Biometrics is the science of identifying people through their unique physiological or behavioral characteristics. Once a person’s characteristics have been analyzed, stored and verified, it can be used for authentication purposes, such as providing access to a secure facility or computer system.
Types of Biometrics
Physiological: Some methods for this type of biometrics include retinal, fingerprint, and facial recognition.
Retinal – A retinal scanner is a device that projects infrared light to the back of a users’ retina while he is looking through its eyepiece. Some of the light is absorbed by the blood vessels, and some is reflected back. The various intensities of the returning light are measured and assigned a numeric intensity grade. These values are stored in a database for future comparisons. The next time the scanner is used, it will scan and compare the incoming information to the information stored in its database, confirming the person’s identity if it is a match. This type of system is very accurate, with an error rate of less than one in a million.
Fingerprint – A fingerprint scanner is an electronic device that can capture and digitally store an image of a user’s fingerprint. The image is archived and used as a template to identify authorized users. The device uses the whorl, loop, and arch patterns of the fingerprint for identification. Some technologies employed in the use of these scanners are ultrasonic, capacitance, and optical. An ultrasonic scanner uses sound waves to penetrate the outer layer of skin to read the fingerprint, and is unaffected by dirt or contaminants on the sensor surface. An active capacitance scanner uses voltage and the mathematical principles of capacitance to measure the dermal surface, with the skin acting as the plate of a capacitor. This method also eliminates the need for a clean sensor or skin surface. The optical method works similar to a digital camera, and is the most error prone, since a dirty or scratched sensor or poor skin quality can produce errors.
Facial – Facial-recognition software identifies people based on their facial features. These systems use cameras to capture facial properties such as nose length, eye socket depth, mouth size, cheekbone depth, and so forth. Once the visual information is captured, it is stored and compared to a video or still image of the subject from a database, utilizing state-of-the-art algorithms based on skin textures and 3D scans. This biometric system is still evolving and is far from perfect. Low resolution images, poor lighting, facial hair, and unusual facial expressions can render it unreliable.
Behavioral: Some methods for this type of biometrics include keystroke dynamics and voice recognition.
Keystroke – Every individual types differently on a keyboard due to various factors such as typing skill or native language. The speed of typing, the rate at which the user presses and releases keys, and common spelling errors can also “tag” the distinctive patterns of every individual. A biometric template is then created from initial inputs using keystroke algorithms. These systems incorporate key loggers (hardware or software that records all typing) to monitor all keystroke activity. Unfortunately, many factors affect the accuracy of such a system. Mood, health, alcohol impairment, medication, and switching computers can all affect whether such a system can maintain precise performance.
Voice – Voice verification is a system that compares a stored digital template of a users’ voice (usually a passphrase), and compares it with a live spoken sample. If the two samples match, the user is allowed entry to the system. The different characteristics of a person’s voice, such as cadence, tone, and pitch, allow the system to distinguish between different users. This type of biometric system is also subject to erroneous readings caused by background noise or a person’s poor health (a cold often changes the pitch of their voice). Some systems can be “hacked” using a recording of the subject’s voice, unless it requests random phrases at the time of input.
Biometrics is a continually developing technology that can be used to protect your assets, both physical and electronic. A combination of biometric methodologies can almost guarantee that there will be no unauthorized access to your important data or possessions. For data security specifically, a more economical and secure method is using easily available encryption software.