TY - GEN
T1 - Contactless 3D fingerprint identification without 3D reconstruction
AU - Zheng, Qian
AU - Pathak, Ajay Kumar
AU - Pan, Gang
PY - 2018/6/29
Y1 - 2018/6/29
N2 - Recovery of 3D fingerprint data using photometric stereo generates 3D surface normal and albedo, which forms rich 3D fingerprint surface information. These surface normal's are further subjected to the reconstruction process, which integrates the surface normal to generate depth data. Since the source of depth information is essentially the surface normal, it is prudent to examine if this source information can itself be used for 3D fingerprint identification. In addition to avoiding the errors introduced by well-known integrability problem, such an approach can also enable significantly faster identification as the 3D reconstruction is the most computationally complex operation before the template matching. This paper investigates such an approach for 3D fingerprint identification using recovered surface normal and albedo information. We use publicly available 3D fingerprint database from 240 clients for the performance evaluation. The experimental results presented in this paper are highly promising, validates our approach, and indicate promises from matching contactless 3D fingerprints without the 3D surface reconstruction.
AB - Recovery of 3D fingerprint data using photometric stereo generates 3D surface normal and albedo, which forms rich 3D fingerprint surface information. These surface normal's are further subjected to the reconstruction process, which integrates the surface normal to generate depth data. Since the source of depth information is essentially the surface normal, it is prudent to examine if this source information can itself be used for 3D fingerprint identification. In addition to avoiding the errors introduced by well-known integrability problem, such an approach can also enable significantly faster identification as the 3D reconstruction is the most computationally complex operation before the template matching. This paper investigates such an approach for 3D fingerprint identification using recovered surface normal and albedo information. We use publicly available 3D fingerprint database from 240 clients for the performance evaluation. The experimental results presented in this paper are highly promising, validates our approach, and indicate promises from matching contactless 3D fingerprints without the 3D surface reconstruction.
KW - 3D fingerprint identification
KW - 3D fingerprint reconstruction
KW - Biometrics
KW - Contactless fingerprint matching
UR - http://www.scopus.com/inward/record.url?scp=85050530150&partnerID=8YFLogxK
U2 - 10.1109/IWBF.2018.8401566
DO - 10.1109/IWBF.2018.8401566
M3 - Conference article published in proceeding or book
AN - SCOPUS:85050530150
T3 - IWBF 2018 - Proceedings: 2018 6th International Workshop on Biometrics and Forensics
SP - 1
EP - 6
BT - IWBF 2018 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 6th International Workshop on Biometrics and Forensics, IWBF 2018
Y2 - 7 June 2018 through 8 June 2018
ER -