Fingerprint Recognition by Using Pores Extraction

Fingerprints are the oldest and most broadly used form of biometric identification. Everyone is known to have unique, immutable fingerprints.Fingerprint identification and recognition are considered popular technique in many security and law enforcement applications. Many systems rely on the matching of fingerprints using various methods and algorithms based on find position and orientation of ridge endings and bifurcations within the fingerprint image in both levels (level 2 and level 3). In this research, the sweat pores found in fingerprint image (level 3), their position and their type (open and closed) are detected. All they are considered as an important in matching phase by calculating number of pores found in the specified position. In addition, (level 2) we have focuses on another feature (bifurcation) which is found in detected ridges by applying a (3x3) mask to find bifurcation point. Also, the find distances between bifurcation point are found.The numbers of samples were (10) for each person, five with light intensity and the other five with dark. 8 samples used as training samples.The designed model has been trained and tested using a database consisting of (1000) image with selected fingerprint's images for (100) persons. Each person has 10 images with different contrast and moves. Results of fingerprints matching rate in level 3 is (100%) for pores by using Closet Point algorithm and in level 2, is (99.4%) for bifurcations points by using Iterative Closest Point algorithm.The time consumed in level 2 for the whole image (320x240 pixels) matching process is obtain (2) minute, while it consumes (1) minute when it is performed within a selected window of size 160x120. In level3, the time consumed for 160x120 image matching in level 3 was (22) second while it is (47) seconds for the whole image.