To date, biometric protection systems are increasingly used by developing new mathematical authentication algorithms. The range of tasks that is solved with the help of new technologies is quite extensive:
- Protection of law and order and criminalistics;
- Throughput system (SKUD) and restriction of access to public and commercial buildings, private housing (smart home);
- Transfer and receipt confidential information personal and commercial nature;
- Implementation of trade, financial and banking electronic operations;
- Entry to electronic remote and / or local workplace;
- Blocking work modern gadgets and protection of electronic data (cryptation keys);
- Maintaining and accessing government resources;
Conditionally, biometric authentication algorithms can be divided into two main types:
- Static - Dactyloscopy, Rainbow Eye Shell; measurement of the shape of the brush, lines of palms, placement of blood vessels, measurement of the face form in 2D and 3D algorithms;
- Dynamic - handwriting and rhythm of the text set; gait, voice, etc.
Main selection criteria
When choosing a capable installation of measuring the biological parameter of any type, you should pay attention to two parameters:
- FAR - determines the mathematical probability of coincidence of key biological parameters of two different people;
- FRR - determines the degree of probability of failure to access the face that has the right.
If manufacturers have missed these characteristics when presenting their product, it means their system is incapable and lags behind competitors on functionality and fault tolerance.
Also important parameters for comfortable operation are:
- Easy use and the ability to perform identification without stopping in front of the device;
- The read speed of the parameter, processing the information received and the volume of the database of biological reference indicators.
It should be remembered that biological indicators, static at least, and the dynamic are greater, are parameters that are subject to constant changes. The worst indicators for the static system are FAR ~ 0.1%, FRR ~ 6%. If the biometric system has indicators of failures below these values, it is ineffective and incapacitated.
Classification
To date, the market of biometric authentication systems is extremely uneven. In addition, with a rare exception, the manufacturers of security systems produce and the software with closed source codewhich is suitable exclusively to their biometric readers.
Fingerprints
Dactyloscopic analysis is the most common, technically and programmatically perfect method of biometric authentication. The main development condition is a well-developed scientific and theoretical and practical knowledge base. Methodology and system of classification of papillary lines. When scanning key points are the end of the line of the pattern, branching and single points. In particularly reliable scanners, a system of protection against latex gloves with prints is to check the relief of papillary lines and / or finger temperature.
In accordance with the number, character and placement of key points, a unique digital code is generated, which is stored in the database memory. The digitization time and typography typically does not exceed 1-1.5 sec., Depending on the size of the database. This method is one of the most reliable. Promotable authentication algorithms - Veri Finger SKD reliability indicators are FAR - 0.00% ... 0.10%, FRR- 0.30% ... 0.90%. This is enough for reliable and uninterrupted operation of the system in an organization with personnel more than 300 people.
Advantages and disadvantages
The indisputable advantages of this method are considered:
- High accuracy;
- Lower cost of devices and their wide choice;
- Simple and fast scanning procedure.
Of the main disadvantages it should be noted:
- Papillary lines on the fingers are easily damaged, causing errors in the operation of the system and blocking the passage of the employees who have the right to this;
- Dactyloscopic scanners must have a system of protection against the detailed image: Temperature sensors, pressure detectors, etc.
Manufacturers
Foreign companies that are engaged in the production of biometric systems, and devices for the SUD and need to note:
- SECUGEN - Mobile compact USB scanners for access to PCs;
- Bayometric Inc - Production of biometric scanners different types for integrated security systems;
- DigitalPersona, Inc - release of combined lock scanners with integrated door handles.
Domestic companies producing biometric scanners and to them:
- Biolink.
- Sonda
- Smartlock
Scan Eyes.
Rainbow Eye Shell is as unique as papillary lines on hand. Finally, formed in two years, it actually does not change throughout life. The exception is injured and sharp pathology of eye diseases. This is one of the most accurate user authentication methods. Devices produce scanning and primary data processing 300-500 ms, comparison of digitized information on an average power PC is made at a speed of 500,000-150000 comparisons per second. The method does not impose restrictions on maximum number users. Statistics FAR - 0.00% ... 0.10% and FRR- 0.08% ... 0.19% Casia EYR SDK based on the EYR SDK algorithm. According to these calculations, it is recommended to use such admission systems in organizations with more than 3,000 people. AT modern devices x Cameras with 1.3 mp matrix are widely used, which allows you to capture both eyes during scanning, it significantly increases the threshold of false or unauthorized triggers.
Advantages and disadvantages
- Benefits:
- High statistical reliability;
- The image capture can occur at a distance of up to several tens of centimeters, and the physical contact of the person with the outer shell of the scanning mechanism is eliminated;
- Reliable methods eliminating fake - checking the accommodation of pupil, almost completely exclude unauthorized access.
- Disadvantages:
- The price of such systems is significantly higher than dactyloscopic;
- Ready solutions are available only in the execution of large companies.
The main players in the market are: LG, Panasonic, Electronics, OKI, which are working on licenses to Iridian Technologies. The most common product with which you can encounter on russian market are ready-made solutions: BM-ET500, IRIS Access 2200, Oki Irispass. AT lately There are new companies that trust in AOPTIX, SRI International.
Scanning retina eye
An even less common, but a more reliable method - scanning the placement of the capillary network on the retina of the eye. Such a drawing has a stable structure and is unchanged throughout life. However, the very high cost and complexity of the scanning system, as well as the need for a long time not to move, make such a biometric system available only for government agencies with an increased protection system.
Face recognition
There are two main scanning algorithm:
2D is the most inefficient method that gives multiple statistical errors. It is to measure the distance between the main bodies of the face. Does not require the use of expensive equipment, only the camera and the corresponding software is sufficient. Recently, I received significant distribution in social networks.
3D - this method is radically different from the previous one. It is more accurate, to identify the object, even there is no need to stop in front of the camera. Comparison with information listed in the database is made through serial shooting, which is made on the go. To prepare the data on the client, the object turns the head in front of the camera and the program forms a 3D image with which the original merges.
The main manufacturers of software and specialized equipment on the market are: Geometrix, Inc., Genex Technologies, Cognitec Systems GmbH, BioScrypt. From Russian manufacturers, Artec Group, Vocord, ITV can be noted.
Scanning hands
Also divided into two drastically different methods:
- Scanning a brush patterns under infrared radiation;
- The geometry of the hands - the method occurred from forensic and recently goes into the past. It consists in measuring the distance between the joints of the fingers.
The choice of a suitable biometric system and its integration into a skid depends on the specific requirements of the organization's security system. For the most part, the level of protection against fake biometric systems is quite high, so for organizations with an average level of admission (secrecy), there are enough budget dactylocopic authentication systems.
The attractiveness of additive technologies is difficult to overestimate. Therefore, the auxiliary equipment for three-dimensional printing today takes such popularity. In a limited budget, you can make a 3D scanner with your own hands. To do this, use sweater agents and aggregates or simply turn an ordinary smartphone into the scanner.
Making 3D scanner using webcam
In order to make a homemade 3D scanner, you will need:
- high-quality webcam;
- linear laser, that is, a device emitting a laser beam (for high-quality scanning it is better for the beam to be as thinner as possible);
- different fasteners, including angle for calibration;
- special software For processing scanned pictures and data.
Note that without appropriate for you, it will not be possible to create a digital model of objects and objects. Therefore, initially take care of the availability of special programs. For example, the basic are considered to be DAVID-LASERSCANNER and TRIGLES, but they need to use the rotating surface.
Start from the calibration angle. To create it, type the template (it is included in the program kit). Place it in such a way that it creates an angle of 90 degrees. It is important that the correct scale is observed during the press. To do this, use the calibration scale. Camera calibration is made in automatic or manual mode, it also provides for software.
To scan it, it will need to be placed in a calibration corner, and on the contrary to install a webcam. It is important to put an object exactly in the center of the image on the screen. In the webcam settings you need to disable all automatic adjustments. Also with their help the color of the laser beam is installed. Pushing "Start", smooth movements are performed. The beam needs to circle the subject from all sides. This will be the first scan cycle. In the future, it is necessary to change the position of the laser to cover all the points unprocessed in the previous time.
Upon completion of all processes, scanning stops and the "display in 3D" mode is selected in the program. If you do not have a laser at hand, it can be replaced by a bright light source. It will provide the shadow line projection. True, in this case, change the settings in the program that will correspond to these parameters.
We make a three-dimensional scanner of two webcams
If you need high digitization accuracy, you will need to use two webcasts. In this case, the light source is replaced by the second chamber. 3D scanner with your own hands from two cameras allows you to minimize the time of calculations on the points falling into the laser band.
Making 3D scanner from the projector and webcam
To do this, you will need:
- projector;
- webcam;
- dAVID-LASERSCANNER program;
- tripods for webcam and projector;
- calibration panel (secure two small chipboard sheet at an angle of 90 degrees and cover with dry glue paper sheets with pre-printed templates in advance);
- rotary table (can be built from the old gym "Grace" and several pins).
To scan the object, we have it vertically and make 7-8 scans, rotating it in a circle. We combine the resulting scans. After that, change the position of the object and do the same procedure. We combine the scans of two soulins of the subject. By clicking on the "Fuzing" button, we get a three-dimensional model of the object. It can be saved in any selected format, and then process data using:
- DEELSM LANSAKER;
- Easylast;
- Last Dateign & Engineering;
- Forma 2000;
- Shoeemaster QS.
We make a three-dimensional scanner from the game console
Xbox One is a prefix that has already been equipped with a second-generation cycle and can be applied as a three-dimensional scanner. If you have a regular gaming controller, you can make a 3D scanner from Kinect using the following programs:
- Kinact FusiOn. Creates supercase models, reading data from the Kinect sensors.
- SkaNect. With it, 3D images of premises with all objects that are located in them are created. To create a three-dimensional model of the surrounding space, you just need to rotate the device around itself. In order to detail individual objects, it is necessary to restart the chamber on them.
Making 3D scanner from smartphone
How to make a 3D scanner from ordinary mobile device? Today for this uses various software products. With their help, the smartphone turns into a full-fledged three-dimensional scanner. The most popular program algorithms:
- Mobilefusion. It monitors the position of the subject using a regular camera, after which it takes a photograph. A three-dimensional model is obtained from the snapshot. Works for different platforms and OS.
- Helps in creating three-dimensional photos of any objects, after which it sends them to the 3D printer.
- AUTODESK 123D SATCH. With this program, three-dimensional models of buildings, people and other items can be created and print on additive devices that can be photographed from all corners and sides.
Such systems do not need hardware modifications or connecting to the Internet. To get started, you just need to run mobile app and spend the phone around the object that is scanned.
The will of the Fate entered one of the Russian universities. A diploma was decided to investigate one of the directions of biometric identification - identification in the figure of the veins of the palm. At the initial stages, it was conceived to protect the working model of the device (but everything was not so simple).
To begin with, it was necessary to understand how it represents this method And at the expense of what funds is carried out by its implementation. As already written in the article:
"The veins pattern is formed due to the fact that hemoglobin blood absorbs IR radiation. As a result, the degree of reflection decreases, and veins are visible on the chamber in the form of black lines. Special program Based on the received data creates a digital convolution. No contact person with scanning device is required. "
Figure below shows a graph of the absorption of IR radiation with saturated blood oxygen and blood without oxygen.
Choosing a method
There are two methods for obtaining images of the walls of the palm. Reflection method (reflection) allows you to place all the components of the device in one case, due to which the size decreases. The psychological barrier is also reduced (you do not need to drink your hand anywhere). The IR light transmission method is to install an IR illumination from the back of the palm, and the camera itself is installed on the palm side and receives IR radiation passing through the entire palm. Using the transmitting method, the resulting images are more detailed.
Iron
As the basis for obtaining images of the drawing of the city of Palm, the reflection method was selected. To begin with a paper sheet, a preliminary device model was drawn. It should measure the distance from the device itself to the palm of the identifiable, measuring the surface temperature of the hand (for statistics, I wanted to have such a functionality), turn on and off the IR illumination. As a camera stopped on the Logitech B910 webcam. Tests of device models, and there were already 4, they revealed the shortcomings of the three previous cameras. The first model was with the camera LinkSprite Jpeg Color Camera TTL Interface. The process of transmitting the image on the PC was limited to the speed of the port, it was necessary to refuse it. Another 2 noname webcams were also checked, but the resulting images were very poor quality. Next I will describe only the latest model.A list of components was drawn up:
- Arduino Project Enclosure - a small housing for Arduino projects
- Infrared Thermometer - MLX90614 - IR thermometer
- USB HUB - to connect the camera and Arduino one cable
- Orduino Nano - ATMEGA168
- Infrared Proximity Sensor - Sharp GP2Y0A21YK - IR Distance Sensor (from 10 to 80cm)
- 2 transistors
- 2 resistors 4.7kom, 6-470
- 6 IR diodes 850nm
- Phototransistor (for measuring the illumination of the device with a sneaker, did not participate in experiments, but was planned)
- USB B Connector
- IR Filter 850nm.
- USB cable
In the assembly, the device has the form:
BY
Arduino.
Managing all the components of the device (except for the camera) must implement Arduino. This is a distance sensor, a temperature sensor and an IR illumination. In the aggregate, the work algorithm is as follows:- the PC comes to the start of the start of the distance sensor and the temperature sensor, followed by the output of their values \u200b\u200bin the port
- as soon as the sensor value coincides with the listed values \u200b\u200bin the PC, the distance sensor is turned off (its ray makes the illumination on the resulting images)
- includes IR illumination
- a picture is taken
- iR illumination turns off
- transition to the initial stage
PC.
The images obtained from the device must be further processed and recognized. To write software for PC, MATLAB software was selected. The standard Matlab Neural Network Toolbox was used as the recognition tool. But before applying for learning and recognition, they need to be processed, namely to allocate basic signs.Received images from the device:
After adjusting the algorithm for itself:
Not bad, yes? This is the same image, but already something. Now it remains to fill the database with reference images and train the network. After learning it turned out that the invariance of images is very small. It turned out that in my hand the veins are clearly distinguished, and in some hands the quality of images to put it mildly "not very". On each hand, 100 reference images were collected. As a negative sample, roaming images and images without hands (elbows, fists, etc.) were used. The hands of two people were included in the sample.
After learning, they started checking the device. Errors of the second kind have a place to be. I had to reduce the threshold of coincidence with the references. In the investigation, sometimes false skips were. In practice, the mistakes of the second kind happened because of the wrong arrangement of the hand in front of the device, or by lubricating the removable images due to the movement of the hands. There was not enough time to experiment.
findings
Here you can only say that the topic is interesting and she did not fully revealed due to the lack of time and the necessary skills. Also did not have time to check the device on the "non-residential comparison". The phototransistor was not activated (for adjusting the camera under lighting) and the temperature sensor (although the information was obtained from it, but did not take into account).The source code for Arduino and Matlab is, but it is ashamed to show, for wrote \\ rewriting very rushing time, along the way, alterating the device.
Used sources
- Nadort, A. The Hand Vein Pattern Used As a Biometric Feature: Literature Thesis for Master Of Science Programmed Physics of Life / A. Nadort, - Amsterdam: Medical Natural Sciences At The Free University, 2007. - 179 p.
- FUKSIS, R. Palm Vein Biometrics Based on Palm Infrared Imaging and Complex Matched Filtering: The 12th ACM Workshop on Multimedia and Security / R. FUKSIS, M. Pudzs, M. Greitans, - Rome, 2009. - 27 s.
Frequently asked Questions
For the right and high-quality analysis on the chiromantia, first of all, photos of your palms are needed.
You need to make 5 photos:
- - right hand (do not cut your fingers and bracelets in the photo)
- - left hand
- - the rib from the maizin of the right hand
- - the rib from the mother's mother's
- - Right palm - without fingers
For the perfect photo, of course, need digital camera. The greater permission will be the camera, the better. If you have powerful camera on the cell phone, put on the macro shot.
The smallest lines on the palm should be visible in the photo, a papillar pattern should be viewed. After all, only in this case it is possible to see a complete picture of lines and signs and, accordingly, make a more complete analysis and give the most detailed information In your photo.
Photo I advise you to do during the day, when sunlight. Alternatively, near the window.
We must disable the flash and translate the camera to close shooting mode (macro), which is usually highlighted by the flower icon. In the process of the photo you need to fix the hand that you take off and the camera itself. Any movements in the shooting process is better excluded.
Photo is better to do on a light background. The windowsill will fit)))
The thumb is assigned to the side, but without effort, and the 4 remaining fingers are carried free, but do not press each other and do not spread !!!
We do the photo so that the whole palm and fingers are visible.
Also make photos and edges palm.
When photographing the edge of the palm - do not strain your hand, it should be in a free state.
It is too a good optionBut the scan also needs to be done correctly.
To begin with, we put the scan resolution of at least 300 PC / inch (better more, but do not overdo it !!!)
Put your hand on the glass as well as I told with the photo. That is, the big finger is assigned to the side, and the rest are kept free. Hand in glass can not be pressed !!! To keep it, of course, it is not necessary, but also not to press, otherwise the image will be unreadable.
If you have a strong palm center or well-developed bugs interfere with the scan in such a way that all lines are visible, I advise the photo!
Do not squeeze the photo and send a real size!
