INTRODUCTION• Fingerprint is most popular,reliable and oldest biometric sign of identity• Touchless fingerprint system is a remote sensing technique to process fingerprint pattern, considered as a viable alternative to touch based fingerprint system• New generation of touchless live scan devices is 3D touchless finger print system
TERMINOLOGY• Fingerprint can be looked at from different levels1) GLOBAL LEVEL• Singularity points called core and delta points Core and delta points marked on sketches of the two fingerprint patterns loop and whorl
2) LOCAL LEVELMinutiae details in terms of ridgesRidge bifurcation Ridge termination Representation of minutiae
TOUCHLESS VERSUS TOUCH-BASED TOUCHLESS TOUCH-BASED SKIN DISTORTION NO YES SLIPPAGE,SMEARING NO YES FINGERPRINT RESIDUE NO YES CAPTURE AREA LARGE SMALL TOLERANCE ON SKIN CONDITION LARGE SMALL HIGH LOWUSER COMFORT LEVEL
IMAGE ACQUISITION MULTIPLE VIEW SYSTEMFigure 1. Fingerprint acquisition using a set of cameras surrounding the finger
• Multiple view enables the capture of full nail to nail fingerprints increasing the usable area• From each acquired image a silhouette is extracted.• The 5 silhouettes are then projected into the 3D space and we get the 3D shape of finger by knowing the position and orientation of each camera within a reference coordinate system.
Figure 2 Fingerprint acquisition obtained by combining a single line scan camera and two mirrors
3D FINGERPRINT UNWRAPPING• Unwrapping a 3D object refers to the unfolding the 3D object onto a flat 2D plane. UNWRAPPING METHOD PARAMETRIC NON PARAMETRIC
PARAMETRIC UNWRAPPING USING CYLINDRICAL MODEL• Human fingers vary in shape, like the shape of the middle finger is often more cylindrical than the thumb.• Human fingers can be closely approximated by cylinders.• Human fingers also vary in size and the cylindrical model can also capture this size variability in the vertical direction, but not in the horizontal direction.• Cylindrical model is a reasonable choice for parametric unwrapping of3D fingerprints.
T Parametric unwrapping using a cylindrical model (top down view). Point (x,y,z) on the 3D finger is projectedFigure 3 to ( ,z) on the 2D plane.
• Mathematically, let the origin be positioned at the bottom of the finger, centered at the principle axis of the finger.• Let T be a point on the surface of the 3D finger: x T = Y z
• This 3D point is then projected (transformed) onto the cylindrical surface to obtain the corresponding 2D coordinates S = z Where
NON PARAMETRIC -UNWRAPPING• Non-parametric unwrapping, does not involve any projection on parametric models.• The unwrapping directly applies to the object to preserve local distances or angular relations.• Guarantees the variability in both shape and size of fingers is preserved.• Less distortion compared to parametric unwrapping
PREPROCESSING STEPSa) Computation of local ridge frequency and local ridge orientationb) Enhancement of the fingerprint imagec) Segmentationd) Detection of singularities FEATURE EXTRACTIONa) Conversion of preprocessed fingerprint image into binary imageb) Thinning
FINGERPRINT MATCHING• MINUTIAE-BASED APPROACH :- Analogous with the way that forensic experts compare fingerprints• The minutiae sets of the two fingerprints to be compared are first aligned, requiring displacement and rotation to be computed• Region of tolerance around the minutiae position is defined in order to compensate for the variations that may appear in the minutiae position due to noise and distortion
DISADVANTAGE• Lower contrast between ridges and valleys due to motion blur of hand tremble , camera background noise and small depth of field• Unwrapping technique has distortion upto some extent• Compatibility with contact-based 2D rolled fingerprint image
REFERENCE• Intelligent biometrics technique in finger print and face recognition by L.C Jain, U.halice, S.B lee,S.T Sutsui,I.Hayashi• Tabassi E., Wilson C., and Watson C., “Fingerprint Image Quality,” Tech. Rep. 7151, National Institute of Standards and Technology (NIST), August 2004.• Parziale G. and Diaz-Santana E., “The Surround Imager: Multi-Camera Touchless Device to Acquire 3D Rolled- Equivalent Fingerprints,” in Proceedings of IAPR International Conference on Biometrics (ICB),Hong Kong, China, January 2006, pp. 244–250.