Theoretical Framework for Biometrics
Theoretical Framework
In the Theory of Fingerprint Verification of A.J. Zeelenberg, 1993 the skin on inside of a finger is covered with pattern of ridges and valleys. Already centuries ago it was studied whether these patterns are different among individuals. Indeed every person is believed to have unique fingerprints. This makes fingerprints suitable for verification of the identity of their owner. Although some fingerprint recognition systems do the comparison on the basis of actual recognition of the pattern, most systems use only specific characteristics in the pattern of ridges. These characteristics are a consequence from the fact that the papillary ridges in the fingerprint pattern are not continuous lines but lines that end, split into forks (called bifurcation), or form an island. These special points are called minutiae and although in general a fingerprint contains about hundred minutiae, the fingerprint area that is scanned by a sensor usually contains about 30 to 40 minutiae.
For over a hundred years of law enforcement agency all over the world use minutiae to accurately identify persons. For a positive identification that stands in European courts, at least 12 minutiae have to be identified in the fingerprint. The choice of 12 minutiae is often referred to as “the 12 point rule”. This 12 point rule is not based on statistical calculations but is empirically defined based on the assumption that even when a population of tenths of millions of persons are considered, no two persons will have 12 coinciding minutiae in their fingerprints. Most commercially available fingerprint scanners give a positive match when 8 minutiae are found. Manufacturers claim a FAR of one in a million based on these 8 minutiae, which seems reasonable.
(http://cryptome.org/fake-prints.htm, 19 May 2002)
Figure 1 shows a simplified way of how a fingerprint system works. Feature
Extraction
Capture and
Enhancement
In the Theory of Fingerprint Verification of A.J. Zeelenberg, 1993 the skin on inside of a finger is covered with pattern of ridges and valleys. Already centuries ago it was studied whether these patterns are different among individuals. Indeed every person is believed to have unique fingerprints. This makes fingerprints suitable for verification of the identity of their owner. Although some fingerprint recognition systems do the comparison on the basis of actual recognition of the pattern, most systems use only specific characteristics in the pattern of ridges. These characteristics are a consequence from the fact that the papillary ridges in the fingerprint pattern are not continuous lines but lines that end, split into forks (called bifurcation), or form an island. These special points are called minutiae and although in general a fingerprint contains about hundred minutiae, the fingerprint area that is scanned by a sensor usually contains about 30 to 40 minutiae.
For over a hundred years of law enforcement agency all over the world use minutiae to accurately identify persons. For a positive identification that stands in European courts, at least 12 minutiae have to be identified in the fingerprint. The choice of 12 minutiae is often referred to as “the 12 point rule”. This 12 point rule is not based on statistical calculations but is empirically defined based on the assumption that even when a population of tenths of millions of persons are considered, no two persons will have 12 coinciding minutiae in their fingerprints. Most commercially available fingerprint scanners give a positive match when 8 minutiae are found. Manufacturers claim a FAR of one in a million based on these 8 minutiae, which seems reasonable.
(http://cryptome.org/fake-prints.htm, 19 May 2002)
Figure 1 shows a simplified way of how a fingerprint system works. Feature
Extraction
Capture and
Enhancement