Terminology - biometry, fingerprint recognition

Biometry

Biometry or biometrics is a general term of characteristics or process description. It is a biological or behaviouristic characteristic which can be measured and used for automatic identification of a person. Research and testing of biometric systems are going on for a couple of decades already, but have only recently become public with the emergence of more powerful applications. Biometric systems are gradually becoming part of our lives and our daily activities. A lot of companies use biometric devices for access control, time and attendance control or logical access control, because they want to improve the security.

A typical biometric system is made of five components: a sensor is used to collect data and to transform them to a digital form. Algorithm monitors the quality of digital data and makes a biometric template. A data storage unit stores information (biometric templates) which will be used for comparison with ’live’ biometric templates. Algorithm compares biometric templates, which are stored in a storage unit, to the live templates. Then there is also a decision making process which uses the result of algorithm matching to switch algorithm on/switch off, authorise/forbid the performance of a function.

Biometry types

The most commonly used biometric techniques are: fingerprint, face, iris, voice, signature and hand geometry. They are many more but they are in different development stages and have a negligible percentage of the biometric market share. It could be said that there is no ideal biometric technique that could apply to all needs for authentication. When the decision is made that the biometric solution will be used, various factors should be considered (location, security risk, security level, number of users, etc.) and based on the results, the best technique or combination of techniques should be used. And it is essential to know that not all biometric techniques are equally mature.

What are the advantages of a biometric solution?

Increasing use and advanced development of various technological systems had resulted in increased use of passwords and PIN codes. Every day, passwords are used on ATM machines, mobile phones, computers, access control devices, etc. Passwords that are hard to guess must be as long and complicated as possible. For security reasons, passwords should never be written down, told to a third person and should be changed at least every three months. Because of all these authentication methods, an average user has to remember 10-30 passwords, most of which are rarely used. If the password is lost or forgotten it causes unwanted cost and inconvenience. Obviously, people are forced to sacrifice security due to the capacity of their memory.

Another way of authentication is the usage of RFID card or smart card solutions. This method is convenient only for certain types of applications, such as access control and time & attendance solutions. With this method, the user doesn't need to remember a password but there is a high probability of losing the card. Misuse is highly probable. There is no certainty that the card will indeed be used by it's owner.

There is a solution that is based on natural methods. People have different body characteristics, such as the face shape or the tone of voice. Biometry has become, with the help of ever more advanced technology, a tool for differentiating physical characteristics. Only recently has the technology (computers, microprocessors, biometric sensors, etc.) matured to the point that biometric solutions can be widely used. Now it is the right time that PINs, passwords and card systems are replaced by biometric solutions.

SSO (Single-Sign-On)

is a wizard that replaces user names, passwords, etc. with the fingerprint authentication. It enables the password replacement in Windows logon, screensaver, html access, IM programs, e-bank and many more. In its advanced version it manages the login to domain and has many other advanced functions that will significantly alleviate the work of an IT manager. .

Bioengine

is the expression for a firmware of Biomodule cryptoprocessor. Its functions are the extraction of the image from the sensor, its reconstruction and processing. It contains the algorithm that converts the image of a fingerprint to a biometric template, which can not be converted back to the image. It also encrypts the data and matches the templates.

Match-on-card

is fingerprint recognition technology where the matching is done fully in the smartcard. Not only the templates but also the matching algorithm is stored in the smartcard. By using this method a template is never stored in the database of a device. It is also fully acceptable from the point of data protection since every user stores the templates only on the smartcard.

FingerChip™

is a thermal sensor brand name of Atmel. This is the sensor that we use in our solutions. It measures the difference in temperature between the valleys and ridges of the fingerprint. This sensor is at the moment the best solution on the market as it is resistant to moist, temperature differences, dust, dirt, electrical disturbances, etc. The producer guarantees at least 1.000.000 swipes in the sensor's lifetime.

FRR (False Rejection Rate)

is the level of possibility that verifying of the fingerprint which is already stored in the database will not succeed.

FAR (False Acceptance Rate)

is the level of possibility that the fingerprint will be verified as some other fingerprint stored in the database.

Enrollment

is the procedure of inputting a fingerprint in a database or a smartcard.

Template

is a term for a group of characteristics that were taken from a fingerprint and converted to a special type of file. Templates are additionally encrypted and can not be converted back to the image of a fingerprint. For fingerprint matching the templates are used.