Securing fingerprint templates using fused structures
Mulagala Sandhya, Munaga V. N. K. Prasad · IET Biometrics · 2016
This study presents a protection method for fingerprint templates by using fused structures at the feature level. The authors compute two transformed features from minutiae points: namely, local structure and distant structure. These structures are represented as bit‐strings. A fusion on bit‐strings is done at the feature level to produce a cancelable template. An equal error rate (EER) of 2.19, 1.6 and 6.14% on Fingerprint Verification Competition (FVC) 2002 Database (DB)1 through DB3 databases and an EER of 11.89, 12.71, 17.6% on FVC 2004 DB1 through DB3 proves the tenability of the proposed method.