Biometric systems and applications

Biometric may be described as the mechanized way of finding the personality of a living being depending on the person's behavior and physiology. While examining the description of biometrics, there are various definite provisions that should be explained to thoroughly know about biometric technology. The concept of 'mechanized way' refers to three fundamental modes of ascertaining identity by connecting to the biometric equipment. They are (1) the method to scan and obtain a digital or analog picture of the living being's feature (2) handing out and assessment of the picture with the existing information and (3) checking the details with the application structure. The word 'living' being is of utmost importance while describing biometrics. This means the word does not include a latex finger, prosthetic eye, plastered hand or a digital audio record. Biometric equipment records digital and analog picture to find out a live feature in a person. The word living distinguishes biometric from forensic verification though the fundamental techniques surpasses both fields. (Biometrics Technology)

Face Recognition

The certification of a person through biometric technology has gained momentum in the recent times due to the emerging needs in e-banking, e-commerce and the increased level of security assessment after the series of terrorist attacks around the world. Identification of the face has received much recognition in comparison to other methods of biometric certification due to its non- invasive and user friendly methods. The limitations of identification of a face include the picture received from an open surrounding, huge postures, etc. This makes identification of a face difficult precisely and steadily. Identification of a person through surveillance is a simple method. A photograph of an individual gives enough data compared to information taken from a drawing. The outline of the face and the organic contours in a face leads to the identification of a person. To get good information from a facial image a closely taken picture is the best. This gives information like posture, expression, background and the brightness of the face. (Li, 50-51)

Face recognition is conducted to understand the objective of a person that may cause possible trouble in various surroundings like an office or a common place. Other than the above gestures, a face can be recognized according to the style of the movement of the head. Face recognition is of importance in both household as well as industrial settings. Burglar bells can be customized with familiar faces, so that it can detect strange faces and keep the details for proof. This technique is of interest for marketing companies to find the nature of customers for a particular product and their behavior while shopping. The right to use various resources in an office and at home like telephone, computer or other equipments can be restricted using face identification technique. (Jain, 217)

Identification of a face should be conducted in an automatic manner from normal circumstance to get a picture of reflex actions of a person. This is against the identification of a face in a security enabled organization where a person stands in front of the detecting machine to prove his identity. The machine should catch the movements of a person in uncontrolled settings and process the information to arrive at conclusions quickly. This is the method of identifying face from video pictures from various angles and movements. The method of processing should be easy for standard processors to give out results quickly. A low priced image catcher is enough for this purpose, though a colored one will be of added advantage. (Jain, 221-222) low accuracy, high rejection method is used to find a lesser number of faces from a huge amount of doubtful information whereas a high accuracy, low rejection method is used to find more number, say, a thousand or a million people using a lesser quality of picture information. In a complex situation the posture and the amount of light on the face reflects on the quality of outcome from the biometric equipment. Another complexity while viewing a huge number of people is their posture like sitting, standing or moving positions. Moreover, the direction towards which a face is moving is another constraint. A low decree of picture will be recorded in the case of one permanent camera. Though this is adequate to capture picture of familiar people, other challenges for an automated identification is the different kind of dressing of a person, make up, ornaments, fashion of hair, scar in the face, age, etc. The automated equipment should be upgraded from time to keep an account of strangers and familiar people who visit the setting under surveillance. A video picture with voice is highly appreciable to register the presence of a person. (Jain, 223-224)

Fingerprints

Fingerprints are one of the ancient biometric methods of verifying identity of a person. In the beginning of the 20th century, fingerprints were used as signature to verify a personal legally. The assessment of fingerprints by a person is a difficult method and takes a long time and requires the work of a specialist. Thus mechanized method of fingerprint recognition came up. There are varied forms of information hidden in fingerprints. A detailed morphological analysis of fingerprints indicates that they remain the same from the time of birth till the time of death of a person. Secondly, fingerprint of a person is distinct. Fingerprints are not the same in any two individuals. This is also the case with twins though they have other similar features. The light colored region of a fingerprint is known as ridges and the dark colored region is known as valleys. These are collectively called the Galton characteristics. Ridges in a fingerprint may be describes as one curved portion. A fingerprint is formed by the joining of many ridges. The last part of the ridges and the small figures produced as a result of the passage of ridges is known as minutiae. The features found by Galton are classified into four; the first and last part of the ridges, enclosures, forks and islands. E.R.Henry is another name to reckon with fingerprints. He classified the figure of a fingerprint into various sections so clearly that government agencies use them to identify a person. (Jain, 4-7)

The sections that reveal the identity of a person are Arch (a), Tented Arch (T), Right Loop (R), Whorl (W) and Left Loop (LT). Both Henry's classification and Galton's characteristics are used to arrive at a conclusion on the identity of a person through fingerprints. Today mechanized methods are used to match verify the identity of fingerprint using computers. AFAS (Automatic Fingerprint Authentication System) and AFIS (Automatic Fingerprint Identification System) are the two prevalent method of verifying fingerprints. On entering a fingerprint into mechanized equipment, it gives out a set of results and a grade representing the level of identity of the fingerprint with each person verified in the available information. If there is no total identity to a person, the given set of results decreases the space between the targeted person and the fingerprint. Verification of fingerprint was widely used to in the police departments to find criminals. Off late, it is used to restrict admission to various offices and to provide security in financial institutions. Thus fingerprint identification equipments are widely available in the marketplace. (Jain, 8-10)

Iris

Identification of a live person from the iris gives precise result. The amount of error from such an identification method is least or nil. This is a new technology in comparison to identification using face or fingerprint. The uniqueness of an iris is studied depending on the color of the ring around the pupil. The iris has nearly two hundred and sixty six unique features which is a good base for a biometric verification. The iris constitutes tissues with elastic connective nature. The trabecular meshwork in an iris separates the iris with features like freckles, furrows, corona, rings and striations. The technology used to identify a person from his iris employs around 173 features to substantiate the identity. (U.S. General Accounting Office (b), 11) the iris remains the same during the entire life of a person unless it undergoes a wound. The nature of an iris is recorded using a high resolution camera to give a black and white picture. The border of the picture is scrutinized to find a match to the given information. The evident regions of the iris are later transformed into a pattern which is used to recognize the person. (U.S. General Accounting Office, 47)

Though the accuracy of identification from iris is almost perfect, the distance between the person and the camera is a vital factor. A lengthy distance from the camera will lead to a change in the resolution and the posture of the eye. The setting under which the picture of the iris is taken is another factor to substantiate the perfection of identity. The angle from which a picture of iris is taken results in a change if the posture is not restricted to a change of less than 30 degrees. If there is a major change in the posture of the iris a three dimensional picture of the iris is required to match the identity. (Kanade; Jain; Ratha, 4, 12)

Retina person can be identified from his retina by evaluating the picture of the blood vessels behind the eyes. These blood vessels are called choroidal vasculature. While taking the picture of the retina using infrared rays, the retina is not clearly seen. But certain equipments catch a glimpse of the visual spectrum that assists the evaluation of the retina. The exclusive model of the blood vessels is the identifying factor in a retina. (Bolle, 53) it is not the same even in the case of twins. (U.S. General Accounting Office (a), 27) EyeDentify equipments used to verify retina can check almost three thousand people. Verification of around 1500 people may be conducted in less than five seconds by this equipment. This equipment however cannot capture the image of individuals who has poor vision or severe astigmation. The person under question should be placed near the equipment to get a good picture. This is a costly method to identify retina in comparison to other method of biometric verification. An important aspect of identification through retina is the permanent condition of the retina unless it undergoes a change due to severe mishap. This is a reliable method because a person cannot fix an artificial retina into the eyes. Besides the use of EyeDentify equipment, recent inventions has lead to capture the retina using electromechanical feeler from a close point to learn about its reflective aspects as well as its absorption features. In this method, the retina is first exposed to a 7mW light which enhances the opportunities to know about the model of the veins. One of the disadvantages of the method is the huge price of the feeler. (Bolle, 53)

Another method used to get a picture of the retina is by passing light through the pupil of the eye. The model of the blood vessels is recorded at 400 spots to create a 96 byte model. Identification of a person through retina is employed in important security settings and government buildings. Biometrics using retina is one of the most difficult methods. Ancient techniques required individuals to look at a circling green light for about ten to twelve seconds at a distance of half an inch from the eyes and remain constant. But latest technology requires a person to stay constant only for about 1.5 seconds and the picture of retina can be capture at a distance of about one meter. This method is not sensitive to cataracts, sun glasses or contact lens. Hence it has an edge over the older method. Retina remains the same throughout life, but may undergo a change if a person undergoes glaucoma, high blood pressure, diabetes or AIDS. Though there are latest techniques to identify retina, certain individuals are reluctant to this method since they are exposed to infrared lights. Some people show reluctance since they feel this test would reveal symptoms of illness. Other than EyeDentify, Retinal Technologies is the enterprise which sells equipment to verify retina. (U.S. General Accounting Office (a), 27)