Hometown Security: Real Time, 3D Finger and Palm-Print Scanner for Entry and Access Portal Security

This proposal directly addresses CIP R&D Theme #3: Entry and Access Portals to prevent unauthorized access to important places and systems by developing a structured light illumination system that acquires a 3-D surface scan of a human subject's hand with sufficiently high resolution as to record the 3-D shape of each finger's fingerprint ridges along with the palm-print. A rolled-equivalent print, which meets FBI and NIST standards, will then be established by morphing the 3-D skin model to a 2-D surface mimicking the process of rolling an inked finger. The proposed system can be used to directly address Theme #3's sub-topics for a) Identification, b) Authentication, c) Authorization, and d) Access Control. The scanning prototype will be a box that the subject inserts their hand into, palm down, with the component cameras working in a low-resolution mode to identify the hand as being left or right as well as recognizing the position and pose of each finger. The system will automatically alert the operator/user if the hand is ill posed with partial or complete occlusion of a fingerprint. Once the hand's pose is correct, the 3-D surface is acquired using flash photography within a 200 milli-second window. An innovative technique is used to create a high-resolution depth map from each digital image obtained using a single structured-light pattern projected by a flash synchronized with a digital camera; overlapping depth maps from multiple cameras and light sources are merged to create a 3-D surface model with detail on both the top and sides of the fingers and palm. Because the proposed scanner is non-contact, there are no optics or scanning surface for an operator to clean between subjects. Furthermore, a calibration pattern will be positioned above the subject's hand, inside the box, such that the cameras can automatically calibrate themselves prior to each scan, and hence eliminate performance drift. Because the proposed system will use commodity cameras, the hardware involved will be low cost with future iterations of the device being compact and easily portable. To facilitate collaboration in commercial production of these scanners, FiashScan3D will relocate their place of business from Dallas, Texas to Somerset Community College as directed by David A. Wiles, Chief of Community & Workforce Development, at 230 Air Port Road, Somerset, KY 42501. To facilitate adequate testing, several Beta Units will be tested by the Kentucky State Police Automated Fingerprint Identification Section in Frankfort, Kentucky.