Fingerprint Science Group

HandShot: A Fast 3-D Imaging System for Capturing Fingerprints, Palm Prints and Hand Geometry

by Latanya Sweeney PhD, Victor Weedn MD JD, Ralph Gross


Taking advantage of plummeting digital camera costs and increasing computer-controlled macro photography capabilities of commercial off-the-shelf chip-cameras as well as the computational expertise of Carnegie Mellon University and the forensic science knowledge of Duquesne University Law School, we present a direct high-contrast, high-resolution imaging system to simultaneously obtain 1000 ppi images of all 10 rolled-equivalent finger and both palm friction ridge patterns and minutiae within less than 5 seconds. We call this the “HandShot ID System.” We obtain well-focused, high-resolution images through the use of multiple cameras. We obtain high-contrast, surface-topology-discriminant images in HandShot through the use of oblique blue light and special spot lights from various angles operating sequentially in millions of a second. We obviate the need for an operator to individually roll fingers through the use of cameras positioned at different angles. By obviating the rolling of fingers, we enable truly high-throughput processing. We also eliminate sequencing errors by having physically distinct right and left hand non-contact placement areas and by simultaneously capturing all 10 digits and both palms. We eliminate the need for routine cleaning of a glass platen by eliminating any glass platen and instead directly image hands in the air –i.e., capture is contactless. For the same reason, we eliminate hygiene concerns.

In order for HandShot to accurately and instantly capture and record friction ridge skin detail on 10 rolled-equivalent fingerprints and both palm prints within 5 seconds, HandShot constructs a visual 3-dimensional model of both hands, including palms, fingerprints, fingertips, and sides of the fingers. Our associated HandShot algorithms: (1) stitch images from multiple cameras together forming a complete 3-D model of both hands using active models to normalize for pose; (2) extract ridge detail based on contrast assessment under varying illumination effects; and, (3) translate 3-D images to standardized formats. Having robust, detailed and extensive images supports traditional, as well as, new and improved applications in criminal, security and commercial identification, recognition and authentication tasks. For example, extraction of 10 rolled equivalent fingerprints from the 3-D model is possible for forensic comparison. Similarly, partial latent prints appearing on the palm, fingers (or sides) can be matched. Hand geometry and other new fast methods can be supported for security and commercial uses.

HandShot complies with significant standards in the field. Image specifications meet or exceed the FBI’s CJIS EFTS as well as the NIST M1 committee standards for fingerprints and palm prints. Output complies with ANSI/NIST-ITL 1-2000 data format interchange standards.

Keywords: fingerprint capture, palm prints, friction ridge patterns, minutiae, hand geometry, AFIS

L. Sweeney, V. Weedn, and R. Gross. A Fast 3-D Imaging System for Capturing Fingerprints, Palm Prints and Hand Geometry: the HandShot ID System. Carnegie Mellon University, School of Computer Science Tech Report CMU-ISRI-05-105. Pittsburgh, PA: 2004. (PDF)

This work was sponsored by the National Institute of Justice. Partners are Carnegie Mellon University and Duquesne University.

Real-time hand tracking and fitting (.mpg)

Related Links

Copyright © 2011. President and Fellows Harvard University.   |   IQSS   |    Data Privacy Lab   |    []