuesto resolve, not the least being who would own the card, are responsible for administrationand so on. These may not be insurmountable problems and perhaps we may see somethingDept. of Electronics & Communication Engg. 29 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09along these lines emerge. A notable challenge in this respect would be packaging such aninitiative in a way that would be truly attractive for users. iv. Internet Transactions: Many immediately of think of on line transactions asbeing an obvious area for biometrics, although there are some significant issues to considerin this context. Assuming device cost could be brought down to level whereby a biometric(and perhaps chip card) reader could be easily incorporated into a standard build PC, westill have the problem of authenticated enrollment and template management, although thereare several approaches one could take to that. Of course, if your credit already incorporated abiometric this would simplify things considerably. It is interesting to note that certaindevice manufactures have collaborated with key encryption providers to provide anenhancement to their existing services. Perhaps we shall see some interestingdevelopments in this area in the near future. v. Telephone Transactions: No doubt many telesales and call center managershave pondered the use of biometrics. It is an attractive possibility to consider, especially forautomated processes. However, voice verification is a difficult area of biometrics,especially if one does not have direct control over the transducers, as indeed you wouldn’twhen dealing with the general public. The variability of telephone handsets coupled to thevariability of line quality and the variability of user environments presents a significantchallenge to voice verification technology, and that is before you even consider thevariability in understanding among users. Perhaps we shall see further developments, whichwill largely overcome these problems. Certainly there is a commercial incentive to do so andI have no doubt that much research is under way in this respect. vi. Public Identity Cards: A biometric incorporated into a multipurpose public IDcards would be useful in a number of scenarios if one could win public support for such ascheme. Unfortunately, in this country as in others there are huge numbers of individuals whodefinitely do not want to be identified. This ensures that any such proposal would quicklybecome a political hot potato and a nightmare for the minister concerned. But we expectdevelopments in this field in the near future. This will solve many border security problems,airport security, access to restricted buildings and sites.Dept. of Electronics & Communication Engg. 30 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 9. CONCLUSION Reliable personal recognition is critical to many business processes. Biometricsrefers to automatic recognition of an individual based on her behavioral and/orphysiological characteristics. The conventional knowledge-based and token-based methodsdo not really provide positive personal recognition because they rely on surrogaterepresentations of the person’s identity (e.g., exclusive knowledge or possession). It is, thus,obvious that any system assuring reliable personal recognition must necessarily involve abiometric component. This is not, however, to state that biometrics alone can deliverreliable personal recognition component. In fact, a sound system design will often entailincorporation of many biometric and non-biometric components (building blocks) to providereliable personal recognition. Biometric-based systems also have some limitations that may have adverseimplications for the security of a system. While some of the limitations of biometrics canbe overcome with the evolution of biometric technology and a careful system design, it isimportant to understand that foolproof personal recognition systems simply do not existand perhaps, never will. Security is a risk management strategy that identifies, controls,eliminates, or minimizes uncertain events that may adversely affect system resources andinformation assets. The security level of a system depends on the requirements (threatmodel) of an application and the cost-benefit analysis. In our opinion, properly implementedbiometric systems are effective deterrents to perpetrators. There are a number of privacy concerns raised about the use of biometrics. A soundtrade-off between security and privacy may be necessary; collectiveaccountability/acceptability standards can only be enforced through common legislation. As biometric technology matures, there will be an increasing interaction among themarket, technology, and the applications. This interaction will be influenced by the addedvalue of the technology, user acceptance, and the credibility of the service provider. It istoo early to predict where and how biometric technology would evolve and get embedded inwhich applications. But it is certain that biometric-based recognition will have a profoundinfluence on the way we conduct our daily business.Dept. of Electronics & Communication Engg. 31 College of Engineering, Kottarakkara
    • Biometrics Seminar Report ’09 10. REFERENCES  A. K. Jain and S. Pankanti, "A Touch of Money", IEEE Spectrum, pp. 22-27, July 2006  A. K. Jain, A. Ross and S. Prabhakar, "An Introduction to Biometric Recognition", IEEE Transactions on Circuits and Systems for Video Technology, Special Issue on Image- and Video-Based Biometrics, Vol. 14, No. 1, pp. 4-20, January 2004.  Anil K. Jain, Ruud Bolle and Sharath Pankanti, "Biometrics: Personal Identification in Networked Society", Kluwer Academic Pub; ISBN: 0792383451  www.biometrics.org  www.biometricnewsportal.com  www.biometricgroup.com  www.scribd.com/doc/12733359/Biometrics-Body-ID  www.scribd.com/doc/3834277/Biometrics  www.neurotechnology.com  www.slideshare.net/ajayspi/biometricsDept. of Electronics & Communication Engg. 32 College of Engineering, Kottarakkara