Most of fingerprint identification systems perform
matching algorithms based on different minutiae details present
in the fingerprint. Usually, minutiae are extracted from the
thinned fingerprint image. Due to the image noise and different
preprocessing methods, the thinned image could results in a large
number of false minutiae which may decrease the performance
of the system. In this paper, some existing algorithms from
different studies were integrated to build Automated Fingerprint
Identification System. A new matching algorithm was proposed
based on feature with two points; minutiae and ridge point.
In addition, each feature extracted assigned to an appropriate
weight according to proposed weights table. The modified system
was tested on FVC2002 DB1 set-A and all four FVC2004 set-
A databases and showed that it is effective and gives excellent
results that exceed the performance of classic minutiae-based
matching algorithm.