Biometrics involves a system for recognizing people based on one or more physical or behavioral traits. In the field of information technology, biometrics are used as a form of access identifier management and access control. Biometric analysis is also used to identify people who are under surveillance (widely used in the USA, as well as in Russia - fingerprints)
Encyclopedic YouTube
1 / 2
✪ The biometric reader does not work well. Why?
✪ Biometric systems. Anatoly Bokov, Sonda Technology
Subtitles
To enter fingerprints into the database in general, Timex biometric terminals. In principle, it offers you 3 fingerprints per finger, roughly speaking 3 submissions. It is recommended to bring your finger, first turning a little to the left, to the middle and to the right, so that the entire surface is covered. Because very often they enter fingerprints incorrectly, and then they wonder why it doesn’t work well. Therefore, there is also a point here that you also need to enter them correctly. The more carefully and correctly you enter them, the fewer problems with prints are possible. There are also times when people, customers, install biometric time and attendance terminals with built-in controllers on a high-traffic turnstile. Well, this is also such a dubious story, because, especially for various industries, people there have such imprints that problems definitely arise.
Basic principles
Biometric data can be divided into two main classes:
- Physiological- relate to the shape of the body. Examples include: fingerprints, facial recognition, DNA, palm of hand, retina, smell, voice.
- Behavioral- related to human behavior. For example, gait and speech. Sometimes the English term is used for this class of biometrics. behaviometrics.
Definitions
Basic definitions used in the field of biometric devices:
- Universality - every person must have a measurable characteristic.
- Uniqueness is how well a person is distinguished from another from a biometric point of view.
- Persistence is a measure of the extent to which selected biometric traits remain unchanged over time, such as during the aging process.
- Collections - ease of measurement.
- Productivity - accuracy, speed and reliability of the technologies used.
- Acceptability is the degree of reliability of the technology.
- Elimination - ease of use replacement.
The biometric system can operate in two modes:
- Verification - one-to-one comparison with a biometric template. Verifies that the person is who he claims to be. Verification can be done by smart card, username or identification number.
- Identification - one-to-many comparison: after “capturing” the biometric data, a connection is made to the biometric database to determine the identity. Personal identification is successful if the biometric sample is already in the database.
The first private and individual application of the biometric system was called enrollment. During the registration process, biometric information from the individual was stored. Subsequently, the biometric information was recorded and compared with the information obtained previously. Please note that if a biometric system is to be secure, it is essential that storage and retrieval within the systems themselves be secure.
- False Acceptance Rate (FAR), or False Match Rate (FMR)
FAR - false admission rate, the probability of false identification, that is, the probability that the bioidentification system mistakenly recognizes the authenticity (for example, by fingerprint) of a user who is not registered in the system
FMR is the probability that the system incorrectly compares an input pattern to an unmatched pattern in the database. - False reject rate (FRR), or false negative rate (FNMR)
FRR - false access refusal rate - the probability that the bio-identification system does not recognize the authenticity of the fingerprint of the user registered in it.
FNMR is the probability that the system will make an error in identifying matches between the input pattern and the corresponding pattern from the database. The system measures the percentage of valid inputs that were received incorrectly. - System operating characteristic, or relative operating characteristic (ROC)
The ROC plot is a visualization of the trade-off between FAR and FRR performance. In general, the matching algorithm makes a decision based on a threshold that determines how close the input sample must be to the template to be considered a match. If the threshold was reduced, there would be fewer false non-matches, but more false accepts. Accordingly, a high threshold will reduce FAR but increase FRR. The line graph shows the differences for high performance (fewer errors - fewer errors). - Equal error rate (EER) or transient error rate (CER) are the rates at which both errors (receive error and reject error) are equivalent. The EER value can be easily obtained from the ROC curve. EER is a quick way to compare the accuracy of instruments with different ROC curves. In general, devices with low EER are the most accurate. The lower the EER, the more accurate the system will be.
- Failure to Enroll Ratio (FTE or FER) is the rate at which attempts to create a template from input data are unsuccessful. Most often this is caused by low quality input data.
- False Hold Rate (FTC) - In automated systems, this is the probability that the system is unable to detect biometric input data when it is presented correctly.
- Template capacity is the maximum number of data sets that can be stored in the system.
As the sensitivity of biometric devices increases, FAR decreases and FRR increases.
Tasks and problems
Confidentiality and separation
Data obtained during biometric registration may be used for purposes for which the registered individual did not consent (was not aware).
Danger for owners of protected data
In cases where thieves cannot gain access to protected property, there is the possibility of tracking and assassinating the bearer of biometric identifiers in order to gain access. If something is protected by a biometric device, the owner could suffer irreversible damage, possibly costing more than the property itself. For example, in 2005, Malaysian car thieves cut off the finger of the owner of a Mercedes-Benz S-Class while trying to steal his car.
The use of biometric data is potentially vulnerable to fraud: biometric data is somehow digitized. A fraudster can connect to the bus leading from the scanner to the processing device and obtain complete information about the scanned object. Then the fraudster will not even need a living person, because, having connected to the bus in the same way, he will be able to carry out all operations on behalf of the scanned person, without using the scanner.
Cancellable biometrics
The advantage of passwords over biometrics is the ability to change them. If your password is stolen or lost, you can cancel it and replace it with a new version. This becomes impossible with some biometric options. If the parameters of someone's face have been stolen from the database, then they cannot be canceled or new ones can be issued. Cancelable biometrics are the way to go, which should include the ability to cancel and replace biometrics. It was first proposed by Ratha et al.
Several cancelable biometrics methods have been developed. The first cancelable biometric system based on fingerprints was designed and created by Tulyakov. . Basically, cancelable biometrics is the distortion of a biometric image or properties before they are agreed upon. The variability of distorted parameters carries with it the possibility of cancellation for a given circuit. Some of the proposed techniques work using their own recognition mechanisms, like the work of Teo and Savvid, while others (Dabba) take advantage of promoting well-presented biometric research for their recognition interfaces. Although security restrictions are increasing, this still makes overrideable models more accessible to biometric technologies.
One of the private solutions may be, for example, not using all biometric parameters. For example, for identification, the pattern of papillary lines of only two fingers (for example, the thumbs of the right and left hands) is used. If necessary (for example, if the pads of two “key” fingers are burned), the data in the system can be corrected so that from a certain moment the valid combination will be the index finger of the left hand and the little finger of the right hand (data that was not previously recorded in the system - and could not be compromised).
International exchange of biometric data
Many countries, including the United States, are already participating in the exchange of biometric data. This statement was made in 2009 in the Committee on Appropriations, Homeland Security Subcommittee on "biometric identification" by Kathleen Kraninger and Robert Mockney:
To ensure we can stop terrorist organizations before they reach the United States, we must take a leadership role in promoting international standards for biometrics. By developing interoperable systems, we will be able to securely transfer information about terrorists between countries, maintaining our security. Just as we are improving the way we cooperate within the US Government to identify and eliminate terrorists and other dangerous individuals, we also have a commitment to our partners abroad to work together to prevent any terrorist activity.
We understand that through biometrics and international cooperation, we can change and expand travel options and protect people around the world from those who would do us harm.
According to an article published by S. Magnuson in National Defense Magazine, the US Department of Homeland Security is under pressure to distribute biometric data. The article says:
Miller (a consultant to the Department of Homeland Security and Security Affairs in America) reports that the United States has bilateral agreements on the exchange of biometric data with 25 countries. Every time a foreign leader has visited Washington over the past few years, the State Department has made sure to negotiate a similar treaty with them.
Legislative regulation in Russia
Article 11 of the Federal Law “On Personal Data” No. 152-FZ of July 27, 2006 regulates the main features of the use of biometric data.
Biometrics in popular culture
Biometric technologies have been featured in popular movies. This alone has already aroused consumer interest in biometrics as a means of identifying a person. The 2003 films "X-Men" and "Hulk" used biometric recognition technologies: in the form of handprint access in the film "X-Men" and fingerprint access in "Hulk".
But this was not so significant until the film “I, Robot” starring Will Smith was released in 2004. The futuristic film demonstrated the development of new technologies, which even today are not yet sufficiently developed. The use of voice and palm recognition technologies in the film captured people's vision of the future, and both of these technologies being used today to secure buildings or information are just two of the possible uses of biometrics.
In 2005, the film “The Island” was released. Twice during the film, clones use biometric data: to break into a house and start a car.
The film "Gattaca" depicts a society in which there are two classes of people: products of genetic engineering created to be superior (the so-called "Valid"), and inferior ordinary people ("Disabled"). People considered "Valid" had great privileges, and access to restricted areas was restricted to such people and controlled by automatic biometric scanners, similar to fingerprint scanners but pricking a finger and obtaining a DNA sample from the blood taken.
In the movie Destroyer, the character Simon Phoenix, played by Wesley Snipes, cuts out a victim's eye to open a door with a retinal scanner.
In DreamWorks' Monsters vs. Aliens, a military assistant infiltrates the zone using biometrics.
INTRODUCTION
Issues of studying living organisms and plant objects, as well as processes occurring at the cellular, molecular and genetic level, are becoming more and more relevant every day. For this purpose, scientific laboratories are developing methods for their study and simulating complex natural phenomena. The most commonly used research methods include experimental and multivariate statistical methods. They are an important and integral part of a laboratory experiment and make it possible to reliably identify the patterns of natural processes occurring, as well as find cause-and-effect relationships between them.
In scientific research, the method of mass observation is effectively used to obtain reliable data. This method is based on the use of a large number of replicates in each experimental group. The material obtained during the laboratory experiment is processed and analyzed, then, based on the data obtained, appropriate conclusions are drawn and certain patterns are established. Of great importance in achieving the greatest accuracy of results and conclusions during an experiment is not only the quality of experimental methods, but also correct statistical processing, since the results obtained can vary significantly within one experimental group. Thus, performing a statistical analysis of experimentally obtained data expands the possibilities for knowledge of biological natural phenomena, contributes to an objective assessment of the results obtained, excluding the possibility of a subjective point of view of the researcher, as well as methodological errors that arise during the experiment, and allows the experimenter to draw accurate and correct conclusions and conclusions regarding the phenomenon being studied.
Item research – computer technology as a method of processing data obtained from laboratory research.
Target research – analyze the capabilities of statistical programs when processing data obtained as a result of a laboratory experiment.
Tasks research:
· Evaluate the methods of mathematical statistics in terms of their capabilities and limits of application when planning and processing a biochemical experiment.
· Study available statistical analysis packages.
· Master the ability to solve problems of applied statistics using Microsoft Excel (using standard functions and data analysis packages) and well-known statistical packages STATISTICA in the field of biochemistry.
Computer technologies are of great importance in statistical data processing. This allows not only to speed up this process several times, but also to produce it at a higher quality level.
THEORETICAL ASPECTS OF THE USE OF COMPUTER TECHNOLOGY WHEN CONDUCTING LABORATORY RESEARCH
Biometrics as a science and its basic concepts
In recent years, computer technologies have been increasingly used to solve and simulate problems. In this regard, the need for highly qualified specialists with a good theoretical basis and experience in working with certain programs has increased. Today, disciplines are appearing in educational institutions that make it possible to develop sustainable skills necessary for processing and presenting the results of scientific activity. The science that deals with the study of methods for collecting and interpreting numerical data is called statistics . This discipline has important practical significance, as it allows one to predict the development of natural, social processes and phenomena. Over time, more specialized branches of this science began to appear. Thus, at the junction of two independent sciences: biology and statistics, there appears biological statistics (or biometrics) . Biometrics is an empirical science that studies data obtained during an experiment by performing some mathematical calculations. Performing these operations without computers and computer technology takes a lot of time. We can see how labor-intensive this process is by considering some of the most used concepts of biometrics when characterizing the trait under study.
Basic concepts of biometrics.
Very often in practical human activity and when processing data obtained during scientific research, an average value is used. This value characterizes the characteristic under study and shows what the value of the variable would be if all objects from the sample had the same value. The arithmetic mean is calculated using the formula:
where x 1 x 2, ..., x k - population options; n is the total number of options.
Median (50% interval limit)- a value that divides the sample in half: the same number of options is located on both sides of the median in the variation series. This value depends on the accumulation of frequencies. Frequencies accumulate until half the sum of frequencies is exceeded. The resulting largest value is the median. The formula by which this value can be calculated is as follows:
,
where x min is the minimum value of the interval limit where the median value is located; i - interval value; N-volume of the population; Σn is the total number up to the interval in which the median value is located; N e is the number of intervals where the median value is located.
Another statistical indicator is fashion. Fashion The value that occurs most frequently is called. The mode can be calculated using the Pearson formula:
,
where Me is the median; M is the average value of the feature.
Standard deviation,- the most important characteristic in a biological experiment. This value is a measure of the scattering of the distribution series and is determined by the formula:
Some experiments require very high experimental accuracy. For example, in medical-biological, toxicometric, etc. The error in these experiments should not be higher than 1%; if the error value exceeds 1%, then the accuracy of the result is unsatisfactory and the number of repetitions must be increased.
However, no matter how hard the researcher tries to accurately carry out all the steps of the experimental procedure, errors still occur in practice that must be taken into account when processing data. There are several types of errors.
Mean error (m x)- an indicator by which the average value of the sample (experimental) population differs from the average value of the general population, if the distribution of the parameter under study tends to a normal value. The main error of the mean is calculated using the formula:
More informative and acceptable for comparison of groups is used coefficient of variability, or variations. The coefficient of variability is the main deviation, expressed as a percentage of the average value, which is calculated by the formula:
Based on the results obtained, a conclusion is drawn about the nature and degree of variation of the trait (Table 1.1).
Table 1.1. The nature of the variability of traits (according to M.L. Dvoretsky)
If the t value is greater than four, then the average value will be reliable and, accordingly, correct conclusions can be formulated.
The percentage of discrepancy between the sample and general averages is also determined - accuracy of experience (p,%), or observational error:
This experimental parameter shows by how many percent one can be mistaken if one asserts that the general average is equal to the obtained sample average.
In statistics, the rationing indicator is important. This indicator is used to evaluate an option relative to the average value of a given group using the following formula:
Depending on the purpose of the study, the value can range from x: ±0.5σ to x±1σ. Options with a value from 0.67σ to 2σ are subnormal if the value is more than x± 2σ , then such options should be classified as anomalies.
In biometrics there is such a thing as representativeness error. This is an error that occurs not during measurements or calculations, but due to random selection when forming a group.
When calculating the error of the arithmetic mean in small groups, the number of observations (P) is the “number of degrees of freedom” - the expression (n-1) is used, and then the formula looks like:
There are a huge number of formulas for calculating experimental errors. Some of them are given below as an example. The formula for calculating the average error of the standard deviation:
Average error of coefficient of variation (C):
Average error of the asymmetry indicator:
Or more precisely: 
Error of kurtosis coefficient:
A comparative analysis of the results obtained comes down to assessing the degree of reliability of the differences observed between them using the following formula:
where t is the reliability criterion. Its value is estimated using Student's probability tables. If the actual t is greater than the tabulated t st , then there is a difference between the two study groups. The difference is significant, reliable and cannot be explained by random reasons.
To compare the results obtained with the expected ones, use the chi-square test (χ 2), which is found by the formula:
where p is the empirical frequency, p’ is the expected frequency. The meaning of the χ 2 test is to find out whether the hypothesis is confirmed or refuted by the experiment. If the values of χ 2 exceed the tabulated value, then it can be argued that the difference between the actual and expected results will be reliable.
Since most biological objects have a huge number of often interrelated characteristics that characterize them, for example, weight, height, age, etc., analysis of variance is used when studying a set of indicators. A relationship in which for each value of the independent variable there is only one value of the dependent variable is called functional. However, in nature such a connection is very rare. Typically, studied objects with the same values of one characteristic have different values for other characteristics. This connection is called correlation. Coefficient correlations shows how one characteristic under study is related to another (Table 2). The correlation coefficient is calculated using the formula:
Table 1.2. Characteristics of the closeness of connections between characteristics
It is also necessary to find the square error of the correlation coefficient:
The obtained indicators of the correlation coefficient are assessed using the Student’s reliability criterion:
Or using the formula
When assessing the relationship between quantities, it is very important to find an analytical equation that will correspond to the nature of the phenomenon being studied in order to predict the behavior of an independent characteristic of an object when the dependent parameter changes. The relationship between variables is called regression. Regression coefficient, which is determined by the following similar formulas:
- regression coefficient Y.X;
regression coefficient X.Y,
And .
The mean square error is also found for the regression coefficient:
These are the basic formulas used in biometrics, which are used when processing data obtained during biochemical research. There are many more statistical formulas, but all of them, as we have already seen, consist of several mathematical operations, which complicates the researcher’s calculations and can lead to numerous errors in the calculations. Correcting these errors can be time-consuming when processing large amounts of data. Thus, computer technology simplifies this routine process several times, which allows for more efficient use of time, and also reduces the likelihood of error, which gives confidence in the correctness of the results obtained and allows one to draw correct conclusions.
Planning and processing of a biochemical experiment
Currently, there is a lot of information and it is quite difficult to navigate this endless stream of knowledge. Then the question arises of how you can obtain the information of interest and select the necessary literature, while spending a minimum amount of time. For this, there are various search engines that significantly reduce the amount of time spent at the preparatory stage. Since before starting to carry out and plan a study, it is necessary to make sure whether this issue has been studied previously, what are the results of the studies conducted and what criteria have already been studied. In order to fully understand the need for information technology in experimental planning, it is necessary to understand what this process is.
Experimental planning is a set of measures aimed at effectively setting up an experiment, the main goal of which is to achieve maximum measurement accuracy while conducting a minimum number of experiments. When planning an experiment, there are several stages:
1. Pre-planning - this stage includes drawing up a work plan and its approval, choosing a topic, formulating a working hypothesis, information processing of the plan and mastering techniques.
This stage eliminates the possibility of duplicating research, ensures the reliability of knowledge and an original approach to solving the problems assigned to the researcher.
2. The actual research process - at this stage, an analytical review of the literature on this problem is carried out, data is accumulated, their systematization and the development of ideas and conducting an experiment. An experiment is a set of actions and observations performed to test the truth or falsity of a hypothesis and establish cause-and-effect relationships between the phenomena being studied.
Thanks to this stage, the researcher can realize how new this topic is and how relevant the results are, and formulate scientific and practical significance.
3. The last stage is to formalize the results of scientific research - compiling reports, writing articles.
Any experiment is based on the performance of an analytical method. Analytical methods have criteria that determine the suitability of the method:
· Specificity – the ability to determine the component for which this research method is intended.
· Accuracy - the quality of measurements reflecting the closeness of the results obtained containing the analyte
· Convergence (reproducibility in a series) is the idea of the closeness to each other of the results of a study performed under the same conditions in a series.
· Reproducibility – the closeness of the results obtained when performing a laboratory analytical study of a sample under various conditions. This parameter reflects the degree of data scatter and allows you to identify random errors.
· Right and wrong - differences from the true meaning
· Sensitivity – the ability of the method to detect the lowest value of the analyte. The magnitude of the difference ratio between the measurement indicators of the device is estimated. The higher the ratio, the higher the sensitivity of the method.
· Limit sensitivity – the concentration of the test substance corresponding to the minimum measurement different from the value of the blank sample.
Interpretation of the research results is carried out manually or using a computer. One way to evaluate the results is to construct a graduated (calibration) curve. The calibration curve shows the close relationship between extinction, light intensity and concentration of a substance in a series of standard solutions. Standard solutions are used to construct a graduated curve.
Constructing a calibration curve:
ü Preparation of standard solutions
ü Preparation of a dilution of a standard substance that covers the range of concentrations being studied and goes beyond the maximum and minimum values.
ü From the main one we prepare stock solutions
ü For each concentration of the standard solution we make 3-5 measurements
ü Using the obtained points, we build a graph.
For greater clarity and accuracy, it is best to build a graph. The graph shows the dependence of optical density on the concentration of the solution. This will be more convenient for the subsequent determination of the concentration of the substance under study in the test samples, which will help to calculate a more correct concentration of working solutions.
Related information.
There are about 14 million iPhones in Russia, a third of them with the Touch ID function. To unlock your smartphone screen, you share biometric data with Apple. Users are increasingly giving away intimate body data. It seems convenient, reliable and helps in the fight against crime. Although the recent Indian incident with journalists who gained access to the biometric data of millions of fellow citizens suggests the opposite. As technology improves, lawmakers around the world are grappling with how to regulate and protect the collection and use of biometric information.
What is biometrics?
Biometrics analyzes the physical and behavioral characteristics of people to identify their identity. In the simplest sense, these are technologies for measuring the human body. There are two categories of biometric measurements: physiological and behavioral.
Physiological measurements There are two types: morphological and biological. Morphology includes fingerprints, the shape of the hand, fingers or face, the pattern of the iris and retina; for biological tests - DNA, saliva, blood or urine tests.
Behavioral measurements- this is voice recognition, handwriting dynamics (movement speed, acceleration, pressure, tilt), keystroke dynamics, way of using objects, gait, sound of steps, gestures.
These measurements can be used in two different ways: for identity verification and for identification.
Verification involves comparing biometric data with a specific template contained in a database or on a portable medium, such as a smart card. This operation allows you to make sure that the person is exactly who he claims to be.
When identification a person's biometric data is compared with the data of other people in the database. Identification is successful if such a biometric sample is already in the database.
Biometrics - a new phenomenon?
Not really. In the 19th century, French lawyer and policeman Alphonse Bertillon began comparing physical characteristics of people to identify criminals. The anthropometry system he developed became the first scientific approach to determining personality in criminology. His developments formed the basis of fingerprinting, a system for identifying a person based on fingerprints. The well-known system was invented by the British officer William Herschel - in 1877, he put forward a hypothesis about the immutability of the papillary pattern on the palms of a person. Fingerprint identification of criminals was first used in 1902.
Behavioral biometrics also has its roots in the 19th century: in the 1860s, telegraph operators used Morse code to recognize each other by the transmission of “dots” and “dashes”.
Where is biometrics used today?
Mainly in the field of national security, healthcare and registration systems. Biometrics are widely used by companies to monitor employees and internal security, banks - to identify clients, corporations and social networks - for commercial purposes.
As in the 19th century, today law enforcement agencies use biometrics to identify criminals. Automated fingerprint identification systems (AFIS) process and store fingerprint images, while automated biometric identification systems (ABIS) contain templates for faces, fingers, and irises. Large cities, airports and borders are already using live face recognition technology, which makes it possible to identify a face in a crowd in real time.
Border control uses electronic and biometric passports, which, in addition to the owner’s photograph, also contain two fingerprints. The biometric infrastructure consists of fingerprint scanners and cameras that speed up border crossings. States are introducing these technologies to control migration flows.
Biometrics are also needed to create ID cards that provide access to health care, civic identification, and voter registration.
A huge number of technologies in the field of collecting biometric data are being developed by IT giants like Google and Facebook. Advertisers use real-time facial recognition technology to show specific ads to customers. Banks and retail stores use biometrics to track criminals and untrustworthy customers. Companies are replacing office locks with iris or fingerprint scanners, and elite clubs are using biometric information to identify important clients.
Last year, Russian banks everywhere began to launch pilot projects using biometric technologies to register users and provide them with online services. For now, in this area, biometric data will work alongside standard security systems, such as a login-password pair.
How reliable are biometrics?
While biometric technologies are far from perfect. Physiological indicators are more stable compared to behavioral indicators: they change less throughout life and are not susceptible to situational factors, such as stress. However, history knows many examples when such measurements are falsely accepted or rejected by recognition systems. For example, a face can be replaced with a high-resolution photograph or video, and fingerprints can be “stolen.” A famous case occurred in 2005 in the British prison Glenochil, where prisoners easily learned to cheat the lock system, based on fingerprinting.
Often the risk of error is associated with the identification conditions. Poor quality photography can significantly increase the risk. Lighting, the intensity of background noise, and the person’s position in space are important. Under ideal laboratory conditions, the error rate in facial recognition ranges from 5 to 10%.
Risks of data leakage
During verification, the data is checked against a biometric template that the person himself stores, for example, on a smart card. Only the user has control over his data. In the case of identification, a person’s data is checked against data from a single centralized database, which means that their carrier does not have any power over them. In such a situation, no one is protected from privacy violations and biometric information falling into the wrong hands.
Thus, it became known that Russian banks handed over client biometrics to the FSB - user data can be used in a completely different way to which the client agreed.
Indian incident
In early January 2018, journalists from The Tribune newspaper in the city of Chandigarh said they bought software that gave access to data from the Indian Aadhaar database from unknown sellers on WhatsApp for just £6. Aadhaar is a large centralized database that stores names, phone numbers, addresses of residents and their biometric data. Aadhaar identity cards are required for Indian citizens to access government services, receive benefits and allowances. Journalists reported that the software they purchased also allows them to print fake ID cards.
While the Unique Identification Authority of India (UIDAI) said journalists only accessed names and addresses that were meaningless without biometrics, the incident once again showed how unreliable such databases can be. Activists have already criticized Aadhaar for the starvation deaths of two Indian citizens who were unable to access the rations they were entitled to because receiving them required Aadhaar authentication.
In August 2017, the Supreme Court ruled that privacy is a right guaranteed by the Indian Constitution. Analysts predict that the decision will force a reconsideration of Aadhaar's crucial role in the lives of Indians.
Biometric data protection: where and how does it work?
Despite the very specific nature of biometric data, there are virtually no legal provisions around the world regarding their protection. Most legal texts talk about the protection of personal data and privacy in a broad sense, but sometimes such legislation is poorly adapted to biometrics.
In Russia, the collection and storage of biometric data is possible only with the consent of the subject of personal data in writing. This paragraph is in the law “On Personal Data”. On July 1, 2017, changes were made to it, and now all sites that collect and store any data about users must add documentation to their resource. The fine for failure to comply with these requirements will range from 10,000 to 75,000 rubles for each violation detected. And in the fall of 2017, the head of Roskomnadzor, Alexander Zharov, called for a ban on biometric identification of minors when they use technical devices.
Over the past 10 years, a number of bills created with an emphasis on biometric data have appeared in the United States, and in May 2018, a new EU law on the protection of personal data (General Data Protection Regulation, GDPR) will come into force in all European Union countries.
USA: three versus forty-seven
There is no single law in the United States that governs the collection and use of personal data, including biometrics. Strict legislation regarding biometrics exists in only three states: Illinois, Texas and Washington.
In 2008, Illinois passed the Biometric Information Privacy Act (BIPA), which established strict requirements for organizations that collect, purchase, or otherwise obtain users' biometric data. The law is aimed against the unrestricted use of biometrics for commercial purposes. Any business that gains access to such data should develop a publicly available data retention policy, limit the transfer or disclosure of biometrics, and protect that data in the same way a company protects other sensitive information. BIPA establishes a right of action for the “aggrieved person” and provides damages of $1,000 for each negligent violation and $5,000 for willful violations. In January 2017, similar bills were considered in Connecticut, New Hampshire, Washington and Alaska, but were adopted only in Washington.
In 2016, a group of Illinois plaintiffs sued Facebook for illegally collecting biometric data. The plaintiffs alleged that the social network's facial recognition feature, which tags photos, illegally collected and stored user data. In 2017, more than thirty lawsuits were filed in Illinois courts against companies that collected employee fingerprints to track work hours.
In general, in 47 US states, companies can use software to identify faces in images without user consent if the image is in the public domain. Facial recognition software already exists that stores can use to identify customers who return items too often or who prefer a certain type of purchase. Thanks to Facebook, employees can immediately get information about customers when they first enter the store, find out who they are, where they are from, and what their income is. From a privacy perspective, this is a violation of anonymity, the principle of user consent, and the appropriateness of using biometric data. But this is not prohibited by law in these states.
The European Union is trying to bring back privacy
This year, the European Union is taking a step towards the confidentiality of biometric information: in May 2018, a unified law on the protection of personal data (General Data Protection Regulation, GDPR), adopted in 2016, comes into force. The main goal of the GDPR is to return European citizens control over their personal data and at the same time simplifying the regulatory framework for companies. This law affects not only the 28 countries of the European Union, but also organizations that have representative offices in EU countries, collect and process personal data, provide services to individuals - citizens of the European Union, and use online registration on websites and applications. Therefore, the law will have a strong impact, in particular, on Russian business.
The law, written with a focus on biometrics, will consolidate and strengthen all previously existing standards for the protection of personal data in European countries. Specifically, the GDPR requires any organization to seek user consent before collecting data. However, the data subject has the right to withdraw his consent at any time. This principle is called the “right to be forgotten.”
Companies that manage biometric information will face huge fines if they fail to keep the data secure. Sanctions can reach 20 million euros or 4% of annual global turnover.
The law says that the use of data must be limited. Personal data should be collected and processed only for “specific, explicit and legitimate purposes” (data minimization principle).
China is building a digital dictatorship
While European countries and organizations prepare for the GDPR to come into force, China continues to develop a social credit system that seems to leave no trace of privacy in the country. By 2020, each resident of China, depending on their behavior, will be assigned a personal rating, which will affect access to government services, the ability to take out a loan, get a job, enroll children in school, shop and travel.
The social credit system is based on collecting as much data as possible about citizens and assessing the trustworthiness of residents based on their financial, social and online behavior. Thus, the rating takes into account credit history, timely payment of fines, compliance with traffic rules, purchasing habits, time spent playing computer games (the more idleness, the lower the rating), compliance with family planning rules, frequency of visits to parents, statements on the Internet, social circle ( spending time with people lower in rating will be unprofitable). For now, participation in the ranking is voluntary, but by 2020 it will be mandatory for all individuals and legal entities.
To collect citizen data, the government has hired eight private companies to develop algorithms to evaluate social credit. Among them is China Rapid Finance, a partner of the tech giant Tencent, which supports the largest messenger WeChat with more than 850 million active users. Another player is Sesame Credit, run by Ant Financial Services Group (AFSG), a subsidiary of Alibaba. AFSG sells insurance and provides loans to small and medium-sized businesses, and also owns the AliPay service, which is used not only for online purchases, but also for restaurants, taxis, school fees, movie tickets and money transfers. To develop the social credit system, Sesame teamed up with other data-gathering platforms Didi Chuxing, Uber's former main Chinese competitor, and Baihe, the country's largest online dating service. It's hard to even imagine how much these companies know about their users.
Through total control of online and offline behavior, the system is expected to push citizens to take actions that the government approves and help increase general “sincerity” and trust. The role of facial recognition systems and other biometric technologies in this project will be huge.
The possibilities of biometrics are increasingly turning into problems: data leakage, cybercrime, “identity theft.” And the growing use of biometric technologies poses new challenges for governments. Will states protect the anonymity of their citizens, or will complete transparency await not only the residents of China, but also everyone who has an account on social networks, uses a telephone, and at least sometimes leaves the house? The development of technology in any case will require the development of a legal framework.
Text: Anna Kozonina
Many Habr readers are probably already familiar with biometric technologies. They are now ubiquitous. In a general sense, biometrics is a system for recognizing people by one or more physical (or behavioral) characteristics. In the field of information technology, biometric data is used as a form of access identifier management and access control. Typically, the operating mode of biometric systems comes down to two main types.
The first is called verification, which is a comparison of the test result with a biometric template. This option helps verify whether the person is who they say they are. Verification can be carried out in various ways, including a smart card, user name or user number. The second mode is identification. Once a specific sample is received, the system is checked against a biometric database to determine identity. There is one important point here - for this mode of operation, the biometric sample must be in the database, and the comparison must be carried out on the “one to many” principle. In general, biometric technologies have enormous potential, which has not yet been fully realized. What is the state of biometric technologies in Russia and the world today?
In a number of cases, their development cannot yet be considered satisfactory. So far, this area is actively developing, although there are already some results (more on this below). In some cases, biometrics are considered not a very reliable method of identification or verification. Thus, in the United States, the Tampa Police Department even uninstalled facial recognition software, considering it not very reliable. But there they talked about the introduction of outdated biometric methods, which do not always show their best side.
However, modern biometric technologies are becoming more accurate and reliable. Many companies and scientific organizations are engaged in research and development in this area. Moreover, priority over time has shifted to contactless methods of biometric recognition. Biometrics are used in many areas, including banking, security and access control systems, visa control systems, police criminal identification systems, collecting visitor statistics and much more. So far, about half of the biometric market is occupied by fingerprint recognition systems. But the situation is gradually changing, developers understand that fingerprinting is not the most reliable way to identify a person (in “MythBusters” they once even showed a way to open a fingerprint lock using fingerprints printed on a printer), so new biometric technologies are gradually becoming more and more popular .
Biometrics: scale
In general, we can say that biometrics has become an integral part of people's lives. In some countries, for example, you cannot obtain a passport or visa without biometric data. Government organizations in various countries believe that biometrics is one of the most effective ways to identify refugees and those who cross the border illegally.Now there are many projects based on biometric technologies. Perhaps one of the most large-scale is the AADHAAR project, implemented in India. It is a biometric identification system that contains data from more than a billion people. The database contains about 10 billion fingerprint templates, two billion iris templates and a billion photographs. Something similar was shown in the science fiction film I Origins. However, identification by iris is a very real technology that is becoming increasingly popular.
All residents of India can obtain an entry in AADHAAR; this is an identification number that is linked to the biometric data of users. It is used in financial transactions, when working with various public and private services. A cloud service for storing scanned documents is also linked to AADHAAR.
Of course, India is not alone in introducing biometric identification. Other states are doing this too. And not only states, but also private companies. According to the analytical agency J"son & Partners Consulting, the global market for biometric systems will reach $40 billion by 2022. Analysts' conclusions are based on revenue indicators of key players depending on segments, taking into account hardware, software and integration.
Another analytical agency, Acuity Research, estimates the growth in the number of biometric electronic documents e-ID to 749 million by 2018. And in total, according to agency specialists, in 2018 there will be about 3.5 billion electronic documents in the world. Already, more than half of the UN member countries issue biometric passports. An example of the implementation of programs for the transition to biometric electronic documents includes government and private contracts in Canada, the USA, Belarus, Ukraine, Moldova, Lithuania, Hungary, Bangladesh, Senegal and other countries.
What about in Russia?
In Russia, biometric technologies are developing quite quickly, more actively than in many countries. For example, the largest banks in the Russian Federation began testing biometric customer identification systems this year. The Central Bank, the Ministry of Telecom and Mass Communications and Rosfinmonitoring are creating their own biometric database; this system will reach the testing stage this year.According to Deputy Chairman of the Central Bank Olga Skorobogatova, the pilot project will allow you to become a client of any bank remotely. To do this, it will be enough to go through the biometric registration procedure once at any credit institution participating in the project.
“Biometrics is a very exciting topic. This is identification, remote identification, the creation of a unified database about individuals, I’m talking more about individuals, which would enable any bank and any organization not to force clients to come on their feet to fill out a large list of documents,” RIA quotes Skorobogatova.
From this experiment to the creation of a national biometric database there is literally one step left.
The banking sector is trying to introduce customer identification systems by voice, photograph, and fingerprints. For example, VTB24 has already tested biometric identification as part of online banking. During the login process of the online banking app, customers were asked to provide their photo and voice sample. Using this data it is planned to carry out identification. After confirming the user's identity, all operations are performed without additional confirmation. Sberbank also shows great interest in biometrics, which has already accredited RecFaces (Comlogic application) as one of its partners in this area.
Similar technologies are used in Promsvyazbank and Home Credit, Tinkoff Bank and a number of other organizations. As for the unified biometric database, the Central Bank, the Ministry of Telecom and Mass Communications and Rosfinmonitoring are simultaneously working on its creation. This project may take several years to complete. A common biometric data base, according to experts, will be useful for the financial and legal sectors, government services, public safety, medicine and more.
Digital biometric profile from RecFaces
Speaking about biometrics in Russia, we cannot fail to mention our development - an information platform for multimodal identification, called Id-Me.Typically, companies that implement biometrics must select several suppliers and invest considerable funds in the creation of a central computing infrastructure, its maintenance, various types of licenses and equipment.
But it’s not just about the investment and the complexity of the process. Each algorithm that is currently offered on the market has a number of its own features and advantages. We at RecFaces have focused specifically on creating a full-fledged integrated platform that uses the world's best achievements in the field of biometrics. Having the opportunity to study algorithms and compare them, we select those solutions that show maximum efficiency.
For example, biometric identification technologies based on a mathematical model of the face are licensed from the Japanese company Toshiba. 3D identification is carried out using solutions from Artec ID and Intel Corporation. There is no doubt that for identification modules based on iris patterns, fingerprints, and palm vein patterns that are being prepared for implementation into the Id-Me platform, RecFaces will choose the most modern and promising technical solutions. Clients will only have to use the “magic” of Id-Me to solve their application problems.
For an outside observer, Id-Me works quite simply. One of the main components of the system is the Id-Box (capture module). It is a small “smart” identification device based on a PC platform in a compact case. It is this element that is responsible for recognizing faces and, in the future, other types of biometric data. It connects to a surveillance camera and other sensors. The system receives an array of data from them, which is then converted into a specialized index, a mathematical model, which is sent to the cloud for comparison with the standard stored there. By working with indexes, the system is not demanding on the “width” of the Internet channel.
This is a universal system that works effectively with various types of images and can use information from a surveillance camera. Id-Box, if necessary, can collect statistics on the number of visitors, including age, gender and emotional state. If a failure occurs, there is no need to worry; inside the box there is its own large hard drive where all important data is stored. In the event of a sudden network shutdown, all information will be saved and the system will continue to operate.
The data collected by Id-box is sent to the cloud, where the system compares the current index with all previous versions. If there is a match, that is, the system recognizes the registered person, the client receives an alert. The service is compatible with the main basic platforms, including the web interface, mobile clients iOS, Android, Windows.
The entire system is securely protected thanks to an encrypted connection. In addition, there is a Firewall and a crypto gateway with crypto routers is provided. An electronic digital signature, anti-virus software and intrusion detection tools certified by FSTEC are used.
Scope of application of Id-Me
The Id-Me biometric platform from RecFaces is designed to be as useful as possible for banks, airports, retail, hotel businesses, sports organizations, and government agencies.Banks can use biometrics to improve security. Here we can give as an example a possible case of a fraudster attempting to withdraw money from someone else’s card. The ATM camera connected to Id-Me identifies the face of the person trying to do this. If this information does not match what is contained in the database, the withdrawal of funds is blocked almost instantly. To use this method of protection, you do not even need to equip the ATM with additional equipment.
Similarly, a bank can protect its lending department. A fraudster who is trying to carry out a financial transaction under a false name will not be able to do this if he is being watched by a camera connected to the Id-Me service.
Plus, bank employees can also log in, which is necessary when performing any critical tasks. This function can be useful in many areas. Id-Me, for example, allows you to automate staff time tracking.
Since Id-Me can analyze video streams from surveillance cameras and individual images, the system can be used to collect statistics about visits, movement trajectories and visitor behavior.
Using proprietary recognition technologies from Toshiba and other partners, Id-Me allows you to use biometric identification to determine the buyer's gender, age, and personal preferences, linking all this with CRM. Such a system is also great for recognizing an important client by immediately receiving information about him, the date of his last visit to a store or other site. All this will help you find a common language with a person, instantly determining his preferences.
Examples of such combination of biometric identification with CRM already implemented jointly with RecFaces partners have shown their high marketing effectiveness. We will definitely write about this in detail later.
For the hotel industry, knowing your customers is essential. If a person sees that he is remembered, not only his first and last name, but also his preferences, then, most likely, such a client will return to the “attentive” hotel again and again. And for unwanted guests, you can create a “black list” with the relevant data.
Hotel surveillance cameras will record everything that happens, notifying the administration if an unknown person has entered the room or office space. Hotel employees will be aware that the smart system always knows who went where and why, so there will be less abuse.
Organizers of sports competitions, concerts and other public events can quickly receive information about unwanted elements (for example, hooligan fans) trying to get into the event. Lost child? The system will help you quickly determine how and when this happened, and will also determine where the child is if he is visible. Something went wrong? Security will be immediately alerted.
It will be easier for law enforcement officers to maintain security in schools or public places, or at transport infrastructure facilities, if they receive notifications about suspicious people and events occurring in the surveillance zone. Car thefts, hooliganism - all this can be prevented if you find out about the problem in time.
In general, there are a huge number of ways to use biometric systems. No matter how pretentious it may sound, they are the future. Biometrics is and will be used in a large number of areas. And Id-Me can already be used in most of them. You can learn more about what solutions the company already offers and is preparing to launch, and get acquainted with its comprehensive solutions by visiting the 23rd International Exhibition of Technical Security Equipment and Equipment for Security and Fire Protection
The path of technology that has gone beyond use in law enforcement agencies and replaced graphic and numeric passwords.
To bookmarks
Biometrics were the first to be used by law enforcement agencies and high security services. Nowadays, biometric systems are found in almost all modern devices: cars, laptops, smartphones.
Biometrics are measurable anatomical, physiological and behavioral characteristics that are used to identify an individual. The most common method is fingerprint recognition. But there are other ways - DNA, iris, voice, palms and facial features.
The regulatory, technical and legal framework for biometric technologies is now actively developing. The state initiates the formation of uniform standards to ensure the interaction of autonomous systems. Biometrics committees and departments are being created. Despite the variety of biometric methods, only three areas are mainly used: fingerprint, face and iris recognition.
The development of computer technology makes it possible to use biometrics in many areas of activity: controlling access to premises and devices, confirming financial transactions, ensuring security at airports, identification in schools and hospitals, searching for criminals.
The history of biometrics began three thousand years ago. Artifacts found in Nova Scotia, Babylon and China show that hand and fingerprints were used in ancient times for business transactions and evidence of crimes.
It was only centuries later that people resumed exploring the use of fingerprints and other indicators as a means of identification.
The first people to use biometrics in the modern world were police officers. Until about the mid-1800s, law enforcement officers had to use eye and memory to identify previously arrested criminals. A photograph of a person made the task easier, but could not serve as evidence of guilt.
By the 1920s, the FBI opened the first Department of Identification, creating a central repository of criminal identification data for U.S. law enforcement agencies. In the 1980s, the US government sponsored the creation of automated fingerprint identification systems that became central to police and other law enforcement agencies around the world.
Like a fingerprint, the iris of the eye remains unchanged with age. Its use in biometrics allows the use of contactless identification.
An equally necessary type of biometrics is facial recognition. Initially, this technology was used to ensure safety in crowded places.
In shopping malls, this helps prevent crime and violence. Airports are improving convenience and safety. Device manufacturers are using facial recognition technology to provide users with a new level of biometric security.
More difficult than scanning fingerprints, face or iris, only voiceprint identification. Unique components make voice substitution almost impossible. The history of voice biometric data begins not so long ago. The first real-time identification methods appeared in the late 1990s.
1665
Marcello Malfighi publishes his discovery of the uniqueness of fingerprints.
1858
Indian civil servant William Herschel records each employee's fingerprints on the back of their employment contracts. In this way, Herschel distinguishes employees from other people who may claim to be employees on payday.
1870
French lawyer Alphonse Bertillonage is developing the Bertillonage system - a method for identifying criminals using anthropometric data. The method is based on detailed reports of body measurements, physical descriptions and photographs. The system was used all over the world for 30 years until police realized that some people could have the same parameters.
1880
Scottish surgeon Henry Faulds publishes an article on the usefulness of fingerprints for identification.
1892
Argentine police officer Juan Vucenich begins collecting and cataloging fingerprints. And also uses prints to prove the final guilt of Francisca Rojas in the murder of her neighbor. The police officer determines that her print is identical to a partial blood trail at the crime scene.
In the same year, Francis Galton wrote a detailed study of fingerprints, in which he presented a new classification system.
1896
Inspector General of the Bengal Police Edward Henry, interested in Galton's system, collects a suitcase of fingerprint photographs and improves Galton's classification. Henry divides finger patterns into five basic ones: simple and complex arcs, loops towards the thumb or little finger, and swirls.
Henry's main idea is to encode patterns with numerical formulas. Species were designated by the letters A, T, R, U, W, and subspecies by numbers. Henry's method was the forerunner of the classification system that the FBI and other law enforcement agencies used for many years.
1903
Bertillon's system "breaks down". Two men, later revealed to be twins, were sentenced to forced labor in the United States. It has been established that they have almost identical Bertillonage measurements. But the story is later disputed because it was used to prove the imperfection of bertillonage.
1936
Ophthalmologist Frank Birch proposed using the iris of the eye for personality recognition.
1960
Swedish professor Gunnar Fant publishes a model describing the physiological components of acoustic speech production. The results are based on the analysis of X-rays of individuals making certain sounds.
1964
Woodrow Bledsoe, Helen Chan Wolf, and Charles Bisson develop the initial technology as part of their collective research into pattern recognition. However, Bledsoe is leaving the study, which is being continued by Peter Hart at the Stanford Research Institute.
1965
Woodrow Bledsoe is developing the first semi-automatic facial recognition system under a US government contract.
North American aviation develops first signature recognition system.
1968
A computer consistently outperforms humans in identifying human faces from a database of two thousand photographs.
1969
The FBI begins to develop a system to automate the fingerprint identification process, which becomes a priority and occupies the majority of human resources.
The FBI signs a contract with the National Institute of Standards and Technology (NIST) to study the process of automating human fingerprint identification. NIST identifies two main problems: the first is fingerprint scanning and identifying distinctive features, the second is comparing and contrasting features.
1970
Behavioral components of speech are modeled. Dr. Joseph Purkell expands on the original model developed in 1960. He includes his tongue and jaw. The model provides a more detailed understanding of the complex behavioral and biological components of speech.
1971
Researchers Goldstein, Harmon, and Lesk publish a paper, “Human Face Identification,” that uses 22 relative markers, such as hair color and lip thickness, to automatically recognize faces. The study formed the basis for further study of computer-based facial identification.
1974
The first commercial biometric palm recognition devices appear. The systems are implemented for three main purposes: physical access control, time recording and attendance tracking, and people identification.
1975
The FBI is funding the development of sensors to scan fingerprint patterns to reduce the cost of storing digital information. Early sensors use capacitive methods to collect fingerprint characteristics.
Over the next decades, NIST focuses on developing automated methods for fingerprint digitization and image compression, classification, feature extraction, and feature matching. NIST research resulted in M40, the first computer fingerprint matching algorithm used by the FBI.
1976
US electrical components manufacturer Texas Instruments is developing a speech recognition prototype that is being tested by the US Air Force and the non-profit company Miter Corporation. The latter is engaged in the design, research and development of systems, as well as support of information technology for the US government.
1977
Veripen has been awarded a "Personal Identification Apparatus" patent that captures the dynamic characteristics of a person's signature. Development of the system led to testing of automatic handwriting verification performed by Miter Corporation for the United States Air Force Electronic Systems Division.
1984
The US Army is beginning to use palm recognition in banking.
1985
Ophthalmologists Leonardo Flom and Aran Safir suggest that no two irises are alike.
1986
NIST and the American National Standards Institute (ANSI) are creating the fingerprint pattern data exchange standard ANSI/NBS-I CST 1-1986. This is the first version of existing standards that are now used by law enforcement agencies around the world.
Flom and Safir receive a patent for the use of the iris for identification. Flom approaches Dr. John Dogman with a request to develop an algorithm for identifying a person by his iris.
1987
NIST is forming a group to study and develop the use of speech processing techniques.
1988
The Los Angeles County Sheriff's Department's Lakewood Division uses the first semi-automated facial recognition system against a database of digitized copies.
That same year, Kirby and Sirovich apply principal component analysis—standard methods of linear algebra—to the problem of face recognition. The technology is called Eigenface.
1991
Matthew Turk and Alex Pentland find that the Eigenface residual error can be used to find edges in images. As a result of this discovery, reliable automatic face recognition in real time became possible.
1992
The NSA creates the Biometric Consortium and holds its first meeting in October 1992. Initially, participation in the Consortium is limited to government agencies. However, the organization soon expanded its membership to include the private and academic communities, and developed numerous working groups to initiate and expand efforts in testing, standards development, interoperability, and government collaboration.
Since the beginning of biometrics work in the early 2000s, working groups have been integrated into other organizations, such as INCITS, ISO and the US National Science and Technology Council, to expand and accelerate their activities. The consortium becomes a forum for discussions between government, industry and academia.
1993
The Defense Advanced Research Projects Agency and the Defense Development Program Office are funding the FacE REcognition Technology (FERET) program. The purpose of the incentive is to develop facial recognition algorithms and technologies.
1994
The Integrated Automated Fingerprint Identification System (IAFIS) competition explores three main problems: digital fingerprint acquisition, local sulcus feature extraction, and sulcus feature matching. Lockheed Martin won the competition to create IAFIS for the FBI.
The first Automated Fingerprint Identification System (AFIS) designed to support fingerprint printing is believed to have been built by the Hungarian company RECOWARE. In 1997, the palm and fingerprint identification technology built into RECOderm was purchased by Lockheed Martin Information Systems.
In the same year, the Immigration and Naturalistic Passenger Service Expedited Service (INSPASS) was created based on biometrics. It helped travelers bypass immigration lines at select airports throughout the United States until it went out of business at the end of 2004.
John Daungman develops and patents the first algorithms for computer identification of iris patterns. The patent is called lriScan. Until now, Daugman's algorithms are the basis for public applications of the technology.
1995
The Nuclear Defense Agency and iriScan are creating a joint project that has led to the first commercial product in the iris recognition field.
1996
The Atlanta Olympics are implementing palm access systems to control and secure physical access to the Olympic Village. The system finds information among the data of more than 65 thousand people. More than one million transactions were processed within 28 days.
With NSA funding, NIST is launching an annual NIST Speaker Recognition Assessment to further advance the speaker recognition community.
1997
IAFIS begins work. During the development of the system, scientists considered issues related to the exchange of information between autonomous systems, and also studied the implementation of a national system for identifying fingerprints. IAFIS is used to check people's criminal records and identify latent prints found at crime scenes.
Christoph von der Malsburg and a team of graduate students from the University of Bochum in Germany developed the ZN-Face system, which was then the most reliable due to its ability to recognize faces in low-quality photographs.
The technology was funded by the US Army Research Laboratory. However, large international airports, banks and government agencies used it.
With support from the NSA, the first commercial common biometric standard, the Human Authentication API (HA-API), was published. The goal of the project is to facilitate integration and ensure interchangeability and independence of suppliers. This was a breakthrough for biometric technology providers working together.
1998
The FBI launches a forensic DNA database, the Combined DNA Index System (CODIS). The system provides digital storage and retrieval of DNA markers for law enforcement agencies.
1999
The International Civil Aviation Organization (ICAO) Technical Advisory Group on Machine Readable Travel Documents (TAG or MRTD) has begun research into the compatibility of biometric and machine readable travel documents. The objectives of the study are to create international standards for multiservice data transmission.
year 2000
Several US government agencies are sponsoring Facial Recognition Vendor Testing (FRVT). The tests are conducted by NIST. This marked the first open, large-scale evaluation of several commercially available biometric systems.
Additional assessments took place in 2003 and 2006. The project's goal was to provide law enforcement and the US government with the information needed to determine the best ways to deploy facial recognition technology.
Scientists publish the first research paper describing the use of vessel patterns to recognize people. The article describes the first commercial technology that uses the image of blood vessels on the human hand for identification.
That same year, West Virginia University and the FBI introduced a bachelor's degree program in biometric systems.
January 2001
Facial recognition is being installed at the Super Bowl in Tampa, Florida, to identify wanted people in the stadium. The system didn't find them, but it did mistakenly identify a dozen innocent fans. The media is concerned about the violation of people's privacy when using biometrics.
September 11, 2001
A series of terrorist attacks carried out by the terrorist organization Al-Qaeda renewed scientific interest in the technology. This primarily affected transport systems and bodies ensuring the international movement of people, for example, customs and migration services.
Personal identification when checking documents was not enough, while biometric indicators guarantee accurate recognition of people.
November 2001
An M1 technical committee is being created to accelerate the development of standards for the use of biometrics in the United States and in international standards commissions.
2002
The International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) have established the ISO/IEC JTC1 subcommittee to support the standardization of biometric technologies. The subcommittee develops standards to enable integration and data exchange between autonomous applications and systems.
2003
The International Civil Aviation Organization (ICAO) is adopting a globally harmonized plan for the integration of biometric identification information into passports and other machine readable documents (MRDOs). Facial recognition is chosen as a global interoperable biometric model for computerized identity verification.
In the same year, the European Commission supported the creation of the European Biometric Forum. The project aims to make the EU a world leader in biometrics by removing barriers to decision-making and fragmentation in the market. The Forum also acts as a driving force for coordination, support and strengthening of national authorities.
2004
The US Department of Defense is implementing an Automated Biometric Identification System (ABIS). It is being implemented to improve the US government's ability to track and identify national security threats.
2005 year
US patent on iris recognition concept expires. This opens up marketing opportunities for companies that have developed their own iris recognition algorithms.
2010
The NSA uses biometric data to identify terrorists. This includes using fingerprints from locations associated with the September 11 attacks.
2011
The Government of Panama, working with US Homeland Security Secretary Janet Napolitano, has initiated a pilot program of the FaceFirst facial recognition platform to reduce illegal activity at Tocumen Airport in Panama.
It is known as a center for drug smuggling and organized crime. As a result, the system helped apprehend several Interpol suspects.
Facial identification is increasingly being used for forensic purposes by law enforcement and military personnel. This is often the most effective way to identify dead bodies.
Facial recognition and DNA technology were used to confirm the identity of Osama bin Laden - the founder of the terrorist organization al-Qaeda - after he was killed in a US raid.
year 2013
Apple is introducing Touch ID fingerprint recognition into new smartphones.
2016
Samsung is presenting a device with an iris scanner to increase the level of security for accessing the device.
MasterCard, Visa and other financial institutions include biometric payment authentication.
2017
Retail trade is actively introducing facial recognition technologies. And it is becoming the fastest growing sector in the use of this technology.
In addition, Apple is introducing iPhone X with Face ID facial recognition technology.
