There are a wide variety of techniques used to authenticate one or more individual physical characteristics. The unimodal or, single body trait, technique is most commonly used. This technique utilize a single trait, such as fingerprints, the retina or iris of the eye, hand geometry or facial recognition. Multimodal biometrics combines multiple biometric security techniques, such as voice and speech recognition, and provides an overall stronger authentication process.6
Fingerprint recognition: The images of the ridges and valleys (minutiae) found on the surface of a person’s finger tips can be captured using a scanning process. A mathematical representation (unique data set) of the person’s finger print is created from the image. The hash of this unique data set is saved in the database, not the scanned image of the fingerprint.
This is a one-way function, meaning that mathematically it is impossible to recreate the finger print image from this hashed data set. If the data is stolen or leaked, it cannot be reconverted to an image, and this eliminates the potential of someone recreating the finger print image from the underlying data.7
However, finger print recognition is not completely foolproof. For example, capacitive scanners, which uses tiny capacitors just below the surface of the scanner to track the finger’s friction ridges and valleys, can’t always distinguish between a mold of a person’s finger and the actual finger. An optical scanner, which uses an image sensor to capture the fingertip surface, could be spoofed by someone using a photo of another user’s finger print.
Retina recognition: This technique involves analyzing the layers and unique pattern of blood vessels situated in the back of the eye. This unique pattern can be scanned with the use of a low intensity light source through an optical coupler. Retinal scanning requires a person to look into a receptacle and focus on a given point. Drawbacks to this method include that it may not be very convenient for persons who wear glasses or are concerned about having contact with a reading device that is being used by a multitude of users.8
Iris recognition: The features found in the iris of the eye, or the colored rings of tissue that surrounds the pupil, are unique to every person. Through the use of mathematical recognition techniques, the pattern can be captured and used for authentication. This technique does not require contact with a scanning/reading device, can operate at distance from the user, and requires only a glance. These features make it less intrusive and more convenient for users than the retinal recognition process.9 However, the scanners require a person to stand in front of the device which is sensitive to movements and reflections such as eyelashes or lenses. An untimely blink of an eye will obstruct the scanning device delaying the recognition process and requiring repeated attempts.
Hand geometry: This technique uses the concept of measuring and recording length, width, thickness and surface area of an individual’s hand while placed on a plate. Hand geometry systems use a camera to capture a silhouette image of the hand. The subject’s hand is placed palm down on a plate and guided by five pegs that sense when the hand is in place. The image is evaluated and measured to create a template of the person’s hand characteristics. Hand geometry recognition has been in use the longest, first marketed in the late 1980’s and used at the 1996 Olympic Games to control and protect physical access to the Olympic Village.10
Facial recognition: This technology measures and matches unique characteristics of the face, and is incredibly versatile with a wide range of potential applications. It has the potential to be integrated anywhere you can find a modern camera.
Law enforcement agencies around the world use biometric software to scan faces in CCTV (Closed-circuit television) as well as to identify persons of interest in the field. Border Control uses this technique to verify the identity of travelers.
Face recognition has the ability to gather demographic information and is an emerging technology for the retail industry to help in product marketing. There are those who oppose this technique because of the potential to “profile” and create discriminatory practices.11
Voice and speech recognition: A voice print measures the sound a person makes while speaking, specifically the minutia of the voice. It is not wholly dependent on a spoken code. Speech recognition-- also called voice command-- uses a voice user interface (VUI) technology which allows users to interact and control technologies through speech. In other words, through the use of spoken words, computer software can recognize a voice and word sequence in tandem which can be integrated into security systems to allow or deny access to a restricted location or system. The combination of voice and speech recognition can be a powerful duo simultaneously providing both authentication and hands free interface.12
These are just some of the biometric security techniques that can be used to measure and analyze individual traits. There are many other traits, such as a person’s manner of walking, moving or writing/typing which can also be measured and used for authentication.