Abstract—Fingerprint re-alignment is an important aspect in the subject of fingerprint biometrics because a fingerprint recognition system that does not attempt to re-align an involuntarily rotated fingerprint can create opportunities for falsely rejecting a legitimate subject. It can also create opportunities for security attacks where an illegitimate subject voluntarily rotates their fingerprint, as a way of tempering with the functionality of the recognition system. This manuscript presents a new fingerprint re-alignment solution that uses the True Fingerprint Center Point (TFCP) as a reference, where the TFCP is introduced as the actual center of the fingerprint image foreground. Using the TFCP as a reference point proves – in general – to be fast, independent, and efficient for all types of fingerprints. It, in particular, is useful when dealing with the type of fingerprints that belong to the Plain Arch (PA) fingerprint class. The credibility of this proposed solution is interrogated through its applicability on fingerprint ridge surface estimation and fingerprint template matching.
Index Terms—rotation; re-alignment; plain arch; true fingerprint center point; ridge surface estimation; template matching
All authors are with the Biometrics Research Group, Information Security Competence, Modelling & Digital Sciences Unit, Council for Scientific & Industrial Research (CSIR), P.O. Box 395, Pretoria 0001, South Africa. For correspondence, phone: +27-12-841-3263; fax: +27-12-841-2456; e-mail: imsiza@csir.co.za
Cite: Ishmael S. Msiza, Tendani C. Malumedzha, and Brain Leke-Betechuoh, "A Novel Fingerprint Re-alignment Solution that Uses the TFCP as a Reference," International Journal of Machine Learning and Computing vol. 1, no. 3, pp. 297-304, 2011.