Based on the type of biometric factor they use, biometric solutions can be divided into two categories:
- solutions based on a physiological factor
- solutions based on a behavioral factor
Solutions that are based on a physiological factor are fingerprint recognition, iris recognition and face recognition. Biometric solutions based on a behavioral factor are voice recognition and keystroke dynamics.
Fingerprint recognition is one of the most well known biometrics, and it is by far the most used biometric solution for authentication on computerized systems. Reasons which make fingerprint recognition so popular are:
- easy acquisition of fingerprint images;
- established use and acceptance when compared to other biometrics, which is due to a number of government programs such as implementations by the US Department of Homeland Security;
- there are ten sources of this biometric on each individual, injury is unlikely to disable an individual from identifying himself;
- numerous commercial devices and applications are available.
Iris recognition is relatively young, being only commercially developed the last decade mostly due to previous patent limitations. This makes commercial iris recognition devices and applications expensive, and more difficult to implement when compared to for example fingerprint recognition. However, iris recognition is the biometric solution of the future for a number of reasons:
- the iris does not change with ageing and damage due to injury or disease is less likely than with other biometrics;
- acquisition of a good image is easier than with face recognition or retina scans;
- iris recognition algorithms have unprecedented false match rates and high speeds of comparisson, which makes this biometric suitable for high throughput applications such as airports.
Face recognition has an important advantage to other biometrics, which is that faces can be captured from some distance away even without the subject being aware of it. Therefore face recognition is the biometric solution for surveillance applications. An important difficulty with face recognition is the acquisition of images that can be compared by a matching algorithm, especially in surveillance applications when subjects do not look directly into the camera.
Voice recognition or speaker recognition is especially interesting for telephony based applications. The subject or the speaker is recognized by the unique accoustic features of his voice pattern.
There is often confusion in between speaker recognition (recognizing who is speaking) and speech recognition (recognizing what is being said).
Keyboard dynamics, also known as typing biometrics or keystroke dynamics has an important advantage that it requires no special hardware, it only needs a standard computer keyboard. Furthermore it can be run in the background, while a user is already logged on, to continuously monitor that no unauthorized user takes over the session. Keystroke dynamics is not the best biometric solution in terms of FAR and FRR, however it is ideal for multifactor authentication when combined with user ID and password.
Suitability of biometric solutions
The following 7 criteria are used throughout this website to evaluate the suitability of biometric solutions:
- Universality – the biometric factor, on which the solution is based, must be something each person has.
- Uniqueness – with the factor it must be possible to separate individuals from another, this must be possible and not be overly expensive with existing technical means. Biometric solutions that can differentiate identical twins are of course preferred.
- Permanence – the biometric factor must well resist the effect of time, this concerns ageing but also the likelihood of damage due to injury or disease.
- Collectability – measuring the factor must not be overly expensive or time consuming.
- Acceptability – biometrics are not always very well accepted. This is generally dependent on people’s view and on how invasive a certain technique is. For example retina scans have low acceptability with the general public they require direct contact with the reader and the scans can demonstrate certain medical conditions such as high bloodpressure.
- Circumvention – the biometric factor must be difficult to immitate, the biometric solution must also allow liveness detection.
- Performance – depending on the application speed, accuracy and robustness become more important.