|
||||||||||||||||||||||||||||||||||||
|
AUTOMATIC PORE DETECTION IN FINGERPRINT IDENTIFICATION SYSTEMS USING K-MEANS CLUSTERING TECHNIQUE N. Basha, L. Shulga Applied Acoustics Research Institute; 7A, 9th May Str., Dubna, Moscow Region, 141980, Russia; natalia.basha@niipa.ru, luda.shulga@niipa.ru Pores detection is a stage of automatic fingerprint identification systems. With the advances in fingerprint sensing technology, modern sensors with resolution 1000 dpi allows us to capture high resolution fingerprints, where pores can be clearly detected (average density is about 6-8 pores in one cm of a ridge).We present an automatic pore detection technique, that divides each point of fingerprint in 3 classes: ridge, furrow and pore, using image intensity and k-means clustering algorithm. The obtained results are promising and show the effectiveness of proposed scheme (on a database of 100 fingerprints of 10 persons). Key words: pattern recognition, image analysis, biometrics, fingerprints, pore detection, k-means clustering technique.. |