Identifying unique objects in multiple image collections

• US 20070239683A1
• Filed: 04/07/2006
• Published: 10/11/2007
• Est. Priority Date: 04/07/2006
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Eastman Kodak Company
• Inventors: Gallagher, Andrew
• Granted Patent: US 8,024,343 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/784 the detected or recognised objects being people
• CPC Class Codes: G06F 18/2321 using statistics or function optimisation, e.g. modelling of probability density functions
• CPC Class Codes: G06F 18/2323 based on graph theory, e.g. minimum spanning trees [MST] or graph cuts
• CPC Class Codes: G06V 10/763 Non-hierarchical techniques, e.g. based on statistics of modelling distributions
• CPC Class Codes: G06V 10/7635 based on graphs, e.g. graph cuts or spectral clustering
• CPC Class Codes: G06V 20/30 in albums, collections or shared content, e.g. social network photos or video
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G06V 40/179 metadata assisted face recognition
• CPC Class Codes: Y10S 707/915 Image

Abstract

A method of identifying images containing a unique object found in at least two separate image collections of different users comprising identifying the unique object and providing features for the unique object; at least one user identifying at least two separate image collections produced by separate users that potentially have images of the unique object; and using the features to search the at least two separate collections to identify images that contain the unique object.

Claims

1. A method of identifying images containing a unique object found in at least two separate image collections of different users, comprising:

a) identifying the unique object and providing features for the unique object;

b) at least one user identifying at least two separate image collections produced by separate users that potentially have images of the unique object; and

c) using the features to search the at least two separate collections to identify images that contain the unique object.

2. A method of identifying images having a particular unique object found in at least two separate image collections of different users, comprising:

a) identifying the unique object and providing features for the unique object;

b) at least two users identifying at least two separate image collections produced by separate users that potentially include images of the unique object and authorizing such collections to be available for searching; and

c) using the features to search the at least two authorized separate collections to identify images that contain the unique object.

3. The method of claim 2, wherein a first user identifies a unique object and communicates with a second user to determine that there is a likelihood that the collection of the second user contains the unique object.

4. A method of identifying images having a unique object found in at least two separate collections of different users, comprising:

a) identifying the unique object that is labeled with a label in a first collection and providing features for the unique object;

b) identifying at least two separate image collections produced by separate users that potentially have images of the unique object and authorizing such collections to be available for searching;

c) using the features and the label from the first collection to search the at least two authorized separate collections to identify images that may contain the unique object; and

d) viewing the identified images to determine if they contain the unique object.

5. The method of claim 4, wherein the unique object is a particular person and the label is a name associated with the particular person.

6. The method of claim 5 wherein the features include facial features, image capture location features or image capture time features.

7. The method of claim 5 wherein the particular person is labeled in the second collection and has a different name.

8. The method of claim 6 wherein the features include the different label names.

9. A method of identifying images containing a unique object found in at least two separate image collections of different users, comprising:

a) identifying and labeling the unique object in a first image collection;

b) producing features for the unique object from the first image collection;

c) establishing a link for sharing images or collection data between the first image collection and a second image collection that potentially has one or more images of the unique object; and

d) using the search features to produce at least one label identifying the unique object for an image in the second image collection.

10. A method of identifying images containing a unique object found in at least two separate image collections of different users, comprising:

a) identifying unique objects in a first image collection;

b) identifying unique objects in a second image collection and producing features for the unique object;

c) establishing a link for sharing images or collection data between the first image collection and a second image collection that potentially has one or more images of the unique object; and

c) using the features to produce at least one label identifying the unique object for an image in the first image collection.

11. A method of establishing a connection between two image collections, comprising:

a) providing a first image collection from a first user;

b) providing a second image collection from a different second user;

c) establishing a link between the first and second image collections;

d) detecting unique objects from the first and second image collections and producing features associated with the unique objects;

e) analyzing the features associated with the unique objects to produce a set of distinct unique objects;

12. The method of claim 11, wherein analyzing includes using a label from at least one of the image collections.

13. The method of claim 12, wherein the label is a name associated with a person, the unique objects are human faces and the distinct objects are individual people.

14. The method of claim 13, wherein the social relationship between first user and the second user is used when analyzing the first and second image collections.

15. The method of claim 11, wherein step e) includes using at least one label associated with a unique object from the first image collection and one label associated with a unique object from the second image collection and determining the likelihood that a unique object from the first image collection is the same as a unique object from the second image collection based on the labels.

16. The method of claim 15 wherein the labels are names associated with people, the unique objects are human faces, and the distinct objects are individual people.

17. A method of establishing a connection between two image collections, comprising:

a) providing a first image collection from a first user;

b) providing a second image collection from a different second user;

c) establishing a link between the first and second image collections;

d) detecting unique objects from the first and second image collections and producing features associated with the unique objects;

e) finding a specific unique object in the first image collection that is believed to also appear in the second image collection.

18. The method of claim 17, further comprising:

f) displaying for a user an image portion containing the specific unique object in the first image collection and an image portion believed to contain the specific unique object in the second image collection are displayed.

19. The method of claim 18, further comprising:

g) the user indicating through a user interface if the two image portions both contain the specific unique object.