Fingerprint recognition and iris scanning are the most well-known forms of biometric security. However, facial recognition and (finger and palm) vein pattern recognition are also gaining in popularity. In this article we consider the pros and cons of all these different techniques.
1. Fingerprint recognition
An identification system based on fingerprint recognition looks for specific characteristics in the line pattern on the surface of the finger. The bifurcations, ridge endings and islands that make up this line pattern are stored in the form of an image.
The disadvantage of capturing an image of an external characteristic is that this image can be replicated – even if it is stored in encoded form. An image is still an image, after all, and can therefore be compared. In principle, you can then generate the same code. Fingerprints can already be spoofed* using relatively accessible technology. Another, by no means insignificant, point to consider is that a finger presented for recognition does not necessarily still need to be attached to a body...
In addition, some line patterns are so similar that in practice this can result in a high false acceptance rate.** Fingerprints can also wear away as you get older, if you do a lot of DIY or a particular kind of work, for example. As a result, some people may find that their fingerprints cannot be recognised (false rejection**) or even recorded. There is even a hereditary disorder that results in people being born without fingerprints!
On the other hand, fingerprint identification is already familiar to much of the public and is therefore accepted by a large number of users. The technology is also relatively cheap and easy to use. It should be noted, however, that quality can vary significantly from one fingerprint recognition system to another, with considerable divergence between systems in terms of false acceptance and false rejection rates.
* Biometric spoofing refers to the presentation of a falsified biometric characteristic with the aim of being identified as another person. This may involve using a replicated fingerprint or a contact lens with a falsified iris pattern. The risk of spoofing mainly applies to forms of biometric security based on superficial external characteristics.
** Find out more about false acceptance and false rejection in our article ‘FAR and FRR: security level versus user convenience’.
2. Facial recognition
A facial recognition system analyses the shape and position of different parts of the face to determine a match. Surface features, such as the skin, are also sometimes taken into account.
Facial recognition for security purposes is an offshoot of face detection technology, which is used to identify faces in complex images in which a number of faces may be present. This technology has developed rapidly in recent years and is therefore an excellent candidate if a system is needed for remote recognition. Another plus is that the technology allows ‘negative identification’, or the exclusion of faces, making it a good deal easier to scan a crowd for suspicious individuals.
However, facial recognition also has a number of significant drawbacks. For example, the technology focuses mainly on the face itself, i.e. from the hairline down. As a result, a person usually has to be looking straight at the camera to make recognition possible. And even though the technology is still developing at a rapid pace, the level of security it currently offers does not yet rival that of iris scanning or vein pattern recognition.
3. Iris recognition
When an iris scan is performed a scanner reads out the unique characteristics of an iris, which are then converted into an encrypted (bar)code. Iris scanning is known to be an excellent security technique, especially if it is performed using infrared light.
However, one problem frequently encountered when the technology is introduced is resistance from users. Quite a few people find having their eyes scanned a rather unpleasant experience. You also have to adopt a certain position so the scanner can read your iris, which can cause discomfort. Hygiene is another frequently cited drawback, as many systems require users to place their chin on a chin rest that has been used by countless people before them.
Lastly, it is important to bear in mind that although iris scanning offers a high level of security, this may come at the expense of speed. Incidentally, systems have recently been developed that can read a person’s iris from a (relatively short) distance.
4. Finger vein pattern recognition
In the case of vein pattern recognition the ending points and bifurcations of the veins in the finger are captured in the form of an image, digitised and converted into an encrypted code. This method, combined with the fact that veins are found beneath rather than on the surface of the skin, makes this technology considerably more secure than fingerprint-based identification, as well as faster and more convenient for the user. It is a more expensive method, however.
Another point to bear in mind is that very cold fingers and ‘dead’ fingers (such as those of people suffering from Raynaud’s syndrome) are impossible or difficult to read using finger vein pattern recognition. Perhaps the greatest drawback, however, is that this technology is still relatively unknown.
5. Palm vein pattern recognition
This technique is also based on the recognition of unique vein patterns. However, as more reference points are used than in the case of finger vein pattern recognition, this is an even simpler and more secure identification method.
The technology, which cannot be copied (or only with extreme difficulty), is currently regarded as the best available method in the area of biometric security, alongside iris scanning. Palm scanning is fast and accurate and offers a high level of user convenience.
Access control systems based on palm vein pattern recognition are relatively expensive. For that reason such systems are mainly used within sectors that have exacting demands when it comes to security, such as government, the justice system and the banking sector.
Please note that this recognition method is sometimes confused with hand geometry. However, that is an outdated form of biometrics that is based on the shape of the hand and involves even fewer unique characteristics than fingerprint recognition.