Image classification and information retrieval over wireless digital networks and the internet

• US 7,587,070 B2
• Filed: 06/13/2008
• Issued: 09/08/2009
• 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): PATEL, KANJIBHAI B
• Application Number:
  US12/138,559
• Publication Number:
  US 20090060288A1
• Time in Patent Office: 452 Days
• Field of Search: 382115-118, 382/181, 382/218, 382/260, 382/305, 382/312, 707/104.1, 455/410
• US Class Current: 382/118
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

A method and system for matching an unknown facial image of an individual with an image of a celebrity using facial recognition techniques and human perception is disclosed herein. The invention provides a internet hosted system to find, compare, contrast and identify similar characteristics among two or more individuals using a digital camera, cellular telephone camera, wireless device for the purpose of returning information regarding similar faces to the user The system features classification of unknown facial images from a variety of internet accessible sources, including mobile phones, wireless camera-enabled devices, images obtained from digital cameras or scanners that are uploaded from PCs, third-party applications and databases. Once classified, the matching person'"'"'s name, image and associated meta-data is sent back to the user. The method and system uses human perception techniques to weight the feature vectors.

Claims

1. A system for matching an unknown facial image of an individual with an image of a celebrity, the system comprising:

a mobile communication device comprising means for generating a digital facial image of an individual and means for wireless transmitting the digital facial image;

an image classification server for receiving the digital facial image from the mobile communication device, the image classification server comprising

means for analyzing the digital facial image,

means for processing the digital facial image to generate a processed image,

means for comparing the processed image to a plurality of database processed images to create matches,

means for sorting the matches based on a predicted human perception to determine the best matched image;

means for transmitting the matched images and the perception value to the mobile communication device; and

a wireless network for transmission between the mobile communication device and the image classification server.

2. The system according to claim 1 further comprising means for weighting the voting to provide the predicted human perception.

3. The system according to claim 1 further comprising means for determining if an image of a celebrity should be used to create a feature vector.

4. The system according to claim 1 further comprising means for weighting the feature vector.

5. The system according to claim 4 wherein the vote weighting means comprises a support vector machine to input a source feature vector and a celebrity feature vector to predict a human perception vote.

6. The system according to claim 1 further comprising means for determining which feature vectors are relevant to the determination of the predicted human perception.

7. The system according to claim 1 wherein the analyzing means is a processor configured with facial recognition software.

8. A method for weighting feature vectors for images, the method comprising:

determining the quality of a plurality of matched images to generate preferred plurality of matched images, each of the plurality of matched images comprising a source image and a database image, the source image comprising a plurality of feature vectors and the database image comprising a plurality of feature vectors;

analyzing the preferred plurality of match images to determine which of the plurality of feature vectors are most closely related to human perception to generate a plurality of human perception feature vectors; and

weighting each of the plurality of human perception feature vectors based on the analysis.

9. The method according to claim 8 wherein the each of the source images of the plurality of matched images is provided by a third party, and each of the database images of the plurality of matched images is a celebrity image.

10. The method according to claim 8 wherein the analysis is based on voting by the general public.

11. The method according to claim 8 wherein the plurality of feature vectors are 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.

12. A method for matching images, the method comprising:

wirelessly transmitting a digital facial image from a user over a wireless network to an image classification server;

processing the digital facial image at the image classification server to create a primary feature vector for the digital facial image;

comparing the primary feature vector to a plurality of database feature vectors, each of the plurality of database feature vectors corresponding to a database processed image;

obtaining matched images from the primary feature vector and database feature vectors;

sorting the matched images based on predicted human perception; and

transmitting best matched images to the user.

13. The method according to claim 12 further comprising voting on the matched images.

14. The method according to claim 13 wherein sorting the matched images based on human perception comprises analyzing voting results for the matched images.

15. A method for matching images, the method comprising:

transmitting a digital facial image from a user over a wireless network to an image classification server;

identifying a pair of eyes of the digital facial image;

comparing the digital facial image to a preselected set of database images or an alternative image;

selecting a matching imaging from the preselected set of database images or the alternative image to create a matched images;

displaying the matched images based for voting to collect information on human perception;

analyzing the voting to determine if the matched images are a match based on human perception; and

transmitting the voting results to the user.