28.06.2017

Morpho has released a new 3D Face Reader for face scanning, the accuracy and stability of which is ensured by unique algorithms and the use of 3D biometric templates. Even the illumination of the object does not affect the operation of this device, since face identification is carried out using an IR camera, and the entire identification process takes about 1 second. Additional advantages of the 3D Face reader include direct control of an electric lock or turnstile, integration with any network access control system via Wiegand, LAN, RS232 and RS485, as well as PoE support for power supply and device control via a single Ethernet cable.

The developers paid attention to the design of 3D Face Reader no less than the functional component, resulting in a stylish ergonomic product that provides high-precision, both offline and online, and has a negligible percentage of erroneous admission / denial of access. All Morpho products that can be used for facial, fingerprint or vein identification can be integrated into existing access control systems for gradual upgrades. Moreover, along with the identification mode, all Morpho readers support the verification mode to restrict access to critical objects.

According to the manufacturer of the new face scanner, it is impossible to fool the 3D Face Reader. It is able to identify a person with high accuracy while driving, despite the presence of a headgear, glasses, headphones and other accessories that cover parts of the face, and also excludes the passage of strangers using a photo or video of a person registered in the system. This reader does not reduce its efficiency even with a lack of light on the object, since face identification is carried out in the infrared range. An additional advantage of such an identification device is that a person's hands always remain free, and the "identifier" itself cannot be lost or given to another person.

The simplest way to integrate the Morpho reader with your existing network infrastructure is via TCP / IP, since the same Ethernet cable can be used to configure, manage, and power the ID device. Especially for such cases, the developers of 3D Face Reader have provided support for PoE technology. Multiple Morpho biometric readers can be networked via TCP / IP, which can be used to identify a person at various points of the passage and control them via the SecureAdmin ™ software interface. It is important to note that this software allows you to accumulate biometric face templates created by one 3D Face reader for their use by all readers of the system.

Recently, many articles have appeared on Habré on Google's face identification systems. To be honest, from many of them it carries journalism and, to put it mildly, incompetence. And I wanted to write a good article on biometrics, it's not my first one! There are a couple of good articles on biometrics on Habré - but they are rather short and incomplete. Here I will try to briefly outline the general principles of biometric identification and the modern achievements of mankind in this matter. Including identification by faces.

The article has a sequel, which is, in fact, its prequel.

The article will be based on a joint publication with a colleague in the journal (BDI, 2009), revised to suit modern realities. Habré has no colleagues yet, but he supported the publication of the revised article here. At the time of publication, the article was a brief overview of the modern biometric technology market, which we conducted for ourselves before we launched our product. The evaluative judgments about applicability put forward in the second part of the article are based on the opinions of people who have used and implemented the products, as well as on the opinions of people involved in the production of biometric systems in Russia and Europe.

general information

Let's start with the basics. In 95% of cases, biometrics is essentially mathematical statistics. And matstat is an exact science, the algorithms of which are used everywhere: in radars and in Bayesian systems. Errors of the first and second kind can be taken as the two main characteristics of any biometric system). In the theory of radar, they are usually called "false alarm" or "target miss", and in biometrics, the most established concepts are FAR (False Acceptance Rate) and FRR (False Rejection Rate). The first number characterizes the probability of a false coincidence of the biometric characteristics of two people. The second is the likelihood of denial of access to an authorized person. The system is better, the lower the FRR value for the same FAR values. Sometimes the comparative characteristic EER is also used, which determines the point at which the FRR and FAR graphs intersect. But it is not always representative. More details can be found, for example,.
The following can be noted: if FAR and FRR are not given in the characteristics of the system based on open biometric databases, then whatever the manufacturers declare about its characteristics, this system is most likely incompetent or much weaker than competitors.
But not only FAR and FRR determine the quality of the biometric system. If this were the only way, then the leading technology would be the recognition of people by DNA, for which FAR and FRR tend to zero. But it is obvious that this technology is not applicable at the current stage of human development! We have developed several empirical characteristics that allow us to assess the quality of the system. “Counterfeit resistance” is an empirical characteristic that summarizes how easy it is to trick a biometric identifier. "Resistance to the environment" - a characteristic that empirically evaluates the stability of the system under various external conditions, such as changes in lighting or room temperature. “Ease of use” shows how difficult it is to use a biometric scanner, whether identification is possible “on the go”. An important characteristic is "Speed ​​of operation" and "System cost". Do not forget that a person's biometric characteristics can change over time, so if it is unstable, this is a significant disadvantage.
The abundance of biometric methods is amazing. The main methods using static biometric characteristics of a person are identification by papillary pattern on fingers, iris, face geometry, retina, hand vein pattern, and hand geometry. There is also a family of methods that use dynamic characteristics: identification by voice, handwriting dynamics, heart rate, gait. Below is the distribution of the biometric market a couple of years ago. In every other source, these data fluctuate by 15-20 percent, so this is just an estimate. Also here under the concept of "hand geometry" are two different methods, which will be discussed below.


In this article, we will consider only those characteristics that are applicable in access control systems (ACS) or in related tasks. Due to their superiority, these are primarily static characteristics. Of the dynamic characteristics at the moment, only voice recognition has at least some statistical significance (comparable to the worst static algorithms FAR ~ 0.1%, FRR ~ 6%), but only under ideal conditions.
To get a feel for the probabilities of FAR and FRR, you can estimate how often false matches will occur if you install an identification system at a gateway with N staff. The probability of a false coincidence of a fingerprint received by a scanner for a database of N fingerprints is FAR ∙ N. And every day about N people also pass through the access control point. Then the probability of error for a working day is FAR ∙ (N ∙ N). Of course, depending on the goals of the identification system, the probability of an error per unit of time can vary greatly, but if we accept one error per working day as acceptable, then:
(1)
Then we get that the stable operation of the identification system with FAR = 0.1% = 0.001 is possible with the number of personnel N≈30.

Biometric Scanners

Today, the concept of "biometric algorithm" and "biometric scanner" are not necessarily interconnected. The company can produce these elements individually, or together. The greatest differentiation of scanner manufacturers and software manufacturers has been achieved in the market for finger papillary biometrics. The smallest 3D face scanner on the market. In fact, the level of differentiation largely reflects the development and saturation of the market. The more choice, the more the topic is worked out and brought to perfection. Different scanners have a different set of capabilities. Basically it is a set of tests to check whether a biometric object is tampered with or not. For finger scanners it can be a bump check or temperature check, for eye scanners it can be a pupil accommodation check, for face scanners it can be a face movement.
Scanners greatly influence the received FAR and FRR statistics. In some cases, these numbers can change tens of times, especially in real conditions. Usually, the characteristics of the algorithm are given for some "ideal" base, or just for a well-suited one, where blurry and blurry frames are thrown out. Only a few algorithms honestly indicate both the base and the full FAR / FRR output for it.

And now in more detail about each of the technologies

Fingerprints

Fingerprinting (fingerprint recognition) is the most developed biometric method of personal identification today. The catalyst for the development of the method was its widespread use in forensic science of the 20th century.
Each person has a unique papillary fingerprint pattern, which makes identification possible. Typically, algorithms use characteristic points on fingerprints: the end of the pattern line, branching of the line, single points. Additionally, information about the morphological structure of the fingerprint is involved: the relative position of closed lines of the papillary pattern, "arched" and spiral lines. The features of the papillary pattern are converted into a unique code that preserves the information content of the print image. And it is the “fingerprint codes” that are stored in the database used for search and comparison. The time to translate a fingerprint image into a code and its identification usually does not exceed 1 s, depending on the size of the database. The time spent on raising the hand is not counted.

VeriFinger SDK statistics obtained with the DP U.are.U fingerprint scanner were used as a source of data on FAR and FRR. Over the past 5-10 years, the characteristics of finger recognition have not stepped forward much, so the figures given are a good example of the average value of modern algorithms. The VeriFinger algorithm itself won the International Fingerprint Verification Competition for several years, where finger recognition algorithms competed.

The typical FAR value for the fingerprint recognition method is 0.001%.
From the formula (1) we obtain that the stable operation of the identification system with FAR = 0.001% is possible with the number of personnel N≈300.

The advantages of the method. High reliability - statistical indicators of the method are better than indicators of identification methods by face, voice, painting. Low cost of devices that scan a fingerprint image. A fairly simple procedure for scanning a fingerprint.
Disadvantages: the papillary pattern of the fingerprint is very easily damaged by small scratches and cuts. People who have used scanners in factories with several hundred employees report a high scan failure rate. Many of the scanners are inadequate for dry skin and keep old people out. When communicating at the last MIPS exhibition, the head of the security service of a large chemical enterprise said that their attempt to introduce finger scanners at the enterprise (scanners of various systems were tried) failed - the minimum exposure to chemicals on the employees' fingers caused the security systems of the scanners to fail - the scanners declared the fingers to be fake. There is also a lack of protection against counterfeiting the image of the fingerprint, partly due to the widespread use of the method. Of course, not all scanners can be tricked with the methods of the Legend Busters, but still. For some people with "unsuitable" fingers (especially body temperature, humidity), the probability of denial of access can reach 100%. The number of such people varies from a fraction of a percent for expensive scanners to ten percent for inexpensive ones.
Of course, it is worth noting that a large number of shortcomings are caused by the widespread prevalence of the system, but these shortcomings take place and they appear very often.

Market situation

At the moment, fingerprint recognition systems occupy more than half of the biometric market. Many Russian and foreign companies are engaged in the production of access control systems based on the fingerprint identification method. Due to the fact that this direction is one of the oldest, it is the most widespread and is by far the most developed. Fingerprint scanners have come a really long way to improve. Modern systems are equipped with various sensors (temperature, pressure, etc.), which increase the degree of protection against counterfeiting. The systems are becoming more convenient and compact every day. In fact, the developers have already reached a certain limit in this area, and there is nowhere to develop the method further. In addition, most companies produce ready-made systems that are equipped with everything you need, including software. Integrators in this area simply do not need to assemble the system on their own, since it is unprofitable and will take more time and effort than buying a ready-made and already inexpensive system, the more the choice will be really wide.
Among foreign companies dealing with fingerprint recognition systems, one can mention SecuGen (USB scanners for PCs, scanners that can be installed in enterprises or embedded in locks, SDK and software for connecting the system with a computer); Bayometric Inc. (fingerprint scanners, TAA / Access control systems, fingerprint SDKs, embedded fingerprint modules); DigitalPersona, Inc. (USB-scanners, SDK). In Russia, the following companies work in this area: BioLink (fingerprint scanners, biometric access control devices, software); Sonda (fingerprint scanners, biometric access control devices, SDK); SmartLock (fingerprint scanners and modules), etc.

Iris

The iris of the eye is a unique characteristic of humans. The iris pattern is formed at the eighth month of intrauterine development, finally stabilizes at the age of about two years and practically does not change during life, except as a result of severe trauma or sharp pathologies. The method is one of the most accurate among biometric methods.
The iris identification system is logically divided into two parts: a device for capturing an image, its primary processing and transmission to a computer, and a computer that compares the image with images in the database, and sends a command for admission to the executive device.
The time of primary image processing in modern systems is about 300-500ms, the speed of comparing the resulting image with the base is at the level of 50,000-150000 comparisons per second on an ordinary PC. This speed of comparison does not impose restrictions on the application of the method in large organizations when used in access systems. When using specialized computers and search optimization algorithms, it becomes even possible to identify a person among the inhabitants of an entire country.
I can immediately answer that I am somewhat biased and have a positive attitude towards this method, since it was in this field that we launched our startup. A paragraph at the end will be devoted to a small self-promotion.

Statistical characteristics of the method

The characteristics of FAR and FRR for the iris are the best in the class of modern biometric systems (with the possible exception of the retinal recognition method). The article presents the characteristics of the iris recognition library of our algorithm - EyeR SDK, which correspond to the VeriEye algorithm tested on the same bases. We used CASIA databases obtained by their scanner.

The typical FAR value is 0.00001%.
According to the formula (1), N≈3000 is the number of personnel of the organization, at which the identification of the employee is quite stable.
It is worth noting here an important feature that distinguishes the iris recognition system from other systems. In the case of using a camera with a resolution of 1.3MP or more, you can capture two eyes in one frame. Since the probabilities FAR and FRR are statistically independent probabilities, when recognizing from two eyes, the FAR value will be approximately equal to the square of the FAR value for one eye. For example, for FAR 0.001%, when using two eyes, the probability of a false tolerance will be 10-8%, with FRR it is only twice as high as the corresponding FRR value for one eye with FAR = 0.001%.

Advantages and disadvantages of the method

The advantages of the method. Statistical reliability of the algorithm. Capturing an image of the iris can be performed at a distance from a few centimeters to several meters, while there is no physical contact between the person and the device. The iris is protected from damage - which means it will not change over time. It is also possible to use a large number of anti-counterfeiting methods.
Disadvantages of the method. The price of a system based on iris is higher than the price of a system based on finger recognition or facial recognition. Low availability of ready-made solutions. Any integrator who comes to the Russian market today and says “give me a ready-made system” is likely to break off. Most of them sell expensive turnkey systems installed by large companies such as Iridian or LG.

Market situation

At the moment, the share of iris identification technologies in the global biometric market is, according to various estimates, from 6 to 9 percent (while fingerprint recognition technologies occupy over half of the market). It should be noted that from the very beginning of the development of this method, its strengthening in the market was slowed down by the high cost of equipment and components required to assemble an identification system. However, with the development of digital technologies, the cost of a separate system began to decline.
The leader in software development in this area is Iridian Technologies.
Market entry for a large number of manufacturers was limited by the technical complexity of the scanners and, as a result, their high cost, as well as the high price of software due to Iridian's monopoly position in the market. These factors allowed only large companies to develop in the field of iris recognition, most likely already engaged in the production of some components suitable for the identification system (high-resolution optics, miniature cameras with infrared illumination, etc.). Examples of such companies can be LG Electronics, Panasonic, OKI. They entered into an agreement with Iridian Technologies, and as a result of joint work, the following identification systems appeared: Iris Access 2200, BM-ET500, OKI IrisPass. In the future, improved models of systems arose, thanks to the technical capabilities of these companies to independently develop in this area. It should be said that the above companies have also developed their own software, but in the end they prefer Iridian Technologies software in the finished system.
The Russian market is dominated by the products of foreign companies. Although that can hardly be bought. For a long time, the Papillon firm assured everyone that they had iris recognition. But even representatives of RosAtom, their direct purchaser, for whom they made the system, say that this is not true. At some point, another Russian company showed up that made iris scanners. Now I can't remember the name. They bought the algorithm from someone, possibly from the same VeriEye. The scanner itself was a 10-15 year old system, not contactless.
In the last year, a couple of new manufacturers entered the world market due to the expiration of the primary patent for recognizing a person by the eyes. The most trusted of them, in my opinion, is AOptix. At least their previews and documentation are not suspicious. The second company is SRI International. Even at first glance, to a person involved in iris recognition systems, their videos seem very deceitful. Although I would not be surprised if in reality they can do something. And that and that system does not show data on FAR and FRR, and also, apparently, is not protected from counterfeiting.

Face recognition

There are many face geometry recognition methods. All of them are based on the fact that facial features and the shape of the skull of each person are individual. This area of ​​biometrics seems attractive to many, because we recognize each other primarily by their faces. This area is divided into two directions: 2-D recognition and 3-D recognition. Each of them has advantages and disadvantages, but much also depends on the area of ​​application and the requirements for a particular algorithm.
In short, I'll tell you about 2-d and move on to one of the most interesting methods for today - 3-d.

2-D face recognition

2-D face recognition is one of the most statistically ineffective biometrics. It appeared quite a long time ago and was used mainly in forensic science, which contributed to its development. Subsequently, computer interpretations of the method appeared, as a result of which it became more reliable, but, of course, it was inferior and every year more and more inferior to other biometric methods of personality identification. Currently, due to poor statistical indicators, it is used in multimodal or, as it is also called, cross-biometrics, or in social networks.

Statistical characteristics of the method

For FAR and FRR, data from the VeriLook algorithms were used. Again, it has very common characteristics for modern algorithms. Sometimes algorithms with an FRR of 0.1% are flashed with a similar FAR, but the bases on which they were obtained are very doubtful (cut out background, the same facial expression, the same hairstyle, lighting).

The typical FAR value is 0.1%.
From the formula (1) we get N≈30 - the number of the organization's personnel, at which the employee's identification is quite stable.
As you can see, the statistical indicators of the method are quite modest: this negates the advantage of the method that it is possible to carry out covert shooting of faces in crowded places. It's funny to see how, a couple of times a year, another project is funded to detect criminals through video cameras installed in crowded places. Over the past ten years, the statistical characteristics of the algorithm have not improved, and the number of such projects has grown. Although, it is worth noting that the algorithm is quite suitable for guiding a person in a crowd through many cameras.

Advantages and disadvantages of the method

The advantages of the method. With 2-D recognition, unlike most biometric methods, expensive equipment is not required. With the appropriate equipment, the possibility of recognition at significant distances from the camera.
Flaws. Low statistical significance. Lighting requirements are imposed (for example, it is not possible to register the faces of people entering from the street on a sunny day). For many algorithms, any external interference is unacceptable, for example, glasses, a beard, some elements of a hairstyle. A frontal face image is mandatory, with very slight deviations. Many algorithms do not take into account possible changes in facial expressions, that is, the expression must be neutral.

3-D face recognition

The implementation of this method is a rather difficult task. Despite this, there are currently many methods for 3-D face recognition. The methods cannot be compared with each other, since they use different scanners and bases. not all of them issue FAR and FRR; completely different approaches are used.
A transitional method from 2-d to 3-d is a method that implements the accumulation of information about a person. This method has better characteristics than the 2d method, but just like it uses only one camera. When the subject is entered into the base, the subject turns his head and the algorithm joins the image together, creating a 3d template. And during recognition, several frames of the video stream are used. This method is rather experimental and I have never seen an implementation for ACS systems.
The most classic method is the template projection method. It consists in the fact that a mesh is projected onto an object (face). Then the camera takes pictures at a rate of tens of frames per second, and the resulting images are processed by a special program. A ray falling on a curved surface bends - the greater the curvature of the surface, the stronger the bending of the ray. Initially, it used a source of visible light, supplied through the "louvers". Then visible light was replaced by infrared, which has a number of advantages. Usually, at the first stage of processing, images are discarded in which the face is not visible at all or there are foreign objects that interfere with identification. Based on the images obtained, a 3-D model of the face is restored, on which unnecessary interference (hairstyle, beard, mustache and glasses) are highlighted and removed. Then the model is analyzed - anthropometric features are highlighted, which, as a result, are written into a unique code that is entered into the database. Image capture and processing time is 1-2 seconds for the best models.
Also gaining popularity is the 3-d recognition method from the image obtained from several cameras. An example of this is Vocord with its 3D scanner. This method gives positioning accuracy, according to the assurances of the developers, higher than the template projection method. But until I see FAR and FRR at least on their own base - I won't believe it !!! But it has been under development for 3 years already, and progress has not yet been seen at exhibitions.

Statistical indicators of the method

Complete data on FRR and FAR for algorithms of this class are not publicly presented on manufacturers' websites. But for the best models from Bioscript (3D EnrolCam, 3D FastPass), working according to the template projection method with FAR = 0.0047%, FRR is 0.103%.
It is believed that the statistical reliability of the method is comparable to that of the fingerprint identification method.

Advantages and disadvantages of the method

The advantages of the method. No need to contact the scanning device. Low sensitivity to external factors, both on the person himself (the appearance of glasses, a beard, a change in hairstyle), and in his environment (light, head turn). High level of reliability comparable to the fingerprint identification method.
Disadvantages of the method. Expensive equipment. The commercially available systems outperformed even the iris scanners. Changes in facial expressions and interferences on the face worsen the statistical reliability of the method. The method is not yet well developed, especially in comparison with the long-used fingerprinting, which complicates its widespread use.

Market situation

Face recognition is considered to be one of the "three big biometrics" along with fingerprint and iris recognition. I must say that this method is quite common, and for now it is preferred over recognition by the iris of the eye. The share of face geometry recognition technologies in the total volume of the world biometric market can be estimated at 13-18 percent. In Russia, this technology is also showing greater interest than, for example, identification by the iris. As mentioned earlier, there are many 3-D recognition algorithms. Most of the companies prefer to develop ready-made systems, including scanners, servers and software. However, there are those who only offer the consumer the SDK. Today, the following companies developing this technology can be noted: Geometrix, Inc. (3D face scanners, software), Genex Technologies (3D face scanners, software) in the USA, Cognitec Systems GmbH (SDK, special computers, 2D cameras) in Germany, Bioscrypt (3D face scanners, software) - a subsidiary of the American company L- 1 Identity Solutions.
In Russia, the Artec Group companies (3D face scanners and software) work in this direction - a company headquartered in California, and development and production are carried out in Moscow. Also, several Russian companies own 2D face recognition technology - Vocord, ITV, etc.
In the field of 2D face recognition, software is the main subject of development. conventional cameras do an excellent job of capturing facial images. The solution to the problem of face recognition has to some extent reached a dead end - for several years now, there has been practically no improvement in the statistical indicators of algorithms. In this area, there is a systematic "work on errors."
3D face recognition is now a much more attractive area for developers. Many teams work in it and regularly hear about new discoveries. Many works are in the “just about to release” state. But so far there are only old offers on the market, the choice has not changed in recent years.
One of the interesting points that I sometimes think about and which Habr might answer: is the accuracy of kinect enough to create such a system? Projects for pulling a 3d model of a person through it are quite possible.

Hand vein recognition

This is a new technology in the field of biometrics, its widespread use began only 5-10 years ago. An infrared camera takes pictures of the outside or inside of the hand. The pattern of veins is formed due to the fact that the hemoglobin of the blood absorbs IR radiation. As a result, the degree of reflection is reduced and veins are visible on the camera as black lines. A special program based on the received data creates a digital convolution. No human contact with the scanning device is required.
The technology is comparable in reliability with recognition by the iris of the eye, surpassing it in some way, and inferior in some way.
FRR and FAR values ​​are for Palm Vein scanner. According to the developer's data, with a FAR of 0.0008%, the FRR is 0.01%. A more accurate graph for several values ​​is not given by any company.

Advantages and disadvantages of the method

The advantages of the method. No need to contact the scanning device. High reliability - the statistical indicators of the method are comparable to those of the iris. Hidden characteristics: unlike all of the above, this characteristic is very difficult to obtain from a person “on the street”, for example, by photographing him with a camera.
Disadvantages of the method. The illumination of the scanner by the sun's rays and the rays of halogen lamps is unacceptable. Certain age-related diseases, such as arthritis, greatly worsen FAR and FRR. The method is less studied in comparison with other static biometric methods.

Market situation

Hand vein pattern recognition is a fairly new technology, and therefore its share in the world market is small and amounts to about 3%. However, there is an increasing interest in this method. The fact is that, being quite accurate, this method does not require such expensive equipment as, for example, recognition methods based on face geometry or iris. Now many companies are developing in this area. For example, by order of the English company TDSi, software was developed for the PalmVein biometric palm vein reader, presented by Fujitsu. The scanner itself was developed by Fujitsu primarily to combat financial fraud in Japan.
The following companies Veid Pte also work in the field of identification by vein pattern. Ltd. (scanner, software), Hitachi VeinID (scanners)
In Russia, I do not know any companies dealing with this technology.

Retina

Until recently, it was believed that the most reliable method of biometric identification and personality authentication is a method based on scanning the retina of the eye. It contains the best features of identification by the iris and by the veins of the arm. The scanner reads the pattern of capillaries on the surface of the retina. The retina has an immobile structure that does not change over time, except as a result of a disease, such as cataracts.
Retinal scans are performed using low-intensity infrared light directed through the pupil to the blood vessels at the back of the eye. Retinal scanners have become widespread in access control systems for highly classified objects, since they have one of the lowest percentages of denial of access for registered users and there is practically no erroneous access permission.
Unfortunately, a number of difficulties arise when using this biometrics method. The scanner here is a very complex optical system, and a person must not move for a considerable time while the system is guided, which causes unpleasant sensations.
According to EyeDentify, for the ICAM2001 scanner with FAR = 0.001%, the FRR value is 0.4%.

Advantages and disadvantages of the method

Advantages. High level of statistical reliability. Due to the low prevalence of systems, there is little likelihood of developing a way to "cheat" them.
Flaws. Sophisticated system with high turnaround times. The high cost of the system. Lack of a wide supply market and, as a consequence, insufficient intensity of development of the method.

Hand geometry

This method, quite widespread even 10 years ago, and originated from forensic science in recent years, is on the decline. It is based on obtaining the geometric characteristics of the hands: the lengths of the fingers, the width of the palm, etc. This method, like the retina of the eye, is dying, and since it has much lower characteristics, we will not even introduce it more fully.
It is sometimes believed that geometric recognition methods are used in vein recognition systems. But on sale we have never seen such a clearly declared one. And besides, often when recognizing by veins, a picture of only the palm is taken, while when recognizing by geometry, a picture of fingers is taken.

A little self-promotion

At one time, we developed a good eye recognition algorithm. But at that time such a high-tech thing was not needed in this country, and I didn't want to go to bourgeois (where we were invited after the very first article). But suddenly, after a year and a half, there were investors who wanted to build a "biometric portal" for themselves - a system that would eat two eyes and use the color component of the iris (for which the investor had a world patent). Actually, now we are doing this. But this is not an article about self-promotion, this is a short lyrical digression. If anyone is interested there is some information, but sometime in the future, when we enter the market (or not), I will write a few words about the twists and turns of a biometric project in Russia.

conclusions

Even in the class of static biometrics systems, there is a large selection of systems. Which one should you choose? It all depends on the requirements for the security system. The most statistically reliable and tamper-resistant access systems are the iris and arm veins. For the former, there is a wider market for proposals. But this is not the limit. Biometric identification systems can be combined to achieve astronomical accuracy. The cheapest and easiest to use, but with good statistics, are the fingertip systems. 2D face tolerance is convenient and cheap, but has limited applicability due to poor statistics.
Consider the characteristics that each of the systems will have: resistance to counterfeiting, resistance to the environment, ease of use, cost, speed, stability of the biometric feature over time. Let's put the marks from 1 to 10 in each column. The closer the score is to 10, the better the system is in this regard. The principles for choosing grades were described at the very beginning of the article.


Also consider the relationship between FAR and FRR for these systems. This ratio determines the efficiency of the system and the breadth of its use.


It is worth remembering that for the iris, you can increase the accuracy of the system almost quadratically, without loss of time, if you complicate the system by making it two eyes. For the fingerprint method - by combining several fingers, and recognition by veins, by combining two hands, but such an improvement is possible only with an increase in the time spent working with a person.
Summarizing the results for the methods, we can say that for medium and large objects, as well as for objects with a maximum safety requirement, the iris should be used as biometric access and, possibly, recognition by hand veins. For facilities with up to several hundred personnel, fingerprint access will be optimal. 2D face recognition systems are very specific. They may be required in cases where recognition requires the absence of physical contact, but it is impossible to deliver the iris control system. For example, if it is necessary to identify a person without his participation, a hidden camera, or an outdoor detection camera, but this is possible only with a small number of subjects in the base and a small stream of people filmed by the camera.

Note for a young technician

Some manufacturers, for example Neurotechnology, have demo versions of biometrics methods that they release on their website, so you can easily plug them in and play around. For those who decide to delve into the problem more seriously, I can recommend the only book that I have seen in Russian - "A Guide to Biometrics" by R.M. Ball, J.H. Connel, S. Pancanti. There are many algorithms and their mathematical models. Not everything is complete and not everything corresponds to modernity, but the base is not bad and comprehensive.

P.S.

In this opus, I did not go into the problem of authentication, but only touched on identification. In principle, from the characteristics of FAR / FRR and the possibility of forgery, all conclusions on the issue of authentication suggest themselves.

Tags: Add Tags

ZlodeiBaal August 11, 2011 at 09:54 PM

Modern biometric identification methods

• Information Security

Recently, many articles have appeared on Habré on Google's face identification systems. To be honest, from many of them it carries journalism and, to put it mildly, incompetence. And I wanted to write a good article on biometrics, it's not my first one! There are a couple of good articles on biometrics on Habré - but they are rather short and incomplete. Here I will try to briefly outline the general principles of biometric identification and the modern achievements of mankind in this matter. Including identification by faces.

The article has, which is essentially its prequel.

The article will be based on a joint publication with a colleague in the journal (BDI, 2009), revised to suit modern realities. Habré has no colleagues yet, but he supported the publication of the revised article here. At the time of publication, the article was a brief overview of the modern biometric technology market, which we conducted for ourselves before we launched our product. The evaluative judgments about applicability put forward in the second part of the article are based on the opinions of people who have used and implemented the products, as well as on the opinions of people involved in the production of biometric systems in Russia and Europe.

general information

Let's start with the basics. In 95% of cases, biometrics is essentially mathematical statistics. And matstat is an exact science, the algorithms of which are used everywhere: in radars and in Bayesian systems. Errors of the first and second kind can be taken as the two main characteristics of any biometric system). In the theory of radar, they are usually called "false alarm" or "target miss", and in biometrics, the most established concepts are FAR (False Acceptance Rate) and FRR (False Rejection Rate). The first number characterizes the probability of a false coincidence of the biometric characteristics of two people. The second is the likelihood of denial of access to an authorized person. The system is better, the lower the FRR value for the same FAR values. Sometimes the comparative characteristic EER is also used, which determines the point at which the FRR and FAR graphs intersect. But it is not always representative. More details can be found, for example,.
The following can be noted: if FAR and FRR are not given in the characteristics of the system based on open biometric databases, then whatever the manufacturers declare about its characteristics, this system is most likely incompetent or much weaker than competitors.
But not only FAR and FRR determine the quality of the biometric system. If this were the only way, then the leading technology would be the recognition of people by DNA, for which FAR and FRR tend to zero. But it is obvious that this technology is not applicable at the current stage of human development! We have developed several empirical characteristics that allow us to assess the quality of the system. “Counterfeit resistance” is an empirical characteristic that summarizes how easy it is to trick a biometric identifier. "Resistance to the environment" - a characteristic that empirically evaluates the stability of the system under various external conditions, such as changes in lighting or room temperature. “Ease of use” shows how difficult it is to use a biometric scanner, whether identification is possible “on the go”. An important characteristic is "Speed ​​of operation" and "System cost". Do not forget that a person's biometric characteristics can change over time, so if it is unstable, this is a significant disadvantage.
The abundance of biometric methods is amazing. The main methods using static biometric characteristics of a person are identification by papillary pattern on fingers, iris, face geometry, retina, hand vein pattern, and hand geometry. There is also a family of methods that use dynamic characteristics: identification by voice, handwriting dynamics, heart rate, gait. Below is the distribution of the biometric market a couple of years ago. In every other source, these data fluctuate by 15-20 percent, so this is just an estimate. Also here under the concept of "hand geometry" are two different methods, which will be discussed below.


In this article, we will consider only those characteristics that are applicable in access control systems (ACS) or in related tasks. Due to their superiority, these are primarily static characteristics. Of the dynamic characteristics at the moment, only voice recognition has at least some statistical significance (comparable to the worst static algorithms FAR ~ 0.1%, FRR ~ 6%), but only under ideal conditions.
To get a feel for the probabilities of FAR and FRR, you can estimate how often false matches will occur if you install an identification system at a gateway with N staff. The probability of a false coincidence of a fingerprint received by a scanner for a database of N fingerprints is FAR ∙ N. And every day about N people also pass through the access control point. Then the probability of error for a working day is FAR ∙ (N ∙ N). Of course, depending on the goals of the identification system, the probability of an error per unit of time can vary greatly, but if we accept one error per working day as acceptable, then:
(1)
Then we get that the stable operation of the identification system with FAR = 0.1% = 0.001 is possible with the number of personnel N≈30.

Biometric Scanners

Today, the concept of "biometric algorithm" and "biometric scanner" are not necessarily interconnected. The company can produce these elements individually, or together. The greatest differentiation of scanner manufacturers and software manufacturers has been achieved in the market for finger papillary biometrics. The smallest 3D face scanner on the market. In fact, the level of differentiation largely reflects the development and saturation of the market. The more choice, the more the topic is worked out and brought to perfection. Different scanners have a different set of capabilities. Basically it is a set of tests to check whether a biometric object is tampered with or not. For finger scanners it can be a bump check or temperature check, for eye scanners it can be a pupil accommodation check, for face scanners it can be a face movement.
Scanners greatly influence the received FAR and FRR statistics. In some cases, these numbers can change tens of times, especially in real conditions. Usually, the characteristics of the algorithm are given for some "ideal" base, or just for a well-suited one, where blurry and blurry frames are thrown out. Only a few algorithms honestly indicate both the base and the full FAR / FRR output for it.

And now in more detail about each of the technologies

Fingerprints

Fingerprinting (fingerprint recognition) is the most developed biometric method of personal identification today. The catalyst for the development of the method was its widespread use in forensic science of the 20th century.
Each person has a unique papillary fingerprint pattern, which makes identification possible. Typically, algorithms use characteristic points on fingerprints: the end of the pattern line, branching of the line, single points. Additionally, information about the morphological structure of the fingerprint is involved: the relative position of closed lines of the papillary pattern, "arched" and spiral lines. The features of the papillary pattern are converted into a unique code that preserves the information content of the print image. And it is the “fingerprint codes” that are stored in the database used for search and comparison. The time to translate a fingerprint image into a code and its identification usually does not exceed 1 s, depending on the size of the database. The time spent on raising the hand is not counted.

VeriFinger SDK statistics obtained with the DP U.are.U fingerprint scanner were used as a source of data on FAR and FRR. Over the past 5-10 years, the characteristics of finger recognition have not stepped forward much, so the figures given are a good example of the average value of modern algorithms. The VeriFinger algorithm itself won the International Fingerprint Verification Competition for several years, where finger recognition algorithms competed.

The typical FAR value for the fingerprint recognition method is 0.001%.
From the formula (1) we obtain that the stable operation of the identification system with FAR = 0.001% is possible with the number of personnel N≈300.

The advantages of the method. High reliability - statistical indicators of the method are better than indicators of identification methods by face, voice, painting. Low cost of devices that scan a fingerprint image. A fairly simple procedure for scanning a fingerprint.
Disadvantages: the papillary pattern of the fingerprint is very easily damaged by small scratches and cuts. People who have used scanners in factories with several hundred employees report a high scan failure rate. Many of the scanners are inadequate for dry skin and keep old people out. When communicating at the last MIPS exhibition, the head of the security service of a large chemical enterprise said that their attempt to introduce finger scanners at the enterprise (scanners of various systems were tried) failed - the minimum exposure to chemicals on the employees' fingers caused the security systems of the scanners to fail - the scanners declared the fingers to be fake. There is also a lack of protection against counterfeiting the image of the fingerprint, partly due to the widespread use of the method. Of course, not all scanners can be tricked with the methods of the Legend Busters, but still. For some people with "unsuitable" fingers (especially body temperature, humidity), the probability of denial of access can reach 100%. The number of such people varies from a fraction of a percent for expensive scanners to ten percent for inexpensive ones.
Of course, it is worth noting that a large number of shortcomings are caused by the widespread prevalence of the system, but these shortcomings take place and they appear very often.

Market situation

At the moment, fingerprint recognition systems occupy more than half of the biometric market. Many Russian and foreign companies are engaged in the production of access control systems based on the fingerprint identification method. Due to the fact that this direction is one of the oldest, it is the most widespread and is by far the most developed. Fingerprint scanners have come a really long way to improve. Modern systems are equipped with various sensors (temperature, pressure, etc.), which increase the degree of protection against counterfeiting. The systems are becoming more convenient and compact every day. In fact, the developers have already reached a certain limit in this area, and there is nowhere to develop the method further. In addition, most companies produce ready-made systems that are equipped with everything you need, including software. Integrators in this area simply do not need to assemble the system on their own, since it is unprofitable and will take more time and effort than buying a ready-made and already inexpensive system, the more the choice will be really wide.
Among foreign companies dealing with fingerprint recognition systems, one can mention SecuGen (USB scanners for PCs, scanners that can be installed in enterprises or embedded in locks, SDK and software for connecting the system with a computer); Bayometric Inc. (fingerprint scanners, TAA / Access control systems, fingerprint SDKs, embedded fingerprint modules); DigitalPersona, Inc. (USB-scanners, SDK). In Russia, the following companies work in this area: BioLink (fingerprint scanners, biometric access control devices, software); Sonda (fingerprint scanners, biometric access control devices, SDK); SmartLock (fingerprint scanners and modules), etc.

Iris

The iris of the eye is a unique characteristic of humans. The iris pattern is formed at the eighth month of intrauterine development, finally stabilizes at the age of about two years and practically does not change during life, except as a result of severe trauma or sharp pathologies. The method is one of the most accurate among biometric methods.
The iris identification system is logically divided into two parts: a device for capturing an image, its primary processing and transmission to a computer, and a computer that compares the image with images in the database, and sends a command for admission to the executive device.
The time of primary image processing in modern systems is about 300-500ms, the speed of comparing the resulting image with the base is at the level of 50,000-150000 comparisons per second on an ordinary PC. This speed of comparison does not impose restrictions on the application of the method in large organizations when used in access systems. When using specialized computers and search optimization algorithms, it becomes even possible to identify a person among the inhabitants of an entire country.
I can immediately answer that I am somewhat biased and have a positive attitude towards this method, since it was in this field that we launched our startup. A paragraph at the end will be devoted to a small self-promotion.

Statistical characteristics of the method

The characteristics of FAR and FRR for the iris are the best in the class of modern biometric systems (with the possible exception of the retinal recognition method). The article presents the characteristics of the iris recognition library of our algorithm - EyeR SDK, which correspond to the VeriEye algorithm tested on the same bases. We used CASIA databases obtained by their scanner.

The typical FAR value is 0.00001%.
According to the formula (1), N≈3000 is the number of personnel of the organization, at which the identification of the employee is quite stable.
It is worth noting here an important feature that distinguishes the iris recognition system from other systems. In the case of using a camera with a resolution of 1.3MP or more, you can capture two eyes in one frame. Since the probabilities FAR and FRR are statistically independent probabilities, when recognizing from two eyes, the FAR value will be approximately equal to the square of the FAR value for one eye. For example, for FAR 0.001%, when using two eyes, the probability of a false tolerance will be 10-8%, with FRR it is only twice as high as the corresponding FRR value for one eye with FAR = 0.001%.

Advantages and disadvantages of the method

The advantages of the method. Statistical reliability of the algorithm. Capturing an image of the iris can be performed at a distance from a few centimeters to several meters, while there is no physical contact between the person and the device. The iris is protected from damage - which means it will not change over time. It is also possible to use a large number of anti-counterfeiting methods.
Disadvantages of the method. The price of a system based on iris is higher than the price of a system based on finger recognition or facial recognition. Low availability of ready-made solutions. Any integrator who comes to the Russian market today and says “give me a ready-made system” is likely to break off. Most of them sell expensive turnkey systems installed by large companies such as Iridian or LG.

Market situation

At the moment, the share of iris identification technologies in the global biometric market is, according to various estimates, from 6 to 9 percent (while fingerprint recognition technologies occupy over half of the market). It should be noted that from the very beginning of the development of this method, its strengthening in the market was slowed down by the high cost of equipment and components required to assemble an identification system. However, with the development of digital technologies, the cost of a separate system began to decline.
The leader in software development in this area is Iridian Technologies.
Market entry for a large number of manufacturers was limited by the technical complexity of the scanners and, as a result, their high cost, as well as the high price of software due to Iridian's monopoly position in the market. These factors allowed only large companies to develop in the field of iris recognition, most likely already engaged in the production of some components suitable for the identification system (high-resolution optics, miniature cameras with infrared illumination, etc.). Examples of such companies can be LG Electronics, Panasonic, OKI. They entered into an agreement with Iridian Technologies, and as a result of joint work, the following identification systems appeared: Iris Access 2200, BM-ET500, OKI IrisPass. In the future, improved models of systems arose, thanks to the technical capabilities of these companies to independently develop in this area. It should be said that the above companies have also developed their own software, but in the end they prefer Iridian Technologies software in the finished system.
The Russian market is dominated by the products of foreign companies. Although that can hardly be bought. For a long time, the Papillon firm assured everyone that they had iris recognition. But even representatives of RosAtom, their direct purchaser, for whom they made the system, say that this is not true. At some point, another Russian company showed up that made iris scanners. Now I can't remember the name. They bought the algorithm from someone, possibly from the same VeriEye. The scanner itself was a 10-15 year old system, not contactless.
In the last year, a couple of new manufacturers entered the world market due to the expiration of the primary patent for recognizing a person by the eyes. The most trusted of them, in my opinion, is AOptix. At least their previews and documentation are not suspicious. The second company is SRI International. Even at first glance, to a person involved in iris recognition systems, their videos seem very deceitful. Although I would not be surprised if in reality they can do something. And that and that system does not show data on FAR and FRR, and also, apparently, is not protected from counterfeiting.

Face recognition

There are many face geometry recognition methods. All of them are based on the fact that facial features and the shape of the skull of each person are individual. This area of ​​biometrics seems attractive to many, because we recognize each other primarily by their faces. This area is divided into two directions: 2-D recognition and 3-D recognition. Each of them has advantages and disadvantages, but much also depends on the area of ​​application and the requirements for a particular algorithm.
In short, I'll tell you about 2-d and move on to one of the most interesting methods for today - 3-d.

2-D face recognition

2-D face recognition is one of the most statistically ineffective biometrics. It appeared quite a long time ago and was used mainly in forensic science, which contributed to its development. Subsequently, computer interpretations of the method appeared, as a result of which it became more reliable, but, of course, it was inferior and every year more and more inferior to other biometric methods of personality identification. Currently, due to poor statistical indicators, it is used in multimodal or, as it is also called, cross-biometrics, or in social networks.

Statistical characteristics of the method

For FAR and FRR, data from the VeriLook algorithms were used. Again, it has very common characteristics for modern algorithms. Sometimes algorithms with an FRR of 0.1% are flashed with a similar FAR, but the bases on which they were obtained are very doubtful (cut out background, the same facial expression, the same hairstyle, lighting).

The typical FAR value is 0.1%.
From the formula (1) we get N≈30 - the number of the organization's personnel, at which the employee's identification is quite stable.
As you can see, the statistical indicators of the method are quite modest: this negates the advantage of the method that it is possible to carry out covert shooting of faces in crowded places. It's funny to see how, a couple of times a year, another project is funded to detect criminals through video cameras installed in crowded places. Over the past ten years, the statistical characteristics of the algorithm have not improved, and the number of such projects has grown. Although, it is worth noting that the algorithm is quite suitable for guiding a person in a crowd through many cameras.

Advantages and disadvantages of the method

The advantages of the method. With 2-D recognition, unlike most biometric methods, expensive equipment is not required. With the appropriate equipment, the possibility of recognition at significant distances from the camera.
Flaws. Low statistical significance. Lighting requirements are imposed (for example, it is not possible to register the faces of people entering from the street on a sunny day). For many algorithms, any external interference is unacceptable, for example, glasses, a beard, some elements of a hairstyle. A frontal face image is mandatory, with very slight deviations. Many algorithms do not take into account possible changes in facial expressions, that is, the expression must be neutral.

3-D face recognition

The implementation of this method is a rather difficult task. Despite this, there are currently many methods for 3-D face recognition. The methods cannot be compared with each other, since they use different scanners and bases. not all of them issue FAR and FRR; completely different approaches are used.
A transitional method from 2-d to 3-d is a method that implements the accumulation of information about a person. This method has better characteristics than the 2d method, but just like it uses only one camera. When the subject is entered into the base, the subject turns his head and the algorithm joins the image together, creating a 3d template. And during recognition, several frames of the video stream are used. This method is rather experimental and I have never seen an implementation for ACS systems.
The most classic method is the template projection method. It consists in the fact that a mesh is projected onto an object (face). Then the camera takes pictures at a rate of tens of frames per second, and the resulting images are processed by a special program. A ray falling on a curved surface bends - the greater the curvature of the surface, the stronger the bending of the ray. Initially, it used a source of visible light, supplied through the "louvers". Then visible light was replaced by infrared, which has a number of advantages. Usually, at the first stage of processing, images are discarded in which the face is not visible at all or there are foreign objects that interfere with identification. Based on the images obtained, a 3-D model of the face is restored, on which unnecessary interference (hairstyle, beard, mustache and glasses) are highlighted and removed. Then the model is analyzed - anthropometric features are highlighted, which, as a result, are written into a unique code that is entered into the database. Image capture and processing time is 1-2 seconds for the best models.
Also gaining popularity is the 3-d recognition method from the image obtained from several cameras. An example of this is Vocord with its 3D scanner. This method gives positioning accuracy, according to the assurances of the developers, higher than the template projection method. But until I see FAR and FRR at least on their own base - I won't believe it !!! But it has been under development for 3 years already, and progress has not yet been seen at exhibitions.

Statistical indicators of the method

Complete data on FRR and FAR for algorithms of this class are not publicly presented on manufacturers' websites. But for the best models from Bioscript (3D EnrolCam, 3D FastPass), working according to the template projection method with FAR = 0.0047%, FRR is 0.103%.
It is believed that the statistical reliability of the method is comparable to that of the fingerprint identification method.

Advantages and disadvantages of the method

The advantages of the method. No need to contact the scanning device. Low sensitivity to external factors, both on the person himself (the appearance of glasses, a beard, a change in hairstyle), and in his environment (light, head turn). High level of reliability comparable to the fingerprint identification method.
Disadvantages of the method. Expensive equipment. The commercially available systems outperformed even the iris scanners. Changes in facial expressions and interferences on the face worsen the statistical reliability of the method. The method is not yet well developed, especially in comparison with the long-used fingerprinting, which complicates its widespread use.

Market situation

Face recognition is considered to be one of the "three big biometrics" along with fingerprint and iris recognition. I must say that this method is quite common, and for now it is preferred over recognition by the iris of the eye. The share of face geometry recognition technologies in the total volume of the world biometric market can be estimated at 13-18 percent. In Russia, this technology is also showing greater interest than, for example, identification by the iris. As mentioned earlier, there are many 3-D recognition algorithms. Most of the companies prefer to develop ready-made systems, including scanners, servers and software. However, there are those who only offer the consumer the SDK. Today, the following companies developing this technology can be noted: Geometrix, Inc. (3D face scanners, software), Genex Technologies (3D face scanners, software) in the USA, Cognitec Systems GmbH (SDK, special computers, 2D cameras) in Germany, Bioscrypt (3D face scanners, software) - a subsidiary of the American company L- 1 Identity Solutions.
In Russia, the Artec Group companies (3D face scanners and software) work in this direction - a company headquartered in California, and development and production are carried out in Moscow. Also, several Russian companies own 2D face recognition technology - Vocord, ITV, etc.
In the field of 2D face recognition, software is the main subject of development. conventional cameras do an excellent job of capturing facial images. The solution to the problem of face recognition has to some extent reached a dead end - for several years now, there has been practically no improvement in the statistical indicators of algorithms. In this area, there is a systematic "work on errors."
3D face recognition is now a much more attractive area for developers. Many teams work in it and regularly hear about new discoveries. Many works are in the “just about to release” state. But so far there are only old offers on the market, the choice has not changed in recent years.
One of the interesting points that I sometimes think about and which Habr might answer: is the accuracy of kinect enough to create such a system? Projects for pulling a 3d model of a person through it are quite possible.

Hand vein recognition

This is a new technology in the field of biometrics, its widespread use began only 5-10 years ago. An infrared camera takes pictures of the outside or inside of the hand. The pattern of veins is formed due to the fact that the hemoglobin of the blood absorbs IR radiation. As a result, the degree of reflection is reduced and veins are visible on the camera as black lines. A special program based on the received data creates a digital convolution. No human contact with the scanning device is required.
The technology is comparable in reliability with recognition by the iris of the eye, surpassing it in some way, and inferior in some way.
FRR and FAR values ​​are for Palm Vein scanner. According to the developer's data, with a FAR of 0.0008%, the FRR is 0.01%. A more accurate graph for several values ​​is not given by any company.

Advantages and disadvantages of the method

The advantages of the method. No need to contact the scanning device. High reliability - the statistical indicators of the method are comparable to those of the iris. Hidden characteristics: unlike all of the above, this characteristic is very difficult to obtain from a person “on the street”, for example, by photographing him with a camera.
Disadvantages of the method. The illumination of the scanner by the sun's rays and the rays of halogen lamps is unacceptable. Certain age-related diseases, such as arthritis, greatly worsen FAR and FRR. The method is less studied in comparison with other static biometric methods.

Market situation

Hand vein pattern recognition is a fairly new technology, and therefore its share in the world market is small and amounts to about 3%. However, there is an increasing interest in this method. The fact is that, being quite accurate, this method does not require such expensive equipment as, for example, recognition methods based on face geometry or iris. Now many companies are developing in this area. For example, by order of the English company TDSi, software was developed for the PalmVein biometric palm vein reader, presented by Fujitsu. The scanner itself was developed by Fujitsu primarily to combat financial fraud in Japan.
The following companies Veid Pte also work in the field of identification by vein pattern. Ltd. (scanner, software), Hitachi VeinID (scanners)
In Russia, I do not know any companies dealing with this technology.

Retina

Until recently, it was believed that the most reliable method of biometric identification and personality authentication is a method based on scanning the retina of the eye. It contains the best features of identification by the iris and by the veins of the arm. The scanner reads the pattern of capillaries on the surface of the retina. The retina has an immobile structure that does not change over time, except as a result of a disease, such as cataracts.
Retinal scans are performed using low-intensity infrared light directed through the pupil to the blood vessels at the back of the eye. Retinal scanners have become widespread in access control systems for highly classified objects, since they have one of the lowest percentages of denial of access for registered users and there is practically no erroneous access permission.
Unfortunately, a number of difficulties arise when using this biometrics method. The scanner here is a very complex optical system, and a person must not move for a considerable time while the system is guided, which causes unpleasant sensations.
According to EyeDentify, for the ICAM2001 scanner with FAR = 0.001%, the FRR value is 0.4%.

Advantages and disadvantages of the method

Advantages. High level of statistical reliability. Due to the low prevalence of systems, there is little likelihood of developing a way to "cheat" them.
Flaws. Sophisticated system with high turnaround times. The high cost of the system. Lack of a wide supply market and, as a consequence, insufficient intensity of development of the method.

Hand geometry

This method, quite widespread even 10 years ago, and originated from forensic science in recent years, is on the decline. It is based on obtaining the geometric characteristics of the hands: the lengths of the fingers, the width of the palm, etc. This method, like the retina of the eye, is dying, and since it has much lower characteristics, we will not even introduce it more fully.
It is sometimes believed that geometric recognition methods are used in vein recognition systems. But on sale we have never seen such a clearly declared one. And besides, often when recognizing by veins, a picture of only the palm is taken, while when recognizing by geometry, a picture of fingers is taken.

A little self-promotion

At one time, we developed a good eye recognition algorithm. But at that time such a high-tech thing was not needed in this country, and I didn't want to go to bourgeois (where we were invited after the very first article). But suddenly, after a year and a half, there were investors who wanted to build a "biometric portal" for themselves - a system that would eat two eyes and use the color component of the iris (for which the investor had a world patent). Actually, now we are doing this. But this is not an article about self-promotion, this is a short lyrical digression. If anyone is interested there is some information, but sometime in the future, when we enter the market (or not), I will write a few words about the twists and turns of a biometric project in Russia.

conclusions

Even in the class of static biometrics systems, there is a large selection of systems. Which one should you choose? It all depends on the requirements for the security system. The most statistically reliable and tamper-resistant access systems are the iris and arm veins. For the former, there is a wider market for proposals. But this is not the limit. Biometric identification systems can be combined to achieve astronomical accuracy. The cheapest and easiest to use, but with good statistics, are the fingertip systems. 2D face tolerance is convenient and cheap, but has limited applicability due to poor statistics.
Consider the characteristics that each of the systems will have: resistance to counterfeiting, resistance to the environment, ease of use, cost, speed, stability of the biometric feature over time. Let's put the marks from 1 to 10 in each column. The closer the score is to 10, the better the system is in this regard. The principles for choosing grades were described at the very beginning of the article.


Also consider the relationship between FAR and FRR for these systems. This ratio determines the efficiency of the system and the breadth of its use.


It is worth remembering that for the iris, you can increase the accuracy of the system almost quadratically, without loss of time, if you complicate the system by making it two eyes. For the fingerprint method - by combining several fingers, and recognition by veins, by combining two hands, but such an improvement is possible only with an increase in the time spent working with a person.
Summarizing the results for the methods, we can say that for medium and large objects, as well as for objects with a maximum safety requirement, the iris should be used as biometric access and, possibly, recognition by hand veins. For facilities with up to several hundred personnel, fingerprint access will be optimal. 2D face recognition systems are very specific. They may be required in cases where recognition requires the absence of physical contact, but it is impossible to deliver the iris control system. For example, if it is necessary to identify a person without his participation, a hidden camera, or an outdoor detection camera, but this is possible only with a small number of subjects in the base and a small stream of people filmed by the camera.

Note for a young technician

Some manufacturers, for example Neurotechnology, have demo versions of biometrics methods that they release on their website, so you can easily plug them in and play around. For those who decide to delve into the problem more seriously, I can recommend the only book that I have seen in Russian - "A Guide to Biometrics" by R.M. Ball, J.H. Connel, S. Pancanti. There are many algorithms and their mathematical models. Not everything is complete and not everything corresponds to modernity, but the base is not bad and comprehensive.

P.S.

In this opus, I did not go into the problem of authentication, but only touched on identification. In principle, from the characteristics of FAR / FRR and the possibility of forgery, all conclusions on the issue of authentication suggest themselves.

Tags:

• biometrics
• fingerprint scanners

Add tags

Various biometric technologies have been used for many years:

Fingerprint scanners.

The most common technology today is fingerprint recognition. The police have long used fingerprints to identify criminals. This is due to the fact that each fingerprint is unique, which makes it possible to identify each person by their fingerprints. In scanners, the print pattern is converted into a unique code. In this case, not the whole drawing is usually used, but only characteristic places - where the lines, for example, merge, bifurcate or "twist".

Finger scanning takes place using a light source - a laser or, more commonly, an LED. Then the image enters the CCD-matrix (the same as used in digital video cameras) or CMOS-matrix. Finally, the rest of the chips generate a unique code based on the print pattern. This code is stored in a database - either on a remote computer if the scanner is a cheap USB model, or in the device itself. Finally, after a person puts his finger on the scanner, the code of his fingerprint is compared with those stored in the database to identify him. Fingerprint-to-code conversion algorithms are carefully guarded.

The most attractive aspect of using this technology is its price. So, for example, protecting your computer with a mouse-like system costs less than $ 100. All this makes fingerprint scanning the most requested of all biometric technologies.

However, this system also has disadvantages. For example, fingers may be calloused or worn out after hard work; there may be dirt on them; skin may peel off. All this makes it difficult to read the fingerprint. Some people have difficulty with cheap models due to the fact that they have a weak central "curl" on the ball of the finger (devices use it to find the center of the image and generate code). Further, since the images are converted into code, it is possible that two images will get the same code. Because of this, any model is limited in the number of correctly detected prints.

Currently, a new technology is being developed that reads up to five layers of skin deep into the finger, even with worn or dirty hands. Usually, scanners are supplemented with auxiliary systems (for example, code entry) for 100% security.

Below are some of the scanners:

Biometric door lock.

Biometric lock based on attendance.

A cheap protection system for a computer, connected via a USB port.

Scanner using a magnetic card.

Hand scanners.

The second most popular devices are hand scanners. They work by measuring the parameters of a person's hands, usually fingers. This is based on the fact that each person has a different finger length. Even if there are two people with the same length of one finger, the combination of the lengths of different fingers will still be unique.

Early models of these scanners used backlighting or lasers, while modern ones are based on measuring electrical resistance, which creates fewer technical problems and requirements for energy consumption. In such a device, a person places his hand on a metal or plastic substrate, and then it measures the electric current coming from the hand. Thanks to this, the scanner identifies the areas where the hand comes into contact with the substrate, and from them calculates the length of the fingers.

Unlike fingerprint scanners, hand scanners do not require a computer - all the necessary calculations are done internally. Most models also have a keypad where a person can enter data or an identification number. The models on the market can store from 50 to 500 measurements of different hands.

This technology sometimes also has false positives - when users place their palm incorrectly on the scanning device. For this reason, most models require an identification number. In general, such devices are very "unpretentious" and are ideal for people with dirty and worn hands in factories and factories.

Eye scanners.

The iris scanning technology was first proposed in 1936 by the ophthalmologist Frank Burch. He noted that each person's iris is unique. In 1994, John Daugman of Iridian Technologies patented an algorithm to detect such differences.

Eye scanner.

Eye scanners analyze various details of the colored tissue surrounding the pupil. An ordinary video camera is used for scanning. In this case, the user looks at the device, and after scanning the iris, it compares the resulting image with those in the database. The identification process varies, but is usually less than five seconds. At the same time, only a quick glance at the device is enough for scanning.

To prevent the use of fake eyes, some models alter the flow of light entering the eye and monitor the pupil's response - a fixed size indicates a fake.

Currently, these scanners are used in various military and law enforcement agencies. However, they have not become as widespread as fingerprint scanners, despite their higher security. These scanners are usually bulky.

Retinal scanners work in a similar way, but the person must be very close to the camera. This camera captures images of tiny blood vessels in the retina, illuminated by a weak laser. It is believed that such scanners cannot be fooled, and therefore they are installed in areas with very high security. Their high cost and the need to place the eye close to the camera hinders their wider use.

Face recognition.

Never forget a face? Why? This is due to the fact that the face of each person is unique - enough to create a new technology for identifying people. However, there are serious ethical doubts about her.

Unlike all other technologies, where a person must perform some actions, this technology can do everything by itself, and a person does not even know about its existence. Her work is based on taking a picture of the face, followed by calculating the distance between the eyes, the width of the mouth and up to 50 other facial parameters. After that, a search takes place in the database of available images.

Stereoscopic face scanning system.

During the 2000 Sydney Olympics, police identified two drug traffickers from Mexico who were wanted by the United States. She tracked them to the airport and then alerted the US forces who had captured them in Hawaii. These people fell into the hands of the police by accident while visiting the stadium. At this time, the security forces had just installed a face recognition system to prevent terrorist attacks and loaded a huge database of terrorists and criminals into it. Accordingly, after the drug dealers walked through the stadium entrance - along with thousands of other people - the system found them.

A similar system is used by the London police to find criminals. Large supermarkets have also used facial recognition systems to prevent theft, but some have been removed following complaints from lawyers and customers.

Recent studies have shown that even under the best conditions, the error rate can reach 40%, which requires additional forms of identity verification. However, their ability to work with existing video surveillance systems makes them very attractive. To do this, it is enough to install a computer with a recognition program and a base of photographs and connect it to video cameras.

More reliable systems use stereoscopic cameras. This creates a three-dimensional model of the face. Thus, it becomes possible to take into account additional facial features, which significantly reduces the risk of errors.

Most systems are able to handle things like hats, glasses, and sharp angles, but this increases the likelihood of error.

Human rights defenders cite the Fourth Amendment to the US Constitution, which prohibits unwarranted surveillance and arrest. However, law enforcement agencies say this is just an increase in their surveillance capabilities.

Behavior recognition.

A technology that is often overlooked when considering biometric technologies, but which is widely used and gaining popularity, is the method of studying human behavior. Based on this, computer programs track certain behaviors, such as typing on the keyboard or making purchases.

So, lending companies have a history of people's purchases. Accordingly, any unusual orders are monitored by a computer program and a warning is issued to staff for the identity of the purchaser to be verified.

On computer networks, tracking the speed of your typing is an effective way to keep track of unauthorized users. For protection, information is also used about the Internet sites that a person visits, the programs and files used, and the peculiarities of work in general.

The big advantages of this approach are low cost, high efficiency and discreet operation.

Conclusion.

The sharp increase in identity theft and the threat of terrorism, coupled with falling prices for all biometric technologies, will lead to significant growth in the biometrics market over the next 3-5 years. Many technologies are already in use, and they are also being constantly improved to eliminate false identification.

While new inventions tend to be dissatisfied, biometric technology has led to a storm of protests from human rights groups. However, this can only slow down, but not stop the introduction of biometrics. Experts point out that additional work is needed to resolve ethical issues.

To identify citizens in biometric access control systems, parameters are used that are unique for each person.

The most common are systems that identify citizens on the following grounds:

The use of this approach for identification provides a number of advantages: the signs by which identification occurs cannot be lost or forgotten, transferred to third parties, unlike ordinary contactless cards, it is almost impossible to counterfeit or steal. However, there are also a number of disadvantages, which include the inability to 100% authenticity of identification, the relatively high cost of readers and, often, too long the identification process.

The principle of operation of such systems is based on obtaining an image from the scanner of a biometric reader and converting it into a certain template, which is then compared with the existing base. Templates can be stored both in the ACS database and in the built-in memory of the reader or in the memory of the access card.

Pattern Matching Methods and Errors That Occur

During recognition, the image obtained from a scanner or camera is compared with previously registered data (templates). There are two comparison methods:

• Identification
  The resulting image is compared with a large number of templates stored in the system database (one to many). Answers the question - who it is, takes more time, contains more errors.

• Verification
  The resulting image is compared with the registered template of a specific person (one to one). It answers the question - is it the one with whom the fingerprint is compared, much faster in time, more error-free, but requires the input of an additional identifier, for example, a card or a pin code.

At the moment of comparing the resulting image with the template, errors may appear:

• Errors of the first kind
• False Rejection Rate (FRR) when the scanner cannot recognize the registered user. They are not very critical for the security system, they create inconveniences due to the need for secondary verification. The frequency of occurrence is higher than that of errors of the second type.
• Errors of the second kind
  Erroneous acceptance of verification (FAR - False Acceptance Rate), when an unregistered user is determined by the system as registered. Critical for the security system, since an attacker can gain access to the system.

The appearance of FRR and FAR errors is determined by such characteristics as the quality and resolution of scanning, the scanning area, the mathematical algorithms used for comparison, the number of compared parts, and the ratio of these indicators makes it possible to assess the applicability of the selected identification method at the controlled object.

To eliminate errors, various sets of tests are additionally used that determine the reality of the biometric object. For finger scanners, this can be a check of the relief, pressure or temperature of the finger, for eye scanners - checking the pupil accommodation, for face scanners - a thermogram of the face.

Fingerprint

This type of identification is the most studied and cheapest; it is based on obtaining an image of the papillary patterns of the fingers of people, which have the properties of individuality, relative stability and recoverability.

There are two main algorithms for fingerprint recognition: by individual details (characteristic points) and by the relief of the entire surface of the finger, and the original fingerprint cannot be recreated from the digital code obtained as a result of processing.

The variety of biometric fingerprint readers on the market is due to the wide range of sensors (scanners) used to obtain an image. Among the innovative solutions there are contactless readers that do not require touching, and 10-finger readers at the same time, but they are quite expensive in terms of their accuracy and comfort for the user.

The disadvantage of such identification is the dependence of the quality of fingerprint recognition on the state of the finger surface and external conditions (temperature, humidity, dust), the reluctance of some people to leave their fingerprints, as well as the presence of people (about 2% of the total) with congenital poorly distinguished fingerprints.

At the moment, deep integration with BioSmart fingerprint readers is available in the Sigur ACS, which allows entering employee fingerprints into the system within the Sigur ACS software interface.

Human face geometry

In terms of technical implementation, biometric face identification is a more complex and expensive task (compared to fingerprints) and is based on the construction of two-dimensional or three-dimensional face models based on images taken by a video camera. It is the most comfortable and not always noticeable for the user; it does not require physical contact with the device.

When building a two-dimensional model, a flat image is obtained, such systems are more demanding on lighting and the position of the face when scanning, and therefore there are quite a lot of errors.

When building a three-dimensional model, a three-dimensional image is obtained, which allows achieving greater recognition accuracy by minimizing the influence of such factors as changes in skin color (including with the help of cosmetics), wearing a beard or mustache, changing the surface of the face during illness, etc., providing at the same time, sufficient speed of building a 3D-model of the face.

To increase the reliability of face recognition, you can additionally be used a facial thermogram (infrared scan of the face) that compensates for the presence of glasses, a hat, or overheads.

Iris and retina

Iris identification is one of the most reliable but expensive biometric identification technologies. The iris is unique, the most resistant to damage and does not change over time. Glasses and contact lenses do not interfere with image acquisition, even a blind person can be identified in this way.

After receiving the image, the frequency or other data about the pattern of the iris of the eye are extracted, which are saved in the template. This method is quite comfortable, since it does not require physical contact with the device and there is no bright stream of light directed into the eye.

When using a camera with a resolution of more than 1.3 megapixels, you can capture two eyes in one frame, which significantly increases the level of recognition reliability.

It is faster and more comfortable than identification by the retina of the eye and can be used on objects numbering several tens of thousands of people. Currently, it is actively developing due to its prospects.

When identifying by the retina of the eye, a pattern of blood vessels located on the surface of the fundus (retina) is used, obtained by translucent blood vessels on the back wall of the eye with a soft laser beam. The retina is one of the most stable physiological signs of the body, however, this method is very expensive, has low throughput and is not comfortable, since the user has to sit still and look through the eyepiece for several seconds. Currently, retinal identification in ACS is rarely used.

Palm veins drawing and its geometry

Identification by the pattern of the veins of the palm is based on obtaining a template when photographing the outside or inside of the hand with an infrared camera. Due to its non-contact, this method is quite comfortable for the user, while there is practically no possibility of counterfeiting, but vein diseases can complicate or distort the identification result.

The degree of recognition reliability is comparable to that of the iris, although the cost of the equipment is much lower. It is currently being actively researched.

The palm geometry identification method is based on the measurement of individual parameters of the shape of the hand, such as the width of the palm, the radius of the circle inscribed in the center of the palm, the length of the fingers and the height of the hand, and the five basic lines existing on any palm are also taken into account.

The reliability of this method is comparable to fingerprint identification and also strongly depends on the state of the object, since swelling of tissues or bruises of the hand can distort the original structure, the hands can change with age.

The advantages of using it include the lack of influence on the scanning process of temperature, humidity and pollution, although at present, identification by the geometry of the hand in the ACS is rarely used.

Barcode identification

Overview

A barcode is a set of coded numeric or alphanumeric characters in the form of geometric shapes, such as a sequence of black and white stripes. Barcodes are of different types, differing in how much and what type of information can be encoded with their help.

In access control systems, the simplest barcodes are usually used, since, as a rule, there is no need to transmit a large amount of data through an identifier. Among these are Code 39, Code 128, EAN-13.

A barcode can be applied to almost any surface, for example, printed on a sheet of paper with a regular printer, or even displayed electronically on a smartphone screen. In addition, the barcode can be easily transmitted via electronic communication channels, such as e-mail or fax. This makes the use of barcodes in ACS very cheap. On the other hand, such identifiers are not protected from copying in any way, and also have low durability.

Based on these features, barcodes in ACS are usually used as one-time passes for visitors.

Readers. Interaction with Sigur ACS

There are a large number of barcode scanners. Depending on the purpose of their use, they differ in reading technology and execution method, are also characterized by resolution, scanning speed and range, as well as the connection interface.

The ACS has a number of generally accepted interfaces through which various readers can interact with controllers, this is Weigand or Dallas Touch Memory. As a rule, barcode scanners have either a USB output for connection directly to a workplace (computer), or an RS232 interface, with which the scanner can be connected to ACS controllers through a special intermediate interface converter.

Sigur ACS supports both connection options:

• directly to the client site via USB in the keyboard typing emulation mode
• to controllers via RS232 - Weigand converter. An Elsys converter can be used as such a device.

ANDidentification by contactless cards

The most widespread among identifiers used in access control systems are contactless cards. They are convenient to use, they are made in different forms and types, and the use of cryptoalgorithms in some card formats significantly reduces the risk of their copying and forgery.

The main features of this type of identifier are discussed below.

Physical execution

In addition to those presented below, there are also other versions of identifiers (tags, stickers, bolts, flasks, tags), however, they are rarely used in access control systems.

Thick cards (Сlamshell card) Proximity cards in standard sizes with a thickness of 1.6 mm. The most inexpensive identifiers, the reading range is the highest of those presented (for the EM Marine format, there is a version with an increased reading range - up to 1.5 meters). Special stickers can be used for personalization.
Thin cards (ISO card) Proximity cards in standard sizes with a thickness of 0.76 mm. The cost is slightly higher than that of thick cards, but the reading range is lower. Personalization is ideal through direct printing on the cards themselves (sublimation or retransfer).

Keyfobs are usually more expensive than cards, while having a shorter reading range. They can have a corporate design of the performance, but there is practically no possibility of personalizing such identifiers.

Compared to cards, key fobs are more resistant to physical impact - they break less and can be attached to keys.

Bracelets

The cost is slightly higher than that of key fobs, the reading range is about the same. They can be fixed on the human body, have a corporate design.

Due to the ease of wearing, they are usually used in various fitness centers, swimming pools, water parks and other sports facilities.

Communication protocol (format)

Choosing the right format has a direct impact on the security level of the entire system. At the moment, several formats have become widespread, differing in a number of parameters.

It is possible to support several formats at once with one identifier. The cheapest and widest range of readers is the EM Marine format, but it is not copy protected in any way. In contrast, Mifare identifiers are not much more expensive, but they have an internal rewritable memory, the correct use of which, in conjunction with specially configured readers, allows you to organize secure identification.

Video reviews of cards of various formats

Readers used

When organizing an access control system, you should select the equipment and identifiers used, based on the requirements for the security of the system.

The choice of a reader is determined by several basic parameters:

• Supported identifier format.
  Readers can support one or several formats at the same time. When choosing a Mifare identifier, readers need to keep working with it in a secure mode
• Synchronization capability.
  It is used to exclude the influence of readers on each other when installed at a close distance, for example, when installing on thin walls
• Reading range.
  The standard range is no more than 10 cm, but there are readers with an increased range of up to 1.5 meters
• Output interface.
  Due to the longer range and noise immunity, Wiegand is the most preferred. The Dallas Touch Memory (iButton) communication interface has lower performance, and OSDP is still rarely used. There are readers with a proprietary interface that work with controllers only from the same manufacturer. Readers with a USB interface can be used to connect to a PC and insert cards into the system
• Operating conditions.
  Readers can be used both indoors and outdoors, therefore they are produced with different values ​​of operating temperature, moisture resistance and vandal resistance
• Design.
  There is a huge variety of solutions on the market to suit any requirement

Below are examples of various readers popular in Russia with different characteristics.

Name	Ironlogic СP-Z2L	Prox EM Reader	Rosslare AY-K12	Ironlogic Matrix V	Prox 13	Cinintec CN560
Design
Identifier format	EM Marine	EM Marine	EM Marine, remote controls	EM Marine (extended range), Keeloq and CAME remote controls	Mifare classic	Mifare Classic, Mifare DESFire, Mifare Plus
Synchronization capability	No	there is	No	No	there is	—
Reading range	3-6 cm	10-12 cm	8 cm	EM Marine - up to 50 cm, remote controls - up to 10 m	4 - 6 cm	6 - 8 cm
Housing, use	normal, indoors	normal, indoors	moisture resistant, weatherproof, outdoor	moisture and dust proof, vandal proof, outdoor	normal, indoors	normal, indoors
Estimated price, rub	750	3 000	3 500	13 000	4 000	12 500

http://sigursys.com/biometrics_overview.php