Method and system for tagging an image of an individual in a plurality of photos

• US 8,369,570 B2
• Filed: 12/22/2008
• Issued: 02/05/2013
• Est. Priority Date: 09/28/2005
• Status: Active Grant

• Current Assignee: Motorola Solutions, Inc.
• Original Assignee: FaceDouble Incorporated
• Inventors: Myers, Charles A., Shah, Alex
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): ANSARI, TAHMINA N
• Application Number:
  US12/341,318
• Publication Number:
  US 20100235400A1
• Time in Patent Office: 1,506 Days
• Field of Search: None
• US Class Current: 382/103
• CPC Class Codes: G06V 10/95 structured as a network, e.g. client-server architectures
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

A system and method for tagging an image of an individual in a plurality of photos is disclosed herein. A feature vector of an individual is used to analyze a set of photos on a social networking website such as www.facebook.com to determine if an image of the individual is present in a photo of the set of photos. Photos having an image of the individual are tagged preferably by listing a URL or URI for each of the photos in a database.

Claims

1. A method for tagging an image of an individual in a plurality of photos, the method comprising:

providing a first plurality of photos, each of the first plurality of photos comprising an identified facial image of the individual;

processing the facial image of the individual in each of the first plurality of photos to generate a feature vector for the individual, wherein the feature vector is based on a plurality of factors comprising facial expression, hair style, hair color, facial pose, eye color, texture of the face, color of the face and facial hair;

analyzing a second plurality of photos to determine if a facial image of the individual is present in a photo of the second plurality of photos, the analysis comprising determining if facial image in each of the photos of the second plurality of photos matches the feature vector for the individual;

identifying each of the photos of the second plurality of photos having facial image of the individual to create a third plurality of photos; and

tagging each of the photos of the third plurality of photos to identify the facial image of the individual in each of the third plurality of photos, wherein tagging comprises determining the X and Y coordinates of the facial image in the photo, storing the X and Y coordinates of the image in the photo in a database, storing an identifier for the photo in the database, and storing an identifier for the individual in a database.

2. The method according to claim 1 wherein the step of tagging further comprises listing a URL or URI for each of the photos of the third plurality of photos in a database.

3. The method according to claim 1 wherein determining if the facial image in each of the photos of the second plurality of photos matches the feature vector for the individual comprises using a statistical model for comparison of feature vectors weighted based on human perception.

4. The method according to claim 1 wherein the feature vector is weighted based on human perception feedback.

5. A system for tagging an image of an individual in a plurality of photos, the system comprising:

a network;

a database comprising a first plurality of photos of facial image of an individual;

a server engine for processing the first plurality of photos to generate a feature vector for the facial image of the individual;

a second plurality of photos located on a social networking web site;

wherein the server engine is configured to analyze the second plurality of photos to determine if facial image of the individual is present in a photo of the second plurality of photos, the analysis comprising determining if facial image in each of the photos of the second plurality of photos matches the feature vector for the individual;

wherein the server engine is configured to identify each of the photos of the second plurality of photos having an image of the individual to create a third plurality of photos; and

wherein the server engine is configured to tag each of the photos of the third plurality of photos to identify the image of the individual in each of the third plurality of photos, wherein the server engine is configured to determine the X and Y coordinates of the facial image in the photo, store the X and Y coordinates of the image in the photo in a database, store an identifier for the photo in the database, and store an identifier for the individual in a database.

6. The system according to claim 5 wherein the server engine is configured to list a URL or URI for each of the photos of the third plurality of photos in a database.

7. The system according to claim 5 wherein the feature vector is weighted based on human perception feedback.

8. The system according to claim 5 wherein the server engine is configured to determine if the facial image in each of the photos of the second plurality of photos matches the feature vector for the individual comprises using a statistical model for comparison of feature vectors weighted based on human perception the image of the individual is a facial image of the individual.

9. The system according to claim 7 wherein the human perception feedback is collected on the social networking web site.

10. The system according to claim 5 wherein the feature vector is based on a plurality of factors comprising facial expression, hair style, hair color, facial pose, eye color, texture of the face, color of the face and facial hair.

11. A method for tagging an image of an individual in a plurality of photos, the method comprising:

creating a feature vector for an individual from a plurality of reference photos;

analyzing a second plurality of photos to determine if an image of the individual is present in a photo of the second plurality of photos, the analysis comprising determining if an image in each of the photos of the second plurality of photos matches the feature vector for the individual;

identifying each of the photos of the second plurality of photos having an image of the individual to create a third plurality of photos;

determining X and Y coordinates of the facial image of the individual in each photo of the third plurality of photos;

determining the size of the facial image of the individual each photo of the third plurality of photos;

determining the tilt of the image of the individual in each photo of the third plurality of photos;

storing in a database the X and Y coordinates of the facial image in each photo of the third plurality of photos;

storing in the database an identifier for each photo of the third plurality of photos; and

storing in the database an identifier for the individual.

12. The method according to claim 11 wherein each of the second plurality of photos is located on a social networking web site.

13. The method according to claim 11 wherein determining if the facial image in each of the photos of the second plurality of photos matches the feature vector for the individual comprises using a statistical model for comparison of feature vectors weighted based on human perception.

14. The method according to claim 13 wherein the feature vector is weighted based on human perception feedback.

15. The method according to claim 14 wherein the human perception feedback is collected on social networking web site.