Modern methods of biometric identification. How is fingerprint identification performed? Iris identification systems
The search for criminals and the establishment of their involvement in certain criminal acts is a top priority for police departments in all countries of the world. Fingerprints, the so-called papillary pattern, are used as indisputable evidence of the suspect's guilt. As you know, the probability of meeting people with the same lines is simply negligible. But how do we know this? In this we are helped by a special scientific discipline - fingerprinting. This is the same section of forensic science, which in our time is considered the main and most important to study. It is about him that we will talk today.
What is fingerprinting?
It is rather difficult to imagine modern forensic science without this science, and it is even more difficult to understand how the policemen of the eighteenth and nineteenth centuries conducted the investigation of crimes without a database of fingerprints. After all, fingerprinting is a method of identifying a person’s personality, in which the individuality of the prints of his fingers and palms is used.
At the moment, it is on this method that forensics is based, all fingerprint laboratories in the world work using identical technology. Although it can be said that this science is one of the youngest and least studied. Yes, yes, the method referred to in all courts is scientifically unverified. How could this happen? Now we will tell you everything in detail.
The history of fingerprinting
In fact, people have always had the idea that the patterns on the fingertips are different for each person. This was given mystical significance and used for their own purposes in Babylon and China. It was believed that if a person puts a fingerprint under any document, then he is simply obliged to fulfill the terms of the contract. Although it had never occurred to anyone to classify the papillary pattern then.
Many consider the Englishman William Herschel to be the founder of fingerprinting. At the end of the nineteenth century, he worked in India and constantly encountered cases of fraud in the preparation of financial papers. The fact is that the Indians, for the most part, were illiterate people and put just a squiggle under the contracts. However, they did not consider themselves obliged to fulfill their obligations. Therefore, Herschel, remembering the mystical significance of handprints for the Indians, introduced the condition of leaving an imprint under the contract. Surprisingly, the method worked, and Herschel received 100% compliance with the rules and conditions specified in the document. During his work, the Englishman noticed that each print is different from the other and no two are the same.
With the help of the same prints, William saved himself from constant shortfalls in the payment of wages to soldiers who also sent their relatives for money and thus received double or even triple wages. After Herschel ordered them to leave fingerprints on the register, the situation returned to normal. All this was of great interest to the Englishman, who began to seriously study various handprints. The larger the base he accumulated, the more he became convinced of how individual the patterns on a person’s hands are.
The inquisitive Englishman even took fingerprints from criminals in a local prison and put things in order there. After all, earlier many offenses went unpunished due to the inability of Europeans to distinguish Indians by their faces. As soon as the investigation began to pay attention to fingerprints, the problem resolved itself. We can say that fingerprinting was born at this very moment.
Development of fingerprinting
In fairness, it should be said that not only Herschel undertook to study the fingerprints of various people. In parallel, several other people worked on this new method. For example, one of the talented Scottish doctors, G. Faulds, quite by chance noticed fingerprints on clay items made by Japanese craftsmen. He became interested in these drawings and set out to find out how diverse they are and whether they can change throughout life. He took fingerprints from his patients, servants and just acquaintances. To his great surprise, they never repeated. In addition, they perfectly matched the marks left on glass or any other polished surface. These observations even inspired him to write a scientific article, which, however, did not attract public attention.
Not the last role in the development of fingerprinting belongs to the policeman Bertillon. He ordered his officers to fingerprint all detainees and suspects. As a result, he gathered an extensive file cabinet, which helped in solving many crimes. This was the first time in history that fingerprinting in forensic science proved to be a justified and useful method of identifying a person.
Classification of papillary patterns
Over time, databases of fingerprints taken as an experiment accumulated in many police stations, but no one knew how to classify them. In the nineties of the nineteenth century, the brother of Charles Darwin tried to combine all the known developments of various people and classify the patterns on the fingers. Francis Galton applied the basics of higher mathematics in his research and was able to deduce that the probability of coincidence of papillary patterns is one chance in sixty-four billion. It was just an incredible figure for those times.
Galton's classification had some shortcomings, but nevertheless was the first serious scientific work on this topic. The researcher identified four types of papillary lines:
- with triangles;
- without a triangle
- triangle on the right;
- triangle on the left.
The card file collected as a result of this classification was filled unevenly. Therefore, a new, more effective method was required that could be used by the police. Based on his work, Galton published a whole book, where he honestly indicated all the people whose achievements he used.
Edward Henry, an employee of the Indian police, using Galton's book, created his own fingerprint classification system, which is used by modern fingerprinting. This was a huge breakthrough in science and forensics. Henry's developments served as the basis for the work of policemen in British India and immediately increased the efficiency and effectiveness of such a difficult task as investigating crimes by several times.
Henry divided patterns into the following types:
- arcs (simple and fir-shaped);
- loops (radial and ulnar);
- swirls.
In addition, Henry singled out the delta, called the Triangle by Galton, and divided this pattern into several subspecies. The researcher deduced a number of formulas, thanks to which it was possible to effectively and accurately identify a person by fingerprints.
The first application of a new technique in forensics
Fingerprinting was first used in the trial of the Stratton brothers. They were accused of double murder, and the bloody print of one finger served as the main evidence. After checking the matches, the police brought the similarity on eleven points. This turned out to be quite enough to sentence the convicts to hanging. Surprisingly, the judge categorically disagreed with this decision, although he was forced to agree with the jurors.
The use of this technique in litigation as an evidence base caused a flurry of public criticism. First of all, the revealing article was published by Folds, the same doctor who worked on the study of fingerprints. The fact is that Folds referred to some "dampness" of the method. He tried to explain that in many people the patterns on the fingers are quite similar, and the differences are expressed in just a few papillary lines. These differences can only be seen by taking prints in the laboratory. Otherwise, experts may make mistakes.
In addition, Faulds was frightened that the reliability of the method was absolutely not in doubt. Judges, juries, police officers and lawyers everywhere have argued that fingerprinting is the only science that guarantees one hundred percent correct results. It never occurred to anyone to study science, and technology was used very inaccurately by rather illiterate policemen at that time. Nevertheless, forensic science has already realized the convenience of the new method, and it has become used all over the world.
What is fingerprinting actually based on? Why are absolutely all people on the planet so confident in this method? Let's try to figure this out.
In fact, there are not so many serious scientific works on fingerprints. What is the scientific basis for fingerprinting? Experts have only two of them:
- no identical fingerprints have yet been found in any database and file cabinet, even a computer program does not find such matches;
- The patterns on the fingers of identical twins are not identical.
These two facts were enough to turn fingerprinting into an exact science. In fact, over time, experts have more and more questions about it. For example, twenty years ago, an FBI agent sent letters to all American laboratories with fingerprints from the crime scene and handprints of the suspect. What was his surprise when the laboratories gave completely different results. This significantly shook faith in fingerprinting.
Recently, information has been published that fingerprints can change over the course of a lifetime. Previously, criminologists did not have such facts, so at the moment there are all prerequisites for not accepting the results of fingerprinting as one hundred percent proof of the suspect's guilt.
Can nature be fooled?
As soon as fingerprinting began to be used everywhere, the bandits thought about the possibility of deceiving nature, in particular, changing fingerprints. American gangsters were the first to try to do this in the thirties of the last century. Members of one of the gangs, with the help of a surgeon, cut off the skin from their fingers and hoped that they had completely got rid of past prints. But after some time, the wounds healed, and the old drawings appeared again.
Next came John Dillinger. This famous gangster in all the states burned his skin with acid, making the pads of his fingers absolutely smooth. This method also turned out to be ineffective - after a couple of months, papillary lines began to appear on the fingers.
In the thirty-fourth year of the last century, FBI agents were faced with a new attempt to avoid retribution for their crimes. The police found the corpse of a well-known gangster, but the fingerprinting of the hands testified that they had a completely different personality in front of them. The summoned agents examined the victim's hands and found numerous small cuts on them. As it turned out, the criminal was trying to confuse the investigation by scarring. But even such a radical method did not bring the desired result, it was later proved that papillary lines would appear over the cuts again after some time.
After these unsuccessful attempts to deceive nature, the criminals stopped doing radical experiments on their hands.
What is used in the detection of fingerprints at a crime scene?
In modern forensics, several methods for determining fingerprints are used. Most often, experts use the following aids:
- fingerprint powder;
- fluorescent powder;
- iodine couples.
Of course, there are others, at the moment there are more than twelve means that allow you to remove prints from different surfaces. The choice of technology by an expert depends on them.
Where are fingerprints stored?
Criminologists are well aware of such a term as "dactyloscopic card". It is these maps that form the basis of the database of papillary patterns. Usually it contains the personal details of the suspect and the prints of each finger along with the palms. Each print must be extremely clear and understandable, on the reverse side is indicated the criminal article under which the charge is made.
The fingerprint card must also contain the date of the procedure and the data of the person who takes the impressions.
Fingerprint examination: details
The appointment of a fingerprint examination is under the jurisdiction of the investigators. Under the law, they can take fingerprints and handwriting samples from suspects. All these actions are carried out in the interests of the investigation in order to identify the person's identity.
Passing fingerprinting is a fairly simple and unpretentious process. Printing ink is applied to clean and dry hands with a roller. Further, the investigator seems to roll the pads of his fingers over the fingerprint card, after receiving all the prints, the paint can be washed off with warm water and soap. Now in large cities it is becoming quite common to take fingerprints with the help of modern technical means. A special device scans the fingertips and immediately creates an electronic fingerprint card in the database. This excludes minor inaccuracies and errors.
Universal fingerprinting: myth or reality
In recent years, information about universal fingerprinting has been found in the media every now and then. This idea periodically appears in the minds of the governments of different countries. Moreover, for the first time this idea arose in the nineteenth century in England and has not yet been realized in any country in the world. After all, this proposal causes a lot of controversy among ordinary citizens. On the one hand, it will become easier to investigate crimes, and on the other hand, it violates personal human rights. In the end, universal fingerprinting remains just a possible method among many others, which, if applied, will reduce the level of world crime.
How many different passwords do you have to keep in mind - two, three, maybe more? What happens if you forget your password? Using a lot of passwords is at least inconvenient. And one in all applications is unsafe. Of course, you can partially solve the problem if you use the system on cards (contactless, smart or iButton). But after all, the card can be lost, it can be stolen, and the code typed on the keyboard can be spied on or picked up. The methods widely used today only partially solve the problem of protection against unauthorized access to a room or to computer information. The only indisputably reliable and convenient identifier can only be the user himself, his unique biometric features - the shape of the limbs, fingerprints, face, eyes, voice, etc. Biometrics is definitely the future. And yet, not so far away.
Leader among biometric identification systems
According to Western experts, 80% of the biometrics market today is occupied by fingerprint identification devices (Fig. 1). What caused it?
Rice. one.
The undisputed leadership of fingerprint identification systems.
Firstly, this is one of the most accessible and inexpensive methods, which was widely used even before the advent of computers and television. Today, the cost of some fingerprint identification systems has already exceeded the $100 mark, while the cost of devices based on other technologies hover around $1,000.
Secondly, the fingerprint identification technique is easy to use, convenient and devoid of psychological barriers that exist, for example, in systems that require exposure of the eye to a light beam.
In addition, the fact that many later identification methods are protected by a patent played an important role. For example, IriScan is the exclusive owner of iris identification technology. And the methods of fingerprinting have been known to mankind since time immemorial and have been intensively used and are being used in forensic science.
Three Approaches
To date, there are three main approaches to the implementation of fingerprint identification systems.
Let's describe them in order of appearance. The earliest and most common method today is based on the use of optics - a prism and several lenses with a built-in light source. The structure of such a system is shown in Figure 2.
Rice. 2.
Functional diagram of the SONY FIU system.
Light incident on the prism is reflected from the surface in contact with the user's finger and exits through the other side of the prism, hitting an optical sensor (usually a monochrome CCD video camera) where an image is formed.
In addition to the optical system, this SONY model has an integrated processor (Hitachi H8 with 4 MB flash memory for 1000 users), RAM for internal data processing, and a DES encryption system.
The disadvantages of such a system are obvious. Reflection strongly depends on the parameters of the skin - dryness, the presence of oil, gasoline, and other chemical elements. For example, in people with dry skin, the effect of blurring the image is observed. The result is a high proportion of false positives.
An alternative method uses the technique of measuring the electric field of a finger using a semiconductor wafer. When the user inserts a finger into the sensor, it acts as one of the capacitor plates (Figure 3). The other capacitor plate is the sensor surface, which consists of a silicon chip containing 90,000 capacitor plates with a 500-dpi reading step. The result is an 8-bit bitmap of the ridges and valleys of the finger.
Rice. 3.
Identification system based on a semiconductor wafer.
Naturally, in this case, the fat balance of the skin and the degree of cleanliness of the user's hands do not play any role. In addition, in this case, a much more compact system is obtained.
If we talk about the disadvantages of the method, then the silicon chip requires operation in a sealed shell, and additional coatings reduce the sensitivity of the system. In addition, strong external electromagnetic radiation may have some effect on the image.
There is another way to implement systems. It was developed by "Who? Vision Systems". At the heart of their TactileSense system is an electro-optical polymer. This material is sensitive to the difference in electric field between the ridges and troughs of the skin. The electric field gradient is converted into a high-resolution optical image, which is then converted into a digital format, which in turn can be transferred to a PC via a parallel port or USB interface. The method is also insensitive to the condition of the skin and the degree of its contamination, including chemical. However, the reader has a miniature size and can be built into, for example, a computer keyboard. According to manufacturers, the system has a colossally low cost (at the level of several tens of dollars).
Table 1. Various technological implementations of fingerprint identification systems
| Properties | Optical system | semiconductor technology | Electro-optical polymer |
| small size | No | Yes | Yes |
| Susceptibility to dry skin | No | Yes | Yes |
| Surface strength | average | low | high |
| Energy consumption | the average | low | low |
| Price | average | high | low |
In the simplest case, when processing an image, characteristic points are selected on it (for example, the coordinates of the end or bifurcation of papillary lines, the junction of the turns). Up to 70 such points can be identified and each of them can be characterized by two, three, or even more parameters. As a result, up to five hundred values of various characteristics can be obtained from a print.
More complex processing algorithms connect the characteristic points of the image with vectors and describe their properties and mutual position (see Fig. 4). As a rule, a set of data obtained from a fingerprint takes up to 1 KB.
 |
 |
The processing algorithm allows you to store not the image itself, but its "image" (a set of characteristic data).
An interesting question is why not the fingerprint images themselves are stored in the archive, but only some parameters obtained by various image processing methods. The answer is limited resources. The size of each snapshot is not so small, and when it comes to a user base of several thousand people, it can take too much time to download and compare the newly received fingerprint with those stored in the database. And the second reason is privacy. Users like anonymity, they do not want fingerprints to be given to law enforcement without their consent or simply stolen by intruders. Therefore, manufacturers often specifically use non-standard methods for processing and storing the received data.
For security reasons, a number of manufacturers (SONY, Digital Persona, etc.) use encryption tools when transferring data. For example, Digital Persona's U are U system uses a 128-bit key, and, in addition, all forwarded packets are timestamped, which excludes the possibility of retransmission.
Data storage and identification comparison usually takes place in a computer. Almost every hardware manufacturer, along with the system, also supplies unique software, most often adapted for Windows NT. Because Most systems are designed to control access to computer information and are primarily aimed at the average user, the software is simple and does not require special settings.
Typical solutions for protecting PCs from unauthorized access
There are quite a variety of ways to connect papillary pattern readers to a PC. Much depends on the approaches of the manufacturer and the cost of the systems. For example, SONY's FIU (Fingerprint Identification Unit) system is a ready-made complex. The remote unit contains not only a scanner, but also a device for primary processing of information and encryption of transmitted data. The FIU connects directly to the PC's serial port. Less expensive readers usually require additional hardware. For example, the SACcat system from "SAC technologes" is connected to a PC via a video capture card with an ISA connector. The video capture module is inserted into the computer case. A similar device from "Key Tronic" also uses a video capture card, but placed in a separate case, which allows the system to be used with laptops.
Readers can be made as a separate device or built into the keyboard. Such products are manufactured by National Registry, Who? Vision Systems, Digital Persona, etc.
Almost all devices are powered by an external AC source.
Photo 1.
The SACcat system allows you to control access to computer information.
Table 2. Comparative characteristics of a number of protection devices from unauthorized access to computer information using fingerprint identification methods
| Characteristic* | U.are.U by "Digital Persona" | SONY FIO and I/O Software | BioMouse by ABC |
| Type I error ** | 3% | 1% | - |
| Type II error *** | 0,01% | 0,1% | 0,2% |
| Registration time | - | <1 сек | 20 - 30 sec |
| Identification time | <1 сек | 0.3 sec | <1 сек |
| Availability of an external capture device | No | No | No |
| Encryption | Yes | Yes | Yes |
| Ability to store data | No | Yes | No |
| Source of power | USB | external | external |
| Connection | USB | serial port | parallel port |
| Price with software | 200 | 650 | 300 |
** Type I error (false reject rate) - the probability that a registered user will be denied access.
*** Error of the second kind (false acceptance rate) - the probability with which the system allows the admission of an unregistered user.
Typical solutions for protecting premises from unauthorized access
Devices for controlling access to premises are more bulky than computer readers. First, there is no need to save space on your desktop. Secondly, the readers must be autonomous, therefore, in addition to the scanner, a decision-making and information storage device, a keyboard (to increase the degree of security) and a liquid crystal display (for ease of setup and operation) are placed in one case. If necessary, a card reader (smart, magnetic, etc.) can be connected to the system. There are also more exotic models. For example, SONY has placed a speaker in the body of the device, and Mytec believes that the future lies in the integration of biometrics and iButton tablets.
In addition, devices for protecting premises from unauthorized access should provide a simple connection of electric locks and alarm sensors. They should be easily networked (presence of RS-485 interfaces). For example, if an object has several inputs, then all devices must be networked so that there is a single base. The number of system users in this case increases dramatically (up to 50,000 in the Finger Scan system).
In all devices present in this segment of the market, only optics are used. New technologies are being introduced into security systems very slowly.
All devices shown are designed for indoor use only. The surface of the scanner must be clean, so dusty warehouses, gas stations, etc. are a priori excluded. The most common applications are banking systems (access to safes, vaults of valuables), access control to various clubs and country residences, e-commerce systems.
Photo 2.
The Veriprint 2000 system allows you to control access to premises.
Table 3. Comparative characteristics of a number of devices for protecting against unauthorized access to premises using fingerprint identification methods.
| Characteristic | Finger Scan by "Identix" | Veriprint 2100 from "Biometric ID" |
| Type I error | 1% | 0,01% |
| Type II error | 0,0001% | 0,01% |
| Registration time | 25 sec | <5 сек |
| Identification time | 1 sec | 1 sec |
| Interface | RS232, RS485, TTL, I/O alarms | RS232, RS485, TTL |
| Max. number of users | 50,000 (network version) | 8 000 |
| Flash memory | 512 kV or 1.5 MV | 2 MV or 8 MV |
| Addition | LCD display, keypad | LCD display, keypad |
In the very near future, we should expect a reduction in the cost of fingerprint identification systems and, as a result, their wider distribution. Most likely, it is precisely because of their already relatively low cost, availability and ease of use that such systems will be bundled with computer equipment.
Biometric readers are ideal for building fast and convenient systems for restricting access to information, for e-commerce systems and Internet sites. And although modern equipment does not fully meet all the requirements, the price is still quite high, and the reliability does not always correspond to the declared one (this, for example, is shown in a test study of the Network Computing magazine), a number of computer equipment manufacturers are already integrating biometrics into their systems. This, for example, was announced by Compaq at the recent CeBIT-99 exhibition.
Experience shows that a surge in interest from computer companies tends to lead to increased investment in research and development and, as a result, to the emergence of new more versatile technical solutions.
Nikulin Oleg Yurievich
Olga GUREEVA
Introduction
In ancient Babylon and China, fingerprints were used as a way to authenticate a person. Various government documents were “signed” with fingerprints, their impressions were left on clay tablets and seals.
At the end of the 19th century, fingerprints began to be used in forensic science. The first algorithms for comparing fingerprints in different areas of the papillary pattern appeared. In more than a hundred years of using this technology for identification purposes, there has never been a situation where two people would have exactly the same fingerprints. However, it should be noted that there is still no scientific proof of the uniqueness of the papillary pattern of the human finger. The uniqueness of fingerprints is an empirical observation, and the lack of proof of the hypothesis in this case is explained by the exceptional complexity of its proof.
Today, in connection with the development of electronic technologies, fingerprint identification has been used not only in forensics, but also in a wide variety of areas that require effective security. First of all, these areas were:
Access control systems;
Information security (access to the network, to a personal computer, mobile phone);
Accounting for working hours and registration of visitors;
Biometric fingerprint identification.
FingerChip technology
According to the American consulting company International Biometric Group, the market for biometric systems will double between 2006 and 2010, with an annual turnover of $5.74 billion. fingerprints. The increase in demand for biometric systems is also associated with the increased attention paid today by government agencies and private companies to security issues.
This article discusses various electronic fingerprint scanning technologies, including Atmel's FingerChip thermal scanning technology.
Carrying out electronic payments;
Various social projects where authentication is required;
State projects (crossing state borders, issuing visas, controlling the flow of passengers during air transportation).
The main purpose of identity verification for security purposes is to uniquely identify an individual, that is, to confirm that a person is who they say they are. Authentication must be reliable, inexpensive, fast, and non-violent. These requirements are met by biometric identification technology based on fingerprint scanning.
Fingerprint scanning
Fingerprints are relief lines, the so-called papillary patterns, the structure of which is due to rows of scalloped skin protrusions separated by grooves. These lines form complex skin patterns (arc, loop, curl), which have the following properties:
Individuality (a different set of papillary lines that form a pattern pattern according to their location, configuration, relative position, unique in another pattern);
Relative stability (the invariance of the external structure of the pattern that occurs during the period of intrauterine development of a person and persists throughout his life);
Restorability (with a superficial violation of the skin, papillary lines are restored in their original form). There are several fingerprint recognition algorithms. The most common is the algorithm based on the selection of details. Typically, there are 30 to 40 small details in a print. Each of them is characterized by its position - coordinates, type (fork, end or delta) and orientation (Fig. 1).
A print standard is formed from a set of these characteristics.
Physiologically, a fingerprint is a relief surface of the skin containing pores.
Blood vessels are located directly under the epidermis. The morphology of the fingerprint is closely related to the electrical and thermal characteristics of the skin. This means that to obtain an image of fingerprints, you can use not only paint, but also electromagnetic energy in its various manifestations. Note that scanning
Rice. 1. Fingerprint recognition by selected details
fingerprints with well-defined papillary lines is not an easy task. Since the prints are too small, you have to use rather complicated methods to obtain a high-quality image.
All existing electronic methods for obtaining fingerprints, depending on the physical principles they use, are divided into the following types:
Optical;
capacitive;
RF;
pressure;
ultrasonic;
Optical method
Currently, there are several types of scanners designed to obtain fingerprints using the optical method:
1. FTIR scanners are devices that use the effect of frustrated total internal reflection (Frusted Total Internal Reflection). The effect lies in the fact that when light falls on the interface between two media, the light energy is divided into two parts - one is reflected from the interface, the other penetrates through the interface into the second medium (Fig. 2).
The fraction of the reflected energy depends on the angle of incidence of the light flux. Starting from a certain value of this angle, all light energy is reflected from the interface.
This phenomenon is called total internal reflection. In the case of contact of a denser optical medium (in our case, the surface of a finger) with a less dense one (for example, with the surface of a prism) at the point of total internal reflection, the light beam passes through this boundary. Thus, only beams of light that hit certain points of total internal reflection, to which the papillary pattern of the finger was not applied, will be reflected from the border. To capture the resulting light image of the finger surface, a special sensor is used.
Papillary finger pattern
Light source Depression Comb skin protrusion of skin
Rice. 2. The principle of operation of FTIR scanners
Image sensor (CMOS or CCD, depending on scanner implementation). The leading manufacturers of such scanners are BioLink, Digital Persona, Identix.
2. Fiber optic scanners (Fiber Optic Scanners) are a fiber optic matrix in which all output waveguides are connected to photo sensors. The sensitivity of each sensor makes it possible to record the residual light passing through the finger at the point where the finger touches the matrix surface.
The image of the entire print is formed according to the data read from each photosensor (Fig. 3). The manufacturer of fiber optic scanners is the Elsys consortium.
3. Electro-optical scanners (Electro-Optical Scanners) - the technology is based on the use of a special electro-optical polymer, which includes a light-emitting layer. When a finger is applied to the scanner, the inhomogeneity of the electric field near its surface (the potential difference between the tubercles and depressions of the skin) is reflected in the glow of the layer. Thus, a fingerprint image is formed. The image sensor then converts the resulting image into digital form. This type of scanner is manufactured by Security First Corp.
4. Sweep Optical Scanners - Similar to FTIR devices in almost every way, except that a finger is not simply applied to obtain an image of a fingerprint
to the scanner, but is carried out along a narrow strip - the reader (Fig. 4). As you move your finger, a series of instant photos are taken. At the same time, neighboring frames are taken with some overlap, which makes it possible to significantly reduce the size of the prism used and the scanner itself. To obtain the resulting fingerprint image, specialized software is used. The leading manufacturer of this type of scanners is Cogent Systems.
5. Roller Style Scanners - these devices are the smallest scanners. The print is captured by rolling a transparent thin-walled roller with a finger. Similar to a rolling scanner, as the finger moves, snapshots of fragments of the papillary pattern are taken with some image overlay. When scanning, the simplest optical technology is used: inside a transparent cylinder there is a static light source, a lens and an image sensor. After a complete “scrolling” of the finger, the resulting image of its fingerprint is programmatically collected (Fig. 5).
Rice. S. a) The principle of operation of the roller scanner; b) its implementation
Roller scanners are manufactured by Digital Persona, CASIO Computer, ALPS Electric.
6. Non-contact scanners (Touchless Scanners) - in these devices, the finger does not contact directly with the surface of the scanner. The finger is just applied to the scanner hole and is illuminated from below from different sides by several
light sources. A lens is located in the center of the hole, with the help of which the fingerprint image is projected onto the CMOS camera (Fig. 6).
Scanners of this type are manufactured by Touchless Sensor Technology.
We note a number of shortcomings that are inherent in optical scanners, and indicate which of them have already been fixed:
Impossibility to make them compact. This problem was until recently, but, as can be seen from the figures, this shortcoming is a thing of the past.
Optical modules are quite expensive due to the large number of components and the complex optical system. This disadvantage is also leveled today due to a significant reduction in the cost of image sensors.
There is no effective protection against dummies.
The last drawback is the most significant, despite the fact that many manufacturers have announced the implementation of protection mechanisms at one stage or another of processing a scanned image.
capacitive method
Capacitive scanners are the most common semiconductor devices for capturing fingerprint images today.
Their work is based on the effect of changing the capacitance of the p-n junction of a semiconductor when the ridge of the papillary pattern comes into contact with an element of the semiconductor matrix. There are modifications of capacitive scanners, in which each semiconductor element in the matrix acts as one capacitor plate, and the finger acts as another. When a finger is applied to the sensor between each sensitive element and the protrusion-cavity of the papillary
With a pattern, a capacitance is formed, the value of which is determined by the distance between the relief surface of the finger and the element. The matrix of these containers is converted into a fingerprint image. The leading manufacturers of this type of scanners are Infineon, STMicroelectronics, Veridicom.
The disadvantage of the capacitive method is the same ineffective protection against dummies.
RF method
RF-Field Scanners - these scanners use a matrix of elements, each of which works like a miniature antenna.
The RF module generates a low intensity signal and directs it to the scanned surface of the finger. Each of the sensitive elements of the matrix receives a signal reflected from the papillary pattern. The value of the EMF induced in each miniature antenna depends on the presence or absence of a papillary pattern ridge near it. The stress matrix thus obtained is converted into a digital fingerprint image. Since the method is based on the physiological properties of the skin, it is difficult to deceive it with an imitation of a finger. The disadvantages of the method include the need for high-quality contact between the finger and the transmitter, which can be quite hot. A well-known manufacturer of RF scanners is Authentec.
Push method (pressure)
Pressure Scanners use an array of pressure-sensitive piezoelectric elements in their design.
When you place your finger on the scanning surface, the comb-like protrusions
papillary pattern put pressure on some subset of matrix elements.
The cavities of the skin pattern do not exert any pressure. Thus, the set of voltages received from the piezoelectric elements is converted into a fingerprint image. This method has a number of disadvantages:
Low sensitivity;
Ineffective protection against dummies;
Susceptibility to damage from excessive force.
Pressure-sensitive scanners are manufactured by BMF.
Ultrasonic method
Ultrasonic scanners (Ultrasonic Scanners) scan the surface of the finger with ultrasonic waves. The distances between the wave source and the scalloped protrusions and cavities of the papillary pattern are measured by the echo reflected from them (Fig. 7). The quality of the resulting image is ten times better than any other biometric method on the market. In addition, this method is almost completely protected from dummies, since it allows, in addition to the fingerprint of the papillary finger pattern, to obtain information about some other characteristics (for example,
about the pulse).
The main disadvantage of the ultrasonic method is the high price of this type of scanners compared to optical and semiconductor scanners.
The leading manufacturer of this type of scanner is Ultra-Scan Corporation.
Rice. 7. Working principle of ultrasonic scanner
Table. Specifications for FingerChip Sensors
Feature AT77C102B AT77C104B AT77C10SA
Sensor size, mm 0.4x14 0.4x11.6 0.4x11.6
Matrix size, pixels 8x280 8x232 8x232
Resolution, bp1 (dpi) 500 500 500
Reading speed, frames/s 1780 2130 2130
Overall dimensions, mm 1.64x17.46 1.5x15 1.5x15
Supply voltage, V 3-3.6 2.3-3.6 2.3-3.6
Operating temperature, °С -40...+85 -40...+85 -40...+85
Surface wear resistance, readings 1 million 4 million 4 million
Additional features no yes yes
temperature method
Thermal Scanners - such devices use sensors that consist of pyroelectric elements that allow you to record the temperature difference and convert it into voltage.
When a finger is applied to the scanner, based on the temperature of the protrusions of the papillary pattern touching the pyroelectric elements and the temperature of the air in the depressions, a temperature map of the finger surface is built, which is subsequently converted into a digital image.
The temperature method has many advantages. These include:
High resistance to electrostatic discharge;
Stable operation in a wide temperature range;
Effective protection against dummies.
The disadvantages of this method include the fact that the image quickly disappears.
When applying the finger at the first moment, the temperature difference is significant and the signal level is correspondingly high. After a short time (less than one tenth of a second), the image disappears as the finger and sensor come to thermal equilibrium. It is this feature that was used by Atmel in the temperature scanning technology, which is reflected in the Fingertip microcircuits. Today Atmel is a leading manufacturer of thermal scanners.
FingerChip technology
The FingerChip technology uses a thermal imaging method in combination with a rolling scan, which is used in the optical rolling scanners discussed above. The broaching method makes it possible to significantly reduce the size of the sensitive matrix and make it equal in width to the resulting print, and in length only a few fractions of a millimeter. To obtain an image, you simply need to slide your finger along the narrow strip of the reader. Note that in combination with the temperature method, this method of obtaining fingerprints leaves no traces after scanning due to the short image lifetime.
The small size and low cost of the sensor, combined with effective protection against dummies, as well as reliable operation over a wide temperature range, are the hallmarks of Atmel's thermal scanning technology.
At the moment Atmel produces three types of readers: AT77C102B, AT77C104B, AT77C105A. Their main technical characteristics are presented in the table.
Rice. S. FingerChip Sensor AT77C102B
The FingerChip AT77C102B sensor (Fig. 8) is made according to the 35 µm process technology and combines readout and data conversion circuits on a monolithic rectangular CMOS substrate 1.64x17.46 mm in size. The fingerprint is read by vertical movement of the finger applied to the matrix.
The FingerChip matrix has a size of 8x280, that is, it contains 2240 temperature-sensitive elements. There is also a service non-working column designed for calibration and identification of frames. The matrix pitch is 50x50 µm, which corresponds to a resolution of 500 dots per inch with a sensitive element size of 0.4x14 mm. This allows you to obtain an image of the central region of the fingerprint that meets the requirements of the image quality specification (IQS).
The clock frequency can be set by software up to 2 MHz, providing up to 1780 frames per second, which is sufficient even when moving your finger quickly over the sensor. The resulting print is assembled from a series of frames using Atmel software.
The functional diagram of this device is shown in fig. 9.
The cycle of work for each frame is as follows:
1. Select a column of 280+1 pixels in the matrix. Columns are selected sequentially from left to right with a cyclic return to the beginning. After reset, the output starts from the leftmost column.
2. Each pixel in the column sends its temperature value as an analog signal to the amplifier array.
3. Two rows are selected at the same time (even and odd). Amplified signals from them are fed to 4-bit analog-to-digital converters (ADCs). The resulting analog values can also be used as output data (not shown in the diagram).
4. Two 4-bit digital equivalents are stored in the shift register and sent in parallel as one byte through parallel outputs Be0-3 (even line) and Bo0-3 (odd line).
On fig. 10 shows the output sequence of one frame; in fig. 11 - the sequence of frames in the active mode of operation of the I^erChyr.
In addition to the read function common to all three devices, the AT77C104B and AT77C105A models have additional navigation options (similar to the touch screen) and keystroke emulation, which allows them to be controlled.
The presence of various housings (Fig. 12) provides an opportunity to optimally choose the method of installing the sensor in the device being developed.
PCLK clocks Dots Column 1 Column 2 Column 280 Column 281
12 3 4 1&2 3&4 5&6 7&8 5 6 1119 1120 1&2 3&4 5&6 7&8 1121 1122 1123 1124 1&2 3&4 5&6 7&8
Rice. 10. FingerChip frame output
Constant integration time
Frame n Frame n+1 Frame n+2 Frame n+3
1124 beats 1124 beats 1124 beats 1124 beats
Rice. 11. FingerChip Frame Sequence
and " f Shdddddd And and
Rice. Fig. 12. Variants of PmdegSyr sensor housings according to the method of fastening and connection with the base board: a) CB01 - fastening with the help of an elastomer; b) CB08 - bonding with elastomer; c) CB02 - fastening through a connector for a flexible cable
Benefits of FingerChip Technology
The Pi^erClip technology has differences due to which it can be used in various security systems. The integrated circuit of the sensor is reliably protected from electrostatic discharges with voltage up to 16 kV.
The read strip is resistant to abrasion, withstands significant applied forces and allows you to get more
1 million prints. The operating voltage of the AT77S102V sensor is from 3.0 to 3.6 V, the power consumption is 16 mW at 3.3 V with a frequency of 1 MHz. A "sleep" mode is provided in which the reset function is enabled, the clock generator is stopped, temperature stabilization is turned off and the output signal is turned off, and all output lines are placed in a high impedance state. In sleep mode, the current consumption is limited only by the leakage current. In operating mode, the sensor is completely passive. The temperature of the applied finger is used to obtain the data. In the case when the temperature difference between the finger and the sensor becomes insignificant (less than one degree), then temperature stabilization is turned on to increase the temperature of the microcircuit and increase the temperature difference.
The main advantages of K^erClip sensors are the simultaneous use of the temperature method for obtaining iso-
image processing, the method of frame-by-frame image reconstruction, and the integration of image reading and conversion circuits on a single CMOS substrate. The integration of two circuits on one substrate reduces the cost of the device, its power consumption and increases the speed of operation.
Independent tests have shown that if a person is forced to put their fingerprint in order to gain access, uneven swiping of the sensor or profuse sweating will prevent the fingerprint image from being read.
Development and Debugging Kit
K^erChyr sensors can be purchased separately. However, to extract the standard and compare the sample to the standard, special software is required, which must either be purchased from third parties,
Rice. 13. Biometric module AT77SM0101BCB02VKE
or create your own. In this regard, the use of separate sensors becomes economically feasible only in large-scale production. For fingerprint reader applications in small and medium volume production, Atmel recommends using the AT77SM0101BCB02VKE biometrics module (Figure 13), based on the Atmel AT91RM9200 32-bit microcontroller.
To evaluate the capabilities of the AT77S-M0101BCB02VKE module and develop low-level software, the AT77SM0101BCB02VEK development kit is released (Figure 14). The kit consists of an AT77SM0101BCB02VKE biometrics module, a base board with a power supply and connectors (Ethernet, USB, RS-232, external CompactFlash, SmartMedia, NAND Flash, ISO7816 smart card), patch cables, documentation, demo software for Windows, and Linux SDK for Linux.
Rice. 14. Development kit AT77SM0101BCB02VEK
The debug kit allows you to demonstrate the capabilities of the biometrics module, as well as the development of upper and lower level software.
Summing up all of the above, I would like to note that today we are witnessing the rapid development of biometric technologies. In the field of fingerprint imaging, until recently there were only two technologies - optical FTIR and capacitive, with their own advantages and disadvantages.
Scanners using FingerChip technology not only get rid of the shortcomings of previous generation devices, but also acquired a number of particularly attractive features, such as extremely small size and low price. ■
Literature
1. Bishop P. Atmel FingerChip Technology for Biometric Security. Atmel White Paper. www.at-mel.com.
2. Maltoni D., Maio D., Jain A. K., Prabhakar S. Handbook of Fingerprint Recognition. Springer, New York, 2003.
3. Zadorozhny B. Fingerprint identification // PC Magazine/Russian Edition. 2004. No. 1.
The concept of "authentication" characterizes the verification of authenticity, for example: is Vasya Pupkin really Vasya or is it, perhaps, some kind of Petya? Is he who he claims to be? The authentication process can be done in one of three possible ways:
- based on what you know, for example, a code combination (password);
- based on what you have: key, magnetic card, key fob;
- what you are: papillary patterns, facial geometry, eye structure.
It is the third point that contains biometric authentication, which, with the development of technology, is becoming more and more relevant. How it works, what are the advantages, disadvantages and how safe it is, let's take a closer look...
A Brief History of Biometrics
Missing a lot of facts, historical events and details, the use of human biometric parameters began long before the advent of technical means. Another 100 BC. e. a certain Chinese emperor put his fingerprint as a seal on especially important prehistoric artifacts. In the 1800s, Alphonse Bertillon developed a system for identifying criminals based on their anatomical characteristics.
Over time, the police in the UK, France, and the USA began to trace intruders and suspected criminals by their fingerprints. Later, the technology found its way into the FBI. Fingerprints were the first complete human recognition system.
At the present time, biometrics has become more extensive and are a means of additional protection for technical means or a security element that is used in, for admission to a protected area, premises, etc.
Varieties of biometric authentication
Currently widely used: human fingers, face and eyes, as well as voice - these are the "three pillars" on which modern biometric user authentication rests:
There are quite a few of them, however, there are three main types of fingerprint scanners in use today:
- capacitive - measure the electrical signals coming from our fingers. The capacitive difference between the raised part of the imprint and its depression is analyzed, after which a “map” of the imprint is formed and compared with the original one;
- ultrasonic - scan the surface of the finger by sound waves that are sent to the finger, reflected and processed;
- optical - photograph a fingerprint and perform a comparison for compliance.
Difficulties in scanning can arise if hands are wet or dirty, if there is an injury (cuts, burns), if a person is disabled (arms, hands, fingers are missing).
Iris authentication
Another and fairly common form of biometric authentication is iris scanners. The patterns in our eyes are unique and do not change during a person's life, which allows us to authenticate a particular person. The verification process is quite complex, as a large number of points are analyzed, compared to fingerprint scanners, which indicates the reliability of the system.
However, in this case, it may be difficult for people with glasses or contact lenses - they will need to be removed for the scanner to work correctly.
Retinal authentication
An alternative way to use the human eye for biometric authentication is retinal scanning. The scanner shines into the eyeball and displays the structure of the blood vessels, which, like the shell, are unique to each of us.
Biometric voice authentication is making its way into consumer technology and has great promise as well. Voice recognition is now implemented by Google Assistant on Android devices, or Siri on iOS devices, or Alexa on Amazon Echo. Basically right now, it's implemented like this:
- User: "I want to eat"
- Voice assistant: "OK, here is a list of nearby cafes .."
Those. no verification of the authenticity of the user is carried out, however, with the development of technology - only the original user of the device will eat. However, voice authentication technology exists and during the authentication process, intonation, timbre, modulation and other biometric parameters of a person are analyzed.
Difficulties here may arise due to background noise, human mood, age, health, which, as a result, reduces the quality of the method, because of this it is not so widely used.
Human face geometry authentication
The last in this article and one of the most common forms of biometric authentication is face recognition. The technology is quite simple: a person's face is photographed and compared with the original image of the user's face who has access to the device or to the protected area. We can observe a similar technology, referred to as "FaceID", implemented in Apple's iPhone.
We are a bit like mom, dad, or an earlier generation of relatives, and someone like a neighbor ... Be that as it may, each of us has unique facial features, with the exception of twins (although they may have moles in different places).
Despite the fact that the technology is simple in its essence, it is rather complicated in the process of image processing, since a three-dimensional head model is built, contours are highlighted, the distance between facial elements is calculated: eyes, lips, eyebrows, etc.
The method is being actively developed, since it can be used not only for biometric authentication of users or employees, but also to catch criminals and intruders. A number of cameras in public places (train stations, airports, squares, crowded streets, etc.) are installed in combination with this technology, where the scanner has a fairly high speed and recognition accuracy.
How can an attacker fool biometric authentication?
You need to understand that when scanning certain parameters, errors may occur in the recognition algorithm. And at the same time, having certain knowledge, skills and resources, an attacker can evade certain authentication methods.
In the case of a fingerprint scanner, some of them can be fooled by:
- production of a three-dimensional model of a finger from a special material (selected based on the principle of operation of the scanner);
- using the fingers of a sleeping person, unconscious or dead;
Scanners of the iris and retina can easily be fooled by a high-quality photograph of a person printed on colored paper. However, most modern scanners are able to recognize a 2D model and distinguish it from a 3D one, in which case, you need to put a contact lens on the picture, which simulates a glare (light reflection). Watch a visual video demonstrating the process of bypassing the eye scanner on a Samsung Galaxy S8 device:
Voice scanners also have their weaknesses, which arise due to the existence of artificial intelligence and neural networks capable of imitating human voices - such systems have the ability to copy any human voice and reproduce it in a matter of seconds.
Human face scanners are not inferior in terms of vulnerability, since some of these systems, an attacker can deceive using a photograph of a person, as, for example, in the case of the Samsung Galaxy Note 8:
Getting access through a face scanner will not be difficult for twins, for example, Face ID in iPhone - it looks like this:
The main advantage and disadvantage of biometric authentication
A clear advantage of the system is convenience, due to the fact that you do not need to remember the code combination (password) or pattern sequence, think about
A clear drawback is security, due to the fact that there are a lot of vulnerabilities and the recognition system is not 100% reliable. At the same time, biometric parameters (fingerprint or iris pattern) cannot be changed, unlike a password or PIN. This is a significant drawback, because if once the data gets to an attacker, we expose ourselves to serious risks.
Considering how common biometric recognition technology is now in modern smartphones, there are several recommendations that allow you to increase the level of protection to some extent:
- most of the fingerprints we leave on the surface are thumb and index, so it's best to use your other fingers to authenticate on your smartphone;
- despite the presence of biometric verification, application or PIN - a prerequisite for complete security.
"Scientific and technical articles"- selection scientific and technical articles electronic themes: novelties electronic components, scientific developments in the field of radio engineering and electronics, articles on stories development of radio engineering and electronics, new technologies and construction methods and development electronic devices, promising technologies future, aspects and dynamics of development of all areas of radio engineering and electronics, exhibition reviews electronic topics.
The desire to protect one's own life, home, property and finances from encroachment is characteristic of every person. But the usual methods of identity verification - presenting a passport or a handwritten signature - are not reliable enough, since documents can be lost, stolen or forged using modern technologies, and signatures can be falsified. Life makes us look for new, more reliable methods.
Introduction
In the light of recent events taking place in the world, especially in connection with the growing activity of international terrorism, more and more attention is being paid to security issues. One of the most important sections of security is the identification of an individual. The task of identifying a person becomes critical even in many everyday situations. Increasingly, we are faced with cases of fraud by persons impersonating others when trying to enter hotel rooms, gaining access to a computer network or making an online purchase.
Biometric identification
One of the possible methods of identification is the biometric authentication of the subject, based on the measurement of unique and permanent personal parameters. The main characteristics of a person can be divided into two groups - behavioral and physiological. Behavioral characteristics include, for example, the manner of speaking, the style of working on a computer keyboard or handwriting, and the group of unique physiological parameters includes fingerprints, palm geometry, iris or retina, facial appearance. Practical methods of biometrics rely more on physiological characteristics, since behavioral ones are still subject to changes depending on the state of the person. For example, a cold can change not only the timbre of the voice, but also the manner of speech: even talkative people avoid unnecessary conversations.
At the same time, many parts of the human body are quite unique and can be used for identification. So, when looking for a friend in a crowd, we use some general face recognition algorithm implemented by our intellect. A more specific simplified algorithm is quite feasible with the help of a computer. A person's face is captured by a camera and certain facial shapes are matched against information held in a database.
The human eye is also a collection of many unique data. By focusing the camera accordingly, the eye can be "drawn" for comparison with the image of the iris sample. And it is possible, using an illuminating scanner, to compare the light reflected from the fundus with a "cast" of the retina. The hand is no less unique. Biometric characteristics are the geometry and topology of its surface. Fingerprints play a special role.
Fingerprints were legally accepted for personal identification over a century ago, and fingerprint identification has been actively used in criminology since the twenties of the last century. They are unique to each individual, cannot be changed, and are used where identity errors are unacceptable, such as in criminal law or when organizing access with the highest level of protection.
Historically, optical sensor systems have been used for fingerprinting, but for a long time they remained very expensive, bulky and not reliable enough. In the late 1990s, the advent of low-cost, different-principle fingerprint data collection devices led to the progress of fingerprint identification technologies from limited use to widespread use in a number of new areas.
Fingerprint scanning technologies
As already mentioned, the oldest technology is optical. Scanning a fingerprint with mini-cameras on a CCD or CMOS chip has significantly reduced the cost of identification systems. But this way of imprinting faces some intractable problems: the resulting image depends on the ambient light, distortion is possible at the edges of the image, the sensor can be "fooled" relatively easily (some cheap sensors can be "fooled" with a printed copy made on a regular copier). There are problems with the size of the scanner. The sensor cannot be smaller than the focal length of the camera. Among the main advantages of optical systems, one can once again mention the relatively low price and practical invulnerability to the effects of electrostatic discharge.
Absolutely new is the technology of using the electromagnetic field. The sensor emits a weak electromagnetic signal that follows the ridges and troughs of the fingerprint and takes into account changes in this signal to create an image of the fingerprint. This scanning principle allows you to view the pattern of the skin under the layer of dead cells, which leads to good results in the recognition of pale or faded prints. There remains the problem of the lack of an acceptable relationship between the size of the sensor and its resolution.
Another promising technology that should be mentioned is ultrasonic. The 3D ultrasound scanner measures the broken surface of the finger with a kind of radar. This scanning method can be especially useful, for example, in healthcare. It does not require any sensor readers to be touched with sterile hands, and the print is easily read even through the surgeon's rubber or plastic gloves. The main disadvantage of ultrasound technology is its high cost and long scanning time.
There are other methods, either used in the past or just being developed, but the volume of the journal article does not allow us to consider them in more detail. Let us dwell on one of the most promising methods.
Capacitive fingerprint scanning
Capacitive fingerprint scanners are made on a silicon wafer that contains an area of microcapacitors. They are arranged evenly in a square or rectangular matrix. Rectangular sensors are considered more suitable as they more closely match the shape of the print. In addition, the area on which the fingerprint image is read is expanding, therefore, the amount of information received is increasing. Among the sensors available on the market today, STMicroelectronics' TouchChip sensors have the largest reading area. The field of the chip has a size of 256 x 360 capacitors, that is, the amount of information about the imprint exceeds 92 Kb. One capacitor occupies a square area of 50 x 50 µm. From these capacitors, a sensor is formed that captures the image of a print with a resolution of about 500 dpi.
Typically, the entire silicon region is protected by a coating specially developed and patented by the sensor manufacturer. This is a very hard and durable layer that can protect silicon circuits, but is so thin that it allows the finger to get as close to them as possible. Some vendors prove the quality of the coating by publishing test results stating that the protective layer has withstood over a million contacts.
Before proceeding to a detailed description of capacitive technology, let's find out what advantages and disadvantages follow from the fact that the finger is in close proximity to the IC chip.
The disadvantage may be the possibility of damage to the sensor by electrostatic discharge. In conventional microcircuits, this danger is eliminated by the casing, but the fingerprint sensor can only be covered by an extremely thin coating. To divert the discharge, additional measures are applied, such as grounding. In modern sensors, this technology is so advanced that fingerprint scanners are able to withstand discharges in excess of 15 kV (a discharge of this magnitude, for example, from electrified clothing, is very unlikely).
But almost direct contact with the crystal provides some advantages. For example, it becomes easier to distinguish a real live fingerprint from a fake or dead one. There are a large number of characteristics of a living fingerprint that can be measured (eg temperature, blood pressure, pulse). By combining these measurements and putting them into practice, you can get a fingerprint scanner that is more resistant to fraud. The use of the appropriate software further enhances the scanner's ability to resist spoofing attempts.
There are two main types of capacitive scanning - passive and active. Both are based on charging and discharging capacitors depending on the distance from the skin of the finger at each individual point in the field and reading the corresponding value. This is possible because the dimensions of the ridges and depressions on the skin are quite large. The average ridge width is about 450 µm. The relatively small size of the capacitor modules (50 x 50 µm) makes it possible to notice and fix capacitance differences even at close points on the skin.
Passive scanning principle
In passive silicon scanners, each cell has only one of the capacitor plates. Another plate forms the surface of the finger. Scanning consists of two stages. In the first stage, when a finger touches the surface of the chip, the sensor plates are charged (usually a whole row at a time) and the voltage values on each of them are stored in the so-called sample-and-hold circuits. In the second step, when the finger is removed, the rows of sensor plates are discharged and another set of sample-and-hold circuits stores the residual voltage across the plates. The difference between the charging and residual voltages of the plate is proportional to the capacitance of the sensor cell. Sequentially, row by row, scanned and digitized cells create an image of a fingerprint. This way of accessing the wafers minimizes the need for sample-and-hold circuits to two for each row.
Such a scanner allows varying within certain limits the values of the charging and discharging potentials, as well as the delay time between the scanning stages, in order to provide the possibility of reading a fingerprint in various states (wet, dry). But even with such regulation, image control cannot be as complete as with active technology, where both capacitor plates are controlled.
Active scanning principle
The sensor cell contains both capacitor plates connected in an active capacitive backfeed circuit through an inverter (inverting amplifier) that acts as a charge accumulator: one plate is connected to the input of the inverter and the other to the output (see Fig. 1). The function of the storage device is to convert the reverse power capacitance into an output voltage that can be digitized.
Rice. 1. Active capacitive scanning
An active sensor, like a passive one, works in two stages. At the first stage, the "Reset" key closes the input and output of the inverter, resetting the circuit to its initial state. In the second stage, a calibrated charge is applied to the capacitor plate connected to the accumulator input, creating an electromagnetic field between the plates. The skin of the finger interacts with the field, changing the active capacitance. Depending on the presence of a ridge or trough of the imprint, the capacitance of the capacitor decreases or increases accordingly. The value of this resulting capacitance is digitized.
Since each of the sensor cells has its own charge accumulator, the pixels of the "picture" are addressed by random access. This allows you to use additional processing features of the fingerprint image (for example, viewing only the selected area or previewing - faster, but with a lower resolution).
Active scanning technology provides much higher resistance to external influences, has a higher signal-to-noise ratio, and therefore the sensors are able to perceive a wider range of fingerprint parameters, regardless of the condition of the finger.
Imaging and fingerprint recognition
The image of a fingerprint is usually stored in binary code, where each pixel of the pattern is described by 8 bits, that is, 256 shades of gray. In advanced scanning systems, the digital image of the print is processed using a special image enhancement algorithm. This algorithm provides feedback to the sensor to adjust scan parameters. When the sensor captures the final image, the algorithm adjusts the contrast and sharpness of the print image for the best quality.
So, after digitization, there is a clear enlarged "picture" of the fingerprint. Such an image is not very suitable for matching fingerprints, because it takes up too much memory (about 90 KB) and its processing during comparison would require increased computing power. Therefore, from this information, it is necessary to make a selection of only those information that are necessary for matching fingerprints. The result of such an operation is called a fingerprint template and has a size of 250...1200 bytes, depending on the identification method.
Fingerprint recognition methods are based on comparison with samples or on the use of characteristic details. Some systems successfully combine both methods. When identifying by sample, the selected parts of the fingerprint image are stored in the database. The recognition algorithm selects the same areas of the newly entered fingerprint and compares it with the available data for authentication. Template size - about 1 Kb.
When identifying by details, only specific places where a feature (detail) is found are extracted from the image. Usually this is either the end of the ridge or its bifurcation (see Fig. 2). The content of the template in this case is relative coordinates and information about the orientation of the part. The recognition algorithm finds and compares the corresponding details. Neither the rotation of the fingerprint nor its parallel translation (shift) affects the functioning of the system, since the algorithm works with relative values. The template size in this case is reduced to about 300 bytes. Processing such a small amount of data is possible even in systems with low processor speed and limited memory.
Recognition algorithms and their markup
There are a fairly large number of algorithms on the market that identify the image by details. It is necessary to find out what are the criteria for their quality.
If we express the correspondence of two compared fingerprint patterns as a percentage, then a perfect match (two patterns of one finger) can be assigned a value of 100%, and an absolute mismatch (two patterns of different fingers) should be indicated by zero (0%). Unfortunately, not all matches are perfect, and mismatches are absolute. Typically, the degree of coincidence does not fall on the extreme points of the scale. There is a problem with inexact and incomplete matches. It is most difficult to compare similar patterns, since the values of the groups of estimated values for matches and mismatches overlap, overlap each other in the middle of the scale. This is a critical area, since in such a case it is impossible to decide exactly whether the patterns matched or not. The way out of such a "schizophrenic" situation is to establish the so-called "threshold", which uniquely determines the value of the assessment that separates the pattern match from the mismatch. This makes it easier to make a decision, but, on the other hand, can lead to errors in the system, since both groups of estimated values may be below the established limit.
Rice. 2. Details of the imprint
Such errors are called misidentification and misidentification, respectively. The degree of such errors is specific to each recognition algorithm and is usually taken into account as FMR (False Match Rate) - the probability of erroneous recognition and FNMR (False Non-Match Rate) - the probability of erroneous non-recognition. In security systems, they are also commonly called FAR (False Accept Rate) - the probability of an erroneous admission and FRR (False Reject Rate) - the probability of a false rejection. FMR and FNMR are mutually opposite: when one value decreases, the other increases (which is equivalent to moving the "threshold" up and down the compliance scale). The quality of recognition algorithms can be assessed by comparing the FMR value with a fixed FNMR or vice versa. Sometimes additional parameters are given for evaluation, for example, the level of equiprobable error - the point on the compliance scale where the values of FMR and FNMR are equal.
| Specifications | Sensors | |||
|---|---|---|---|---|
| TCS1AD | TCS2AF | |||
| Sensor active zone, mm | 18.0x12.8 | 10.4 x 14.4 | ||
| Total area, pixels | 256 x 360 | 208 x 288 | ||
| Pixel area, µm | 50 | |||
| Resolution, dpi | 508 | |||
| Information retrieval frequency, frame/s | 15 | 20 | ||
| Maximum static potential, kV | ±8 | ±15 | ||
| Current consumption | Nominal, mA | 20 | ||
| Stand-by, mA | 7 | |||
| Sleep, mA | 1 | |||
| Case dimensions | Full, mm | 27 x 27 x 4.5 | 27 x 20.4 x 3.5 | |
| Compact, mm | 27x18.4x4.5 | |||
| Connector | Flexible cable | 20-pin flexible connector/Flexible cable | ||
| I/O interface | 8-bit RAM interface | |||
| Environmental characteristics | Working temperature, °С | 0...40 | ||
| Storage temperature, °C | -4...85 | |||
| Humidity | 5...95%RH @ 30°C | |||
The values of the above characteristics are highly dependent on the fingerprint database used in testing the recognition algorithm to evaluate its quality. You can get very good results even with a weak algorithm if only high quality prints are selected for testing. Naturally, even a successful algorithm can give poor results on a database containing only low-quality fingerprints. Therefore, a comparison of recognition algorithms can be carried out only on the condition that the same base is used for their testing. Testing an algorithm, determining its control points - threshold, FMR, FNMR, etc. - is called markup. To obtain useful and realistic marking results, it is necessary to use the largest possible database of fingerprints (at least thousands of people), which would be collected in different regions of the world from representatives of different races, ages and occupations in various conditions (humidity, temperature, etc.). .).
The future is a united module
Fingerprint recognition technology has many advantages, which explains the increasing scope of its application. Already today there are laptops, PDAs, door locks, vending machines and various computer peripherals with built-in fingerprint sensors. Advances in technology are reducing the size and cost of sensors, opening their way to many other uses, such as mobile phones, point of sale terminals or car ignitions.
Rice. 3. Biometric security system STTouchChip
STMicroelectronics offers the ST TouchChip, a turnkey biometric security subsystem that can be easily implemented in general and private products (see Figure 3). TouchChip, PerfectPrint and PerfectMatch are state-of-the-art technologies that provide a full range of typical biometric system functions: fingerprinting, image optimization and access decision making. TouchChip - silicon fingerprint sensor - captures fingerprint images. It is based on the company's patented active capacitive pixel sensor technology, which provides a high signal-to-noise ratio. The PerfectPrint software suite manages the sensor to optimize the fingerprint image based on environmental conditions or skin type. PerfectMatch is a set of software algorithms that solve two essential biometric tasks: extracting templates from a fingerprint image and recognizing the correspondence of live fingerprints to pre-stored images.
PerfectMatch comes with an application programming interface (API) that allows you to integrate TouchChip biometric subsystems into customer designs without detailed knowledge of all system components. This open architecture greatly simplifies the integration of the biometrics system into existing applications and reduces implementation time.
The goal of further development is to combine a fingerprint sensor with a powerful microprocessor and memory. This will make it possible to create a recognition module capable of performing the entire task: from reading a fingerprint to identifying an object - without a computer. Similar projects are already being developed. STMicroelectronics recently announced a device called the TouchChip Trusted Fingerprint Module Biometric Subsystem, due by the end of 2002. Such an integrated module will eliminate the effort currently expended on integrating individual components, which will give an even more significant boost to the entire biometric fingerprint recognition market.
Fingerprint identification will soon become part of our daily lives. Let's hope for the increased security and convenience it will bring.
Publication date: 01.09.2004
Readers' opinions
- vlab / 04.08.2013 - 00:41
An incomplete fingerprint can identify a person - Oleg) / 21.11.2012 - 10:59
Interesting article) Consistently planned, and quite easy to read. It was nice to read. - Anatoly / 18.12.2008 - 14:31
Need a diagram! - Maxim / 08.07.2007 - 19:17
In general, the article looks good. From the point of view of the interested user, a very intelligible interpretation of the topic. If limited to more professional knowledge, then the lack of specifics in the very essence of the problem, for example, the recognition algorithm. The approach used in such systems is very interesting. I would be very happy if such material was also posted on a new site. Good luck!
Where are the boundaries between these categories of consumers?
What is an asset directory
Free programs for Windows free download
Realtek Audio Driver (Realtek HD Audio)
Realtek Audio Driver (Realtek HD Audio)
RK account not found what to do
Does ureaplasma pass by itself (can it pass on its own)?