Biometric Methods A Secure Survey

B.Prasanalakshmi, Dr.A.Kannamal
AbstractBiometrics involves recognizing individuals based
on the features derived from their Physiological and or behavioral characteristics . Biometric systems provide reliable recognition schemes to determine or confirm the individual identity. Applications of these systems include computer systems security, e-banking, credit card, access to buildings in a secure way. Here the person or object itself is a password. User verification systems that use a single Biometric indicator is disturbed by noisy data, restricted degrees of freedom and error rates. Improving the performance of Unimodal biometric systems now becomes problematic. Multibiometric systems tries to overcome these drawbacks by providing multiple evidences to the same identity hence the performance may be increased. Furthermore intruders trying to hack the biometric identity finds it difficult to work on Multibiometric systems. However an effective fusion scheme is necessary to combine multiple information of an identity. This paper addresses a brief overview of biometric methods both Unimodal and Multimodal as well as their advantages and disadvantages are to be presented.

Keywords
Biometrics , Multimodal Biometrics , Identification , Security

B.Biometrics Components. A simple biometric system has four important components. 1) Sensor module which acquires the biometric data of an individual. For example a Fingerprint sensor captures fingerprint impression of a user. 2) Feature Extraction module where the acquired data is processed to extract feature values. For example , the position and orientation of minutiae points in a fingerprint image would be extracted in the feature extraction module of a fingerprint system. 3) Matching module in which feature values are compared against those in the template by generating a matching score . For example , in this module , the number of matching minutae points between the query and the template will be computed and treated as a matching score. 4) Decision making module in which the users identity is established or a claimed identity is either accepted or rejected based on the matching score generated in the matching module C.Identification Vs Verification Identification involves comparing the acquired biometric information against templates corresponding to all users in the database ,while verification involves comparison with only those templates corresponding to claimed identity. Thus identification and verification are two distinct problems having their own inherent complexities. D.Performance analysis.. 1) The performance of a Biometric system can be measured by reporting its False Accept Rate(FAR) and False Reject Rate(FRR) at various thresholds. These two factors are brought together in a receiver operating characteristic (ROC) curve that plots the FRR against the FAR at different thresholds(Alternately, the genuine accept rate, which equals 1-FRR, may be plotted against the FAR), the FAR and FRR are computed by generating all possible genuine and impostor matching scores and then setting a threshold for deciding whether to accept or reject a match. A genuine matching score is obtained when two feature vectors corresponding to the same individual are compared , and an impostor matching score is obtained when feature vectors from two different individuals are compared. E.Biometrics acceptance requirement Any human physiological or behavioral trait serve as a Biometric characteristic as long as it satisfies the following requirements: 1) Universality . Everyone should have it. 2) Distinctiveness . No two should be same.

I.INTRODUCTION IOMETRICS is the word derived from the Greek words bio(life) and metrics(to measure). Biometry refers to the automatic identification of a person based on his or her physiological or behavioral characteristics. This identification is preferred over traditional methods involving passwords and PINS. By replacing PINs , biometric techniques can potentially prevent unauthorized access to or fraudulent use of ATM, cellular phone, smart card, PCs , Workstations, computer networks etc., All Biometric systems are used for real time identification. Human Biometric characteristics are unique. So it can hardly be duplicated. These systems rely on the evidence of Fingerprints, hand geometry , iris, retina, face, hand vein , DNA etc., to either validate or determine an identity.

II.BIOMETRIC SYSTEM A.General A Biometric system is essentially a pattern recognition system that makes a personal identification by determining the authenticity of a specific physiological or behavioral characteristics possessed by the user .Depending on the context the biometric system can be either a verification (authentication ) or an identification system.

B.Prasanalakshmi is with SCSVMV University, Kanchipuram,India Phone:9994050613; e-mail:bplcse@gmail.com Dr.A.Kannamal is now a Asst.Professor in Department of CA at CIT,Coimbatore,India.

Electronic copy available at: http://ssrn.com/abstract=1493905

3) Performance . It should be invariant over a given period of time. 4) Collectability. III.BIOMETRIC SECURITY AN OVERVIEW Since there are number of Biometric methods in use, a brief overview of various biometric characteristics will be given , starting with newer technologies and then progressing to the older ones. Infrared thermo gram (facial, hand, hand vein) The pattern of heat radiated by the human body is captured with an infrared camera. This pattern id is unique for each person and is a non-invasive method. Some drawbacks existing in it are the cost of infrared sensors, the heat generated by other heat-emanating surface near the body sample. Also three-dimensional space and hand orientation are to be considered for further development. Gait Gait recognition is a technology which recognizes individuals by their personal, idiosyncratic manner in which humans walk Gait recognition is most useful for surveillance because it can be captured without the consent of the person being observed. It is also very difficult to hide and to fake. Some factors that might affect gait are stimulants like alcohol and drugs which makes a person unbalanced, physical changes due to pregnancy, accident, after weight gain or weight loss, a persons mood, and the clothes worn by them. Gait recognition can be done using normal surveillance cameras. The peculiar way of walk sustains for every human and it is a spatio-temporal biometrics. This behavioral biometric is a low security identity. The video sequence is considered as an acquisition device for this. The drawback here is that the walking style of a person may not remain the same over a long period of time due to change in body weight or serious brain damage. Some of the advantages are that it can be done from a distance and low cost surveillance cameras can be used. The disadvantages are that it can be easily faked and be affected by the background, clothes and feelings of the person. Keystroke recognition Keystroke Recognition is based on the fact that the manner in which we type our computer keyboard varies from person to person. This technique consists of a keypad or a key board that is connected to a computer. The user is asked to type a set words at different periods of time. Certain conditions need to be met such as one has to type without corrections; in case of a mistype one he has to start again, etc. Some the feature that are extracted are the way a person types, the cumulative typing speed, time that elapses between consecutive strokes, time that each key is held down, frequency of the individual in using other keys on the keyboard, such as the number pad or function keys and the sequence that is utilized when typing a capital letter for example, and whether the individual release the shift key or the letter key first. Using these features a template is created forming a statistical profile of the individuals behavioral characteristics. The advantages of these methods is that they are purely software based, do not require additional hardware, can be integrated with other biometrics, and minimal training is required. Another advantage is that if the template is created for a specific

word and the template is tampered, then the word can be changed and a new template can be formed. The disadvantage is that the method does not ease the burden of remembering the passwords. In addition the technology is in its very early stage and has not been tested on a wide scale. Odor Each object spreads around an odor that is characteristic of its chemical composition .This would be done with an array of chemical sensors, each sensitive to a certain group of compounds. The drawback here is Deodorants and perfumes makes the human odor less distinctive. Ear The shape of ear and the structure of cartilaginous tissue of the pinna are distinctive. Matching the distance of salient points on the pinna from a landmark location of the ear id the suggested method of recognition in this case. This method is not believed to be distinctive. Hand Geometry This recognition system is a fairly simple procedure but is very accurate . The user places his hand on a metal surface, which has guidance pegs on it, to help the user in aligning his hand properly. The device reads about 90 distinct features of the hand such as the length, width, thickness, surface area of the finger along with the palm size. These features are used to form a template and are stored in the database or can be verified against other template stored into the database. One of the main advantages of hand geometry technology is that it has one of the smallest templates in the biometrics field, generally under ten bytes. And also has a high user acceptance and it is non intrusive. The disadvantage of this method is that it has low accuracy and it has a relatively large reader. Some people like children, people with arthritis, missing finger, or large hands might find it difficult to enroll. It is used only for the verification process as of now. Fingerprint Recognition This recognition process takes an image of the fingerprint either using ink or a digital scanner and records the characteristics such as whorls, arches, loops, patterns of ridges, furrows and minutiae. This information is then stored or processed as an encoded algorithm. There are software programs that can be used to map the minutiae points to their relative placements on the finger and then search for similar minutiae information in the database. To save time during the search process, the image is converted into a character string. Therefore, most of the time the image of the fingerprint is never created, just a string of characters that can be used for comparison. The finger print reader requires the person to leave his finger on the reader for just a few seconds, during which time it can identify or verify a person. To make sure that a fake finger is not used, these days the fingerprint reader also checks for blood flow or correctly arrayed ridges at the edges of the fingers. Some of the advantages of this process is that it is a mature technology and user friendly. This method has a very high accuracy, long term stability and supports the option of enrolling multiple fingers to increase the anti-spoofing property. Some of the disadvantages of this method is that the finger print reader might get dirty (because of the

Electronic copy available at: http://ssrn.com/abstract=1493905

contact with the finger), which might affect the quality of the image, or the registered data may vary depending on the skin conditions. Facial Recognition This recognition process measures the overall facial structure such as distance between the eyes, nose, mouth and jaw edges .Using the features, a template is formed and stored in the database. During the verification or identification process the, a person image is obtained using a normal camera or a video camera and a template is formed using the facial features. This template is then compared with the one stored in the database. This system of recognition is currently used only in verification process with much success. To prevent people from faking the system, nowadays the system requires the user to smile, blink or move in a way that is human before verifying. The advantage of this system is that it is non intrusive; it can be operated at a low cost because the normal security cameras can be used to extract images. In addition, in places such as airports where a high level of security is required, the method can operate covertly; i.e. the image is taken without the knowledge of the individual). Facial recognition has a disadvantage of being highly dependent on the quality of the image obtained, which might be affected by appearance and the environment. This system is generally has low accuracy and might cause problems in case of identical twins. Furthermore, due to its ability to work in a covertly fashion, it has the potential of privacy abuse. Iris Recognition This Technology recognizes individuals by analyzing the features that exist in the colored tissue surrounding the pupil which includes rings, furrows and freckles. This is one of the most accurate biometric technologies that are available. In a recent article in Silicon.com, it was stated that in an experiment conducted by the government, which involved facial recognition, iris recognition and fingerprint recognition, it was found that iris recognition was the best among the three for verification process, though its enrollment process was difficult. The iris image can be obtained using a regular video camera and it can be done from further away than that of a retinal scan. In this technology it is very important that the user cooperates to get a clear image. So the device is such that when the user places his head in front of the device, he would be able to see the reflection of his iris in the device, which shows that a clear image can be obtained. The device may vary the light that is shone into the eye and observe the pupil dilate to make sure that system is not fooled by some fake eye. The advantage of this system is that it has a good verification rate and resistance to imposter. Iris data is stable with age, and identical for left and right eyes. The disadvantage of this method is that it is highly intrusive; the enrollment process is somewhat difficult because not all people would be comfortable with the system. Retinal recognition This method examines the blood vessel patterns of the retina which are supposed to be unique for every individual. Basically retina consists of sensory tissue with multiple layers, along with photoreceptors like cones and rods which gather the light rays that are sent to it and transforms it into electrical impulses which are in turn converted by the brain into images . The blood vessel in the retina can observe and

reflect IR light better than the surrounding tissues in the eye. So the device uses IR light to illuminate the retina and when the IR is reflected back, the retinal scanning device uses it to extract unique features of the retina using different algorithms. The size of the pupil, which determines the amount of light that enters the retina, determines the quality of the image. The advantages of this technology are that the retina pattern doesnt change over the life time of a person. Compared to other technologies, it has a good verification rate and it consists of rich unique features. The disadvantage of this method is that there is a public perception that the device might harm the retina, uneasiness of the user when the eye is scanned at a very close distance, the need for the individual to take off glasses on contact lenses, and the requirement of much patience and cooperation by the user. Signature recognition Signature Recognition examines the behavioral aspects in which we sign our name. This technology is based on the behavioral characteristics like change in timing ,pressure, speed, overall size and various directions of strokes during the course of the signing. Even though duplicating signatures seems easy visually, it is not easy to duplicate the behavioral characteristics. The device consists of a pen (stylus) and specialized writing tablet which is connected to a local computer. The user has to sign on the tablet using the pen. This system collects all the information about the features and forms a template, which is then stored in a database. The user has to sign multiple times to make sure a proper enrollment is done. The advantage of this technique is that it is a noninvasive tool, widely accepted and difficult to mimic. The disadvantage is that the signature should not be too long or too short because if it is too long then it might contain too many behavioral data and it will be difficult for the system to identify the consistent and unique data points. If the signature is too short, there might not be enough data point for the system to develop a unique template. It should be made sure that the enrollment process is done under the same environmental condition such as standing up, sitting down or resting ones arm. The system is also difficult for the user to get acclimated, which leads to signature inconsistency. Voice recognition Voice originates from the vocal cords. The gap in the vocal cords contracts and expands as we attempt to communicate. So as the gap narrows when we exhale and widens when the breath passes through, unique sounds are created. Vocal tract basically consists of the laryngeal pharynx, oral pharynx, oral cavity, nasal pharynx and the nasal cavity. In this technique the user is asked to recite a part of text or a list of numbers, using a microphone or any telephone connected to a computer. The computer records the users voice and converts it from analog to digital format. This format is stored and unique features are extracted from them to form a template. More than one sample is taken and comparing various samples and determining the various repeating patterns make a statistical profile. So comparing the statistical profiles does the verification. The advantage of this method is that the existing telephony infrastructure or simple microphone can be used; it is non intrusive and easy to use. The disadvantages are that the system is affected by background noise; it has very low accuracy and it might be affected by the change of voice due to aging, illness or drinking.

DNA recognition This technology uses the DNA pattern of a person to identify them. DNA was discovered in 1868. Since then people have learned how to us and manipulate DNA for various applications. DNA molecules are made of a long string of chemical building blocks called nucleotides. The sources of DNA are Blood, Semen, Tissues, Chemically treated Tissues, Hair Roots, Saliva, Urine, etc. DNA technology has been used extensively in Forensic filed and to establish paternity and other family relationships. Apart from this, it is also used to match organ donors with recipients in transplant programs; determine pedigree for seed, and improve crops. This technology has also been used to rapidly detect and identify new outbreaks of deadly diseases such as SARS. Advantages of DNA are that they can be found in almost every cell in the body, and the technique has an extremely high compared to other biometrics. The disadvantages are that the DNA samples are prone to degradation and contaminations from external sources, and the control and storage of DNA would have to deal with a lot of privacy acts. IV.UNIMODAL VS MULTIMODAL BIOMETRICS A.Problems With Unimodality The shortcoming of unimodal biometrics is that no one technology is suitable for all applications. Therefore, the presence of a multimodal biometric system helps compensate for the following limitations: The usage of certain biometrics makes it susceptible to noisy or bad data, such as inability of a scanner to read dirty fingerprints clearly. This can lead to inaccurate matching, as bad data may lead to a false rejection. Unimodal biometrics is also prone to inter-class similarities within large population groups. In case of identical twins, a facial recognition camera may not be able to distinguish between the two. Some biometric technologies are incompatible with a certain subset of the population. Elderly people and young children may have difficulty enrolling in a fingerprinting system, due to their faded prints or underdeveloped fingerprint ridges Finally, unimodal biometrics are vulnerable to spoofing, where the data can be imitated or forged.

additional biometric available, a greater number of people can be enrolled into the system Using multiple biometrics solves the problem of interclass similarity and the resultant high False Acceptance Rate (FAR). If people with similar hand sizes or similar looking facial features can gain false acceptance, the presence of another biometric such as signature verification can distinguish between the samples. Another advantage of using multimodality is that it solves the problem of data distortion. If the quality of one of the biometric samples is unacceptable, the other can make up for it. If a fingerprint has been scarred and the scanner rejects the distorted sample, having another modality like facial recognition can prevent high False Rejection Rates (FRR). Unimodal Biometrics can be easily spoofed. Placing a high-resolution picture of a fingerprint under the scanner can deceive some systems. However, by using multiple biometrics, even if one modality could be spoofed, the person would still have to be authenticated using the other biometric. Besides, the effort required for forging two or more biometrics is a deterrent to those who wish to do so. V.CONCLUSION We conclude that Biometrics plays a crucial role in the security system of the future. It can also be said that a unimodal biometric systems are not enough and would never fully serve the purpose that we are looking for. So it is best to look at other forms of biometrics such as Multibiometric and combination of Biometrics and technologies such as RFID and smartcards. Multimodal biometric systems elegantly address several problems present in Unimodal systems. By combining multiple sources of information , these systems improve matching performance, increase population coverage, determine spoofing and facilitate indexing. Incorporating user-specific parameters can improve the performance of these systems. with the widespread deployment of the biometric systems several civilian and government applications, it is only a matter of time before multimodal biometric systems begin to impact the way in which identity is established. REFERENCES
[1] [2] A. K. Jain, A. Ross, S. Prabhakar, "An Introduction to Biometric Recognition", IEEE Trans. on Circuits and Systems for Video Technology, Vol. 14, No. 1, pp 4-19, January 2004 A. K. Jain, A. Ross, "Multibiometric Systems", Appeared in Communication of the ACM, Special Issue on Multimodal Interfaces, Vol. 47, No.1, pp. 34-40, January 2004 A. Ross, A. K. Jain, "Information fusion in biometrics", Pattern Recognition Letters 24 (2003) 2115-2125, available at http://www.computerscienceweb.com/ Biometric Institute ,Types Of Biometrics, http://biometricsinstitute .org. D. Maio, D. Maltoni, R. Cappelli, J. L. Wayman, A. K. Jain, "FVC2002: Fingerprint verification competition" in Proc. Int. Conf. Pattern Recognition (ICPR), Quebec City, QC, Canada, August 2002, pp. 744-747 J. Daugman, "How Iris Recognition Works", IEEE Trans. on Circuits and Systems for Video Technology, Vol. 14, No. 1, pp. 21-30, January 2004 Joseph Rice , Vein Biometric Home Page www.ntlworld.com.

B.Advantages of Multimodality It is estimated that approximately 5 percent of any population has unreadable fingerprints, either due to scars or aging or illegible prints. In a civil ID scenario, where millions of people have to be entered in the system, the segment of the population who cannot enroll will face inconveniences. Having multimodal biometric technology can overcome this restriction and ensure lower Failure To Enroll Rate (FTE). Multimodality can also address the problem of aversion to fingerprinting, found in certain parts of the world. Sometimes people associate fingerprints with criminal activity, and are reluctant to submit prints. By having an

[3]
[4] [5]

[6]

[7]

[8]

Kevin.W.Bowyer , Multimodal Vs Unimodal Performance in Biometrics www.Biometrics.org. [9] L. Hong, A. K. Jain, S. Pankanti, "Can multibiometrics improve performance?," in Proc. AutoID'99, Summit, NJ, October 1999, pp. 59-64 [10] L. Hong, A. K. Jain, "Integrating faces and fingerprints for personal identification," IEEE Trans. Pattern Analysis Machine Intell., Vol. 20, pp. 1295-1307, December 1998 [11] L. I. Kuncheva, C. J. Whitaker, C. A. Shipp, R. P. W. Duin, "Is independence good for combining classifiers?," in Proc. Int. Conf. Pattern Recognition (ICPR), Vol. 2, Barcelona, Spain, 2001, pp. 168-171 [12] L. O'Gorman, "Seven issues with human authentication technologies," in Proc. Workshop Automatic Identification Advanced Technologies (AutoID), Tarrytown, NY, Mar. 2002, pp. 185-186 [13] M. Golfarelli, D. Maio, D. Maltoni, "On the error-reject tradeoff in biometric verification systems," IEEE Trans. Pattern Anal. Machine Intell., Vol. 19, pp. 786-796, July 1997 [14] P. J. Philips, P. Grother, R. J. Micheals, D. M. Blackburn, E. Tabassi, J. M. Bone, "FRVT2002: Overview and Summary," available at: http://www.frvt.org/FRVT2002/documents.htm [15] S. Prabhakar, S. Pankanti, A. K. Jain, "Biometric Recognition: Security and Privacy Concerns", IEEE Security & Privacy, March/April 2003, pp. 33-42 [16] Uwe M. Bubeck, Multibiometric Authentication B.Prasanalaskhmi is pursuing her Ph.D in Computer science from Bharathiyar university , has completed her M.E in Computer science and Engineering at Anna university (Chennai) by 2008 also completed her M.phil., in Computer science at Bharathidasan University in the year 2004 and her Master of Computer Applications in the year 2003 with a B.Sc., degree from the same university by year 2000.