Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Electrical and Computer Engineering

Committee Chair

Inanc, Tamer

Author's Keywords

3D minutiae detection; Quality map; Biometric; Minutiae extraction; 3D fingerprint; Curvature detection


Fingerprints--Data processing; Pattern recognition systems; Three-dimensional imaging


Traditionally, fingerprint image acquisition was based on contact. However the conventional touch-based fingerprint acquisition introduces some problems such as distortions and deformations to the fingerprint image. The most recent technology for fingerprint acquisition is touchless or 3D live scans introducing higher quality fingerprint scans. However, there is a need to develop new algorithms to match 3D fingerprints. In this dissertation, a novel methodology is proposed to extract minutiae in the 3D fingerprint scans. The output can be used for 3D fingerprint matching. The proposed method is based on curvature analysis of the surface. The method used to extract minutiae includes the following steps: smoothing; computing the principal curvature; ridges and ravines detection and tracing; cleaning and connecting ridges and ravines; and minutiae detection. First, the ridges and ravines are detected using curvature tensors. Then, ridges and ravines are traced. Post-processing is performed to obtain clean and connected ridges and ravines based on fingerprint pattern. Finally, minutiae are detected using a graph theory concept. A quality map is also introduced for 3D fingerprint scans. Since a degraded area may occur during the scanning process, especially at the edge of the fingerprint, it is critical to be able to determine these areas. Spurious minutiae can be filtered out after applying the quality map. The algorithm is applied to the 3D fingerprint database and the result is very encouraging. To the best of our knowledge, this is the first minutiae extraction methodology proposed for 3D fingerprint scans.