TY - JOUR
T1 - Pre-registration of Translated/Distorted Fingerprints based on Correlation and the orientation field
AU - Lan, Sheng
AU - Guo, Zhenhua
AU - You, Jia
PY - 2020/2
Y1 - 2020/2
N2 - In this study, we present a combined algorithm to register a translated/distorted finger- print against a reference fingerprint based on correlation and the orientation field. In the first step, correlation is adapted to realize rough registration for correction of the whole translation. As the second step, we register the location-aligned distorted fingerprint with the reference fingerprint by using the orientation filed to correct the nonrigid deforma- tion. By registering a translated/distorted fingerprint against a reference fingerprint, we can increase the similarity of the corresponding fingerprints, thereby improving the result of fingerprint matching. Experiments have been carried out on three databases contain- ing many translated/distorted fingerprints, namely, FVC2004 DB1, the Tsinghua Distorted Fingerprint database and the NIST SD27 database. We also compared our algorithm with other registration algorithms, and the experimental results show improvement in finger- print matching when using the proposed algorithm for pre-registration.
AB - In this study, we present a combined algorithm to register a translated/distorted finger- print against a reference fingerprint based on correlation and the orientation field. In the first step, correlation is adapted to realize rough registration for correction of the whole translation. As the second step, we register the location-aligned distorted fingerprint with the reference fingerprint by using the orientation filed to correct the nonrigid deforma- tion. By registering a translated/distorted fingerprint against a reference fingerprint, we can increase the similarity of the corresponding fingerprints, thereby improving the result of fingerprint matching. Experiments have been carried out on three databases contain- ing many translated/distorted fingerprints, namely, FVC2004 DB1, the Tsinghua Distorted Fingerprint database and the NIST SD27 database. We also compared our algorithm with other registration algorithms, and the experimental results show improvement in finger- print matching when using the proposed algorithm for pre-registration.
U2 - 10.1016/j.ins.2020.02.017
DO - 10.1016/j.ins.2020.02.017
M3 - Journal article
SN - 0020-0255
VL - 520
SP - 292
EP - 304
JO - Information Sciences
JF - Information Sciences
ER -