Face recognition system and method

• US 7,221,809 B2
• Filed: 12/17/2002
• Issued: 05/22/2007
• Est. Priority Date: 12/17/2001
• Status: Expired due to Fees

• Current Assignee: Technest Holdings, Inc. (AccelPath, Inc.)
• Original Assignee: Genex Technologies Incorporated (AccelPath, Inc.)
• Inventors: Geng, Z. Jason
• Primary Examiner(s): Bella; Matthew C.
• Assistant Examiner(s): AZARIAN, SEYED H
• Application Number:
  US10/322,315
• Publication Number:
  US 20030123713A1
• Time in Patent Office: 1,617 Days
• Field of Search: 382/103, 382/106, 382107-109, 382115-118, 382/154, 382/318, 382/278, 382280-285, 382/274, 382/255, 382/219, 382/189, 382/182, 345418-419, 348/448, 348/208.4, 348/50, 348/51, 348/207.99, 348/207.1, 348/207, 348/207.2, 352/57
• US Class Current: 382/280
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 10/772 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
• CPC Class Codes: G06V 20/64 Three-dimensional objects
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/174 Facial expression recognition

Abstract

A method of automatically recognizing a human face includes developing a three-dimensional model of a face; and generating a number of two-dimensional images based on the three-dimensional model. The generated two-dimensional images are then enrolled in a database and searched against an input image to identifying the face of the input image.

Claims

1. An improved security surveillance process for comparing a plurality of 2D pictures of individuals comprising a first group with 2D pictures of individuals comprising a second group, said process comprising the steps of:

generating 3D image data for facial pictures of each individual in said first group;

generating a plurality of angularly offset 2D images from said 3D image data for each individual in said first group;

generating a plurality of 2D picture images for individuals in a second group within a surveillance field; and

comparing ones of said plurality of 2D images for an individual of said first group with a 2D image for an individual in said second group.

2. The process of claim 1 additionally including the step of identifying an individual of said second group based upon a best match of facial features with ones of said plurality of 2D images for each individual in said first group.

3. The process of claim 1 wherein said step of comparing comprises the step of optically or electronically comparing facial features data based upon triangular measurements of like data for members of said first and second group respectively.

4. An improved surveillance system for identifying an individual based upon previously recorded 3D facial image data sets for a plurality of individuals comprising a first group of individuals, said system comprising:

a 3D camera for generating 3D facial image data of a plurality of individuals comprising said first group of individuals;

means for generating a plurality of 2D images for each individual in said first group by slicing said 3D facial data for each individual at a plurality of predetermined angular offsets of said 3D image data;

a 2D camera for generating pictures of a plurality of individuals comprising a second predetermined group within an optical field of surveillance; and

a comparator for successively analyzing a plurality of 2D images of individuals of said first group with 2D images of individuals of said second group to identify a possible match of individuals included in both said first and second groups.

5. The improved surveillance system of claim 4 wherein said 2D image generator generates a plurality of 2D images for each individual in said first group, each image being at a predetermined angular offset to a selected reference point of said 3D data set and additionally including positioning means for selecting a predetermined angular offset for facial angles for each successive 2D image.

6. The improved surveillance system of claim 5 wherein said comparator comprises an optical or electrical comparator for comparing predetermined or selectable facial feature data of individuals in said first and second groups of individuals.

7. The improved surveillance system of claim 4 additionally including a memory for storing said plurality of 2D data images generated from said 3D data set for each individual in said first.

8. A method of automatically recognizing a human face, said method comprising:

developing a three-dimensional model of each of a plurality of faces;

generating a plurality of two-dimensional images based on said three-dimensional models, wherein said two-dimensional images illustrate the face of a model from different angles;

developing a database of said two-dimensional images based on said three-dimensional facial models; and

comparing a two-dimensional input image of a target individual with images in said database so as to recognize said target individual.

9. The method of claim 8, wherein said two-dimensional images illustrate the face of a model in different lighting conditions.