Written by: Ethel Macasias
Biometrics is primarly used for identification and verification purposes. Biometric recognition is the study of methods for uniquely recognizing a person based on one or more physical or behavioural characteristic. Biometric recognition is being used in conjunction with various forms of technologies for authentication purposes. Some of these technologies include smart cards, magnetic stripe cards and physical keys.
There are two types of biometrics, physical and behavioural. Behavioural biometrics is usually used for verification and physical biometrics is usually used for determining or verifying ones identity. Various methods of biometrics stem from both the physical and behavioural biometric types. Some of these biometric recognition methods are mentioned below.
Iris patterns are randomly formed, thus a person's left and right eye are different. Scanning of the iris is used to measure the iris pattern in the
coloured part of the eye. The iris is very distinctive and robust, which provides quick results for both identification and verification purposes.
The measuring of blood vessel patterns from the back of an eye is considered to be retinal scanning. However, retinal scanning is not yet readily
available because some find this biometric system to be disturbing, since it involves the person to stand very still while a light source is shined in
The recording of spatial geometry of facial features are used for facial recognition. Since a person's face can be captured from a distance by a camera
many casinos, stores and urban areas use facial recognition to identify potential crimes.
Vocal characteristics are used to identify pass-phrases used by a person. Telephones and microphones are easily available and cheap, which are used as
sound sensors involved with voice recognition. However, voice recognition can be easily affected by background noise. As technology continues to develop
voice recognition is becoming more reliable.
Fingerprint biometrics is widely used by law enforcement agencies. The details of a digital fingerprint is removed for fingerprint pattern analysis.
There are currently three main applications for fingerprint recognition. The first one is used by law enforcement agencies and is known as the
Automated Finger Imaging Systems (AFIS). The other two main
applications are for fraud prevention and access into computers and other physical items.
Hand geometry involves the geometry of the dimensions of a hand. Therefore, spatial geometry is used after the person places their hand on a
scanner. This method of biometrics has a low degree of distinctiveness, therefore it is not widely used for identification purposes.
Dynamic signatures are verified by measuring an individual's signature. This method of biometrics studies the speed, direction, pressure and the total
time it took for the person to create the signature. As well this method can determine where the stylus was raised and lowered from the writing surface.
This method of biometrics examines a person's keystrokes on a keyboard. This method can determine the speed, pressure, total time to type particular
words and time between hits on specific keys. This method is still in the development stage to improve robustness and distinctiveness.
How Biometrics Works
Your fingerprints, retina scans and voice recognition are all biological properties. These unique human characteristics are being used in conjunction with other forms of technology for identification and verification purposes. Most biometric systems measure and record biometric characteristics of a person and matches the results to a database which contains information about individual people.
Biometric systems compare unique human characteristics against a database which contains information of many individuals. The matching of information occurs between a person's human characteristic and data from a database. Biometric systems involve both one-to-one and one-to-many matches. Verification, also known as authentication, uses one-to-one matches, whereas identification uses one-to-many matching.
There are two types of human characteristics used for biometric recognition, physical and behavioural. Fingerprints, hand geometry, facial and iris recognitions are various forms of physiological biometrics which are measured to derive specific data to identify an individual. Voice verification, keystroke dynamics, signatures and an individual's actions are forms of behavioural characteristics, which are also measured to obtain information to directly or indirectly verify a person's identity.
Advantages and Disadvantages
Biometric systems use the unique human characteristics of an individual. For this reason there is an extremely low probability for two humans to share
the same biometric data. Furthermore, biometric characteristics of a person are near impossible to duplicate and can only be altered or lost due to a
serious accident. Biometrics has proven to be a secure method to ensure data integrity, therefore, many banks, businesses and governments have
incorporated biometrics as part of their security procedures.
There are two types of recognition errors, False Accept Rate (FAR) and
False Reject Rate (FRR). FAR is considered to be the most serious recognition
error, which occurs when biometric data does not match, but are accepted as a match by the system. FRR is when biometric data that matches is rejected
by the system. This recognition error can be caused by incorrect alignment or dirt on scanner.
Governments, law enforcement, banks and various businesses are incorporating biometric identifiers to increase security measures.
Last Revised: April 6, 2007