First of all we have to differentiate between definitions:
Finger traces are remains left by the human finger on objects after touching them. In most cases this image is incomplete, and blurry.
Fingerprints are made by rolling the fingertip on the surface, gaining a full and good quality image of the finger. This is mostly used by the police.
Finger marks are left on a flat surface. These are also partial images, but with far better quality than finger traces.
Since most people refer to this technology as ’Fingerprint’ identification, we will also do, for the sake of convenience.
Fingerprint identification has a number of approaches. Certain techniques use emulations of the polices traditional sample identification, others are simple shape identifying tools, and there are also some unique approaches, like ultrasonic scans.
Some recognize whether the tested finger is alive or dead, others do not. Presently this is the most widely used biometric identification method in the world.
In cases of an adequately accurate (with the lowest false acceptance rate) and potentially acceptable fingerprint readers, we often encounter flaws of usage, caused by „inordinate” users. Fingerprint analysis can be a good choice for indoor use, where coverage about proper usage is available, and it runs in a controlled environment.
Sensors can be optical, capacitive or thermo-element based. Their common fault is that without a separate live-finger identification system all of them can be deceived. The newest development is the Lumidgm (we will start testing it in February 2012) technology, which is quite AN unorthodox system, since it analyses not only the surface of the finger, but the patterns of capillaries and fat tissues underneath.