Open AccessPencocokan Citra Sidik Jari Menggunakan Korelasi Silang Ternormalisasi
Author(s) -
Bulkis Kanata
Publication year - 2015
Publication title -
jurnal rekayasa elektrika
Language(s) - English
Resource type - Journals
eISSN - 2252-620X
pISSN - 1412-4785
DOI - 10.17529/jre.v11i4.2405
Subject(s) - fingerprint (computing) , artificial intelligence , matching (statistics) , similarity (geometry) , pattern recognition (psychology) , computer science , cross correlation , computer vision , image (mathematics) , mathematics , template matching , matlab , statistics , operating system
Fingerprint image matching is an important procedure in fingerprint recognition. Robust fingerprint image matching under a variety of different image capture conditions is difficult to achieve, because of changes in finger pressure, variation of the angle, etc. Fingerprint matching is very important for the development of fingerprint system recognition that is sensitive to finger pressure. This paper proposes a fingerprint matching algorithm that enables the so-called fingerprint template (extracted specific part (region of interest (ROI)) of a person’s fingerprints to be matched to the different fingerprint of the same person or different people taken on different time, angle and a different finger pressure using normalized cross-correlation (NCC). This algorithm was implemented in MATLAB. The results showed that the maximum NCC value for ROI of the source fingerprints and targets that was greater than 0.62 indicates a strong correlation or similarity.