Biometric Authentication Systems - Essay Example

Insert Biometrics Advancement in technology and especially electronic technology has brought about great demand for accurate, fast and reliable user authentication and identification systems. Systems require a personal recognition scheme that ensures that services are only accessed by legitimate users. Systems and applications that require such authentication include computer systems, gaining access to buildings, personal computers, laptops, cellular phones and ATMs. Need for reliability has led to heightened concerns about security increased advancements in communication, mobility, and networking. Biometrics recognition is, therefore, the automatic recognition of users based on behavioral and physiological characteristics. Biometrics seeks to establish a users identity based on who she is rather what she possesses like an ID card or what she remembers like passwords. Biometrics gained popularity when body measurements were used by Alphonse Bertillon to identify criminals in the 19th century. It prompted law enforcer to take criminals fingerprints and store them in a database. The stored fingerprints could then be compared with fingerprints lifted from crime scenes. Biometrics are extensively used to identify criminals but has also gained use in employee clearance for sensitive jobs, determining fatherhood, forensics, identification of prisoners. The whole physical body cannot be used as a biometric. For the body part to be used as a biometric then, it has to satisfy several characteristics. Universality – each person in the world possesses the characteristics. Distinctiveness – any two people should be sufficiently different in terms of the characteristic of the body part. Permanence – the characteristic should be sufficiently invariant over a period. Collect-ability – the body characteristic can be,measured quantitatively. A practical biometric system should consider the following before it can be effected. Performance – refers to the achievable recognition speed and accuracy, the resources to be used in achieving the desired recognition accuracy and speed and the environmental and operational factors that affect the accuracy and speed. Acceptability – this is the extent to which users and people are willing to accept the use of the biometric body part in their daily lives. Circumvention – it refers to the ease with which the biometric system can be fraudulently accessed or altered. Types of Biometrics Different body parts are being used as a form of biometric recognition. Biometrics ranges from DNA matching, chemical biometric, visual biometric, eyes biometrics ( iris recognition, retina recognition ), finger geometry recognition, Hand geometry recognition, signature recognition and behavioral biometrics (typing recognition), vein recognition and gait. Chemical Biometrics Chemical biometrics involves the use of and measuring chemical cues such as the order and chemical composition of the human perspiration. DNA matching Deoxyribonucleic acid (DNA) matching is the process used in the identification of persons by their DNA profiles. DNA profiles are hidden sets of numbers that reflect an individuals DNA composition, which can also be used as a persons identifier. DNA profiling is different from a full genome sequencing that is used in parental testing and criminal investigation. Even though 99.9% of human DNA sequences are similar in every person, unless they are identical (monozygotic) twins. DNA matching uses repetitive sequences that are highly variable, known as variable number tandem repeats (VNTRs), particularly short tandem repeats (STRs). In 1984 Sir Alec Jeffreys started research work on the DNA profiling technique at the University of Leicester in England. Advantages They are highly accurate. They can be used to identify both dead and living human beings. Disadvantages It is very easy to steal a piece of DNA from a person, and the piece can be creating contamination and sensitivity issues. There are no real-time recognition technologies as DNA matching requires complex chemical methods and experts. DNA matching is not geared towards on-line noninvasive recognition. Information from DNA patterns of people can be abused resulting in user discriminations especially in workplaces. It creates privacy concerns. Visual Biometrics Visual biometrics uses body parts that can be viewed from the outside body. Advantages They are inexpensive to set up. The body parts used in visual biometrics are very distinct to a person even twins. Most body parts used do not change in an entire lifetime of a person that increases accuracy. Disadvantages Visual biometrics are intrusive and cause user rejections Cannot be effectively used for physically impaired people. People using glasses, broken limbs. Iris Recognition The idea that no two irises are alike, even in twins, was proposed by Dr. Leonard Flom and Dr. Aran Safir in the 1980s. The iris contains many features that are used to distinguish one iris from another. The trabecular meshwork tissue is the primary visible character of the iris and is permanently formed by the eighth month of gestation. The iris is a good biometric because; the iris does no undergo any genetic process ( chaotic morphogenesis), which means that identical twins cannot have same irises. The iris has 266 degrees of freedom that allow large variations thus making the iris distinct from another. The iris is also protected by the eyelid, cornea, and the aqueous humor, and this reduces the chances of damage. Finally, the iris is not subject to aging making it stable in the entire life of a human being. The process of iris recognition involves capturing an image using standard or infrared light, defining the location of the iris and optimizing the captured image. The optimized image is the stored for comparisons. Despite the physiological characteristics of the iris being the major advantage of using iris, iris scanning is also not intrusive to users body. The factors make the scalability and speed of iris-scanning technologies a big advantage. Iris recognition technologies are mostly used in the computing environments to secure personal computers and networks. It is also used in physical security to sensitive data centers by mounting eye scanners on their doors. Iris recognition has also been used in ATMs, border clearance, air travel, airport security and in the transportation sector. Retina Recognition Retina is still the most reliable form of biometric despite its lack of extensive use in most commercial applications. The retina is made up of multiple layers of sensory tissues and millions of photoreceptors that transform light rays into electrical impulses. The blood vessel pattern in the retina forms the foundation for retinal recognition . The retina is not usually or mostly exposed to the external environment due to its back position within the eye making it a very stable biometric. Retinal scanning process has three steps. Signal acquisition and processing ‐ it involves capturing an image of the retina and converting it to a digital format. Matching using a computer system to verify And identify the user. Lastly, the unique features of the retina are represented as a template. The strengths of using retina as a biometric method are; the blood vessel patterns in the retina rarely change in a lifetime, the template size created is only 96 bytes, many data points can be created from the blood vessels and finally the retina is not exposed to external environmental threats due to its location inside the eye. Despite the many strengths, the notable weaknesses of retinal recognition mostly arise from a users perspective. Users perceive retinal scanning as a health threat People also are very uneasy when positioning the eyes very close to the retina scanner. Retinal scanning devices are also very expensive to procure and do not accommodate people wearing glasses. Retina recognition has not been extensively used in the application due to its user invasiveness. Retina recognition is being used at high security facilities like military installations, nuclear facilities, and laboratories to monitor access controls. Face Recognition A facial biometric and recognition system is used to automatically verifying and identify a person from an image or a video. The images are then compared with selected facial features from the database image. There are two approaches to the face recognition problem: geometric that is based on the feature of the face and photometric that is a view based. Face recognition also uses eyebrows, lips, physical shape and locational position of the eyes, chin and nose. Face recognition also uses the overall study of the face image that represent an individuals face as a number of similar recognized faces to distinguish different faces. Facial recognition process has to detect a face in the image taken; the face is then normalized into standard terms of pose, size, and illumination. The features of the face are then extracted and then recognized and identified by comparing with other sources and a facial database. Facial recognition algorithms like principal component analysis (PCA), elastic bunch graph matching (EBGM) and linear discriminant analysis (LDA) are widely used to achieve facial recognition. Face recognition is an uncertain way of clearly verifying the identity of a person because; it is very hard to match face images taken from two different views and those that are under different illumination conditions . Another challenge to face recognition is that, people faces tend to change with time, and other people may alter their face intentionally. Finger Geometry Recognition Finger geometry involves his study of ridge patterns and Valleys on the fingertips. The first seven months of fetal development determine formation of ridge patterns. Fingerprints are effective because they are different, Finger geometry also involves vein recognition that recognizes patterns using images of human hand finger vein and veins found below the skin surface. The vein patterns are acquired by inserting fingers into a verification terminal that has infrared LED light and a monochrome CCD camera. Blood hemoglobin absorbs LED light. The vein system will be made to appear as a dark sequence of lines. Finger vein ID system are harder to manipulate because they only verify the finger of a life human being. Finger geometry recognition are not expensive to implement but face challenges from factors that alter the shape of fingers like genetic factors, aging, environmental, or occupational reasons (manual work ). Hand Geometry Recognition Hand geometry biometrics is based on several measurements taken from the human hand that includes size of palm, shape, width and length of the fingers. Hand geometry recognition is is very simple, relatively easy to use, and inexpensive leading to installations in many places for commercial use. The limitations are that they are not very distinctive and hand geometry-based Recognition systems cannot be used for systems requiring identification of an individual from a large group of people. Other challenges include the constant change of hand properties during the growth of a child, presence of jewelry can alter the properties and limitations in dexterity from diseases like arthritis. Behavioral Biometrics Behavioral biometrics is reliant on skills, style, liking, knowledge, vehicle-skills or different plans used by human beings while completing different every day tasks such as writing, driving an automobile, talking on the phone or using a computer. Advantages They can be collected without the knowledge of the user. They can be collected non-obstructively. It is cost effective because it does not require ant special technologies to collect. They show relatively high accuracy in identity verification. Disadvantages Behaviors of people can change due to emotional changes. People engaging into multiple activities or behaviors being studied might lead to altering the other behaviors leading to behavior distortion. Signature Recognition Signatures are part of behavioral biometric that involves the way human beings sign their names. The sign is then made the characteristic of that person. Signatures can be operated in several ways. Static is group called off-line where users write it on paper, the signature is then digitized using a camera or optical scanner, the biometric system then identifies the signature by studying its shape. Dynamic is an on-line mode where people write their signature in a digitizing tablet or stylus-operated PDAs, which acquires the signature in real time. Techniques used for recognition of the signatures are(DTW) Dynamic Time Warping, Hidden Markov Models (HMM) and Vector Quantization (VQ). Signatures have been adopted in legal, government and business transactions as a Verification method. Challenges to using signatures are that they change over a certain time due to physical and emotional conditions of people, some peoples signatures vary substantially even if written successively and finally, forgers may produce signatures that can fool the recognition systems. Typing Recognition Typing recognition is used in the verification and even sometimes identification of persons. It involves the study of keystrokes while typing using the touch typing method for experienced users and the hunt and peck method that uses two fingers. The features of interest in typing recognition include time spent between hitting different keystrokes, dwell times and strokes in inter-keys.Dwell time is the time a key remains down when it is pressed, total typing speed, rate of using backspace that studies frequency of errors, use of Numpad, the sequence in which user hits the shift key in order to capitalize letters. The force of hitting the keys can also be used for keyboards with special equipment (Ilonen 1999). Typing recognition has also been used to study finger motions of skilled pianists while playing a musical keyboard. Vein Recognition Vein recognition is the process of identifying a person based on behavioral traits and unique physical features. Vein recognition mostly studies the vein patterns in the human fingers. Vein recognition produces higher accuracy than other biometrics like fingerprints because vein patterns are impossible to alter or forge. Vein recognition has gained acceptance due to very low false acceptance and false rejection frequency rates, patterns of the vein only change when a body size changes, thus are relatively reliable and vein recognition devices can be used to read a variety of veins at specific finger or even at the back of the palm. It, therefore, gives a wide selection choice thus ensuring reliability. Gait Gait is the study of people walking patterns in order to acquire biometric information Such as identity, gender, ethnicity and age. The features analyzed in gait include arm swing, rhythm of the walker, bounce, length of steps, the distance ( vertical )between the foot and the head, the utmost total distance between the right foot the left leg and the distance between the head pelvis (Kalyanaraman, 2006) . Advantages Can recognize a user from a distance. It is effectively used in cases where there is low image resolution like those available in CCTV cameras. It is a non-invasive type of biometric. Disadvantages It is not effective in scenarios where users have worn clothes that affect the walking style of a person. Walking surfaces (downhill or uphill ) could alter the walking style of a person leading to wrong results. Conclusion Advancement of technologies today have made reliable biometric systems an important requirement in critical businesses. Biometric systems whether chemical, visual or behavioral offer reliability in user identification and minimizes time wastages. Despite the reliability of older identification systems like fingerprints, new biometric like DNA, iris recognition, body odor, vein and ear recognition are gaining acceptance. The newer technologies are highly accurate, reliable and not intrusive to the users. In order to implement a reliable biometric system, a business should consider the development of that is cost/effective, reliable, accurate and is not user intrusive. Works Cited P. J. Phillips, P. J. Flynn, T. Scruggs, K. W. Bowyer, J. Chang, K. Hoffman, J. Marques, J. Min, W. Worek, “Overview of the face recognition grand challenge,” Computer Vision and Pattern Recognition (CVPR), 1 (2005): 947-954. Detlef Borchers and Robert W. Smith, “Mixed reactions to the facial features recognition technology project of the BKA,” Heise Online, 16 July 2007. http://www.heise.de/english/newsticker/news/92722. F. Jean, A. B. Album, and R. Bergevin, “Towards view-invariant gait modeling: Computing view-normalized body part trajectories,” Pattern Recog., vol. 42, November 2009. Kalyanaraman, S (2006) ‘Biometric authentication systems: a report’, retrieved July 26, 2006 from http://netlab.cs.iitm.ernet.in/cs650/2006/TermPapers/sriramk.pdf. J. L. Wayman, A. K. Jain, D. Maltoni, and D. Maio, Eds., Biometric Systems: Technology, Design, and Performance Evaluation. New York: Springer-Verlag, 2005. Read More