| US 7,512,255 B2 | ||
| Multi-modal face recognition | ||
| Ioannis Kakadiaris, Houston, Tex. (US); and Theoharis Theoharis, Houston, Tex. (US) | ||
| Assigned to Board of Regents, University of Houston, Houston, Tex. (US) | ||
| Filed on Aug. 23, 2004, as Appl. No. 10/923,982. | ||
| Claims priority of provisional application 60/497295, filed on Aug. 22, 2003. | ||
| Prior Publication US 2005/0084140 A1, Apr. 21, 2005 | ||
| Int. Cl. G06K 9/00 (2006.01); G06K 9/54 (2006.01); G06K 7/00 (2006.01) | ||
| U.S. Cl. 382—118 [382/305; 382/312] | 44 Claims |

| 1. A method for positively identifying an individual, comprising:
using an identification system to perform the steps of:
a) acquiring data from an individual's face, using a plurality of input units of one or more modalities, to generate acquired
data;
b) processing the acquired data to generate a geometric image;
c) using a generic geometric model template of the human face;
d) fitting the generic geometric model template to the digital data of a candidate human subject according to a first mathematical
method, to create a fitted model;
e) extracting metadata from the fitted model to describe the individual according to a second mathematical method;
f) storing the metadata in a machine readable format in a storage unit;
g) ranking entries of a database of previously stored metadata in a descending order according to their similarity to a new
set of metadata; and
h) deciding whether the best ranked entry provides positive identification for the candidate human subject.
|