Identifying unique objects in multiple image collections

• US 8,386,505 B2
• Filed: 07/14/2011
• Issued: 02/26/2013
• Est. Priority Date: 04/07/2006
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Eastman Kodak Company
• Inventors: Gallagher, Andrew C.
• Primary Examiner(s): Singh, Amresh
• Application Number:
  US13/182,706
• Publication Number:
  US 20110268323A1
• Time in Patent Office: 593 Days
• Field of Search: 707/737, 707/758, 707/915, 707/999.003, 382/118
• US Class Current: 707/758
• 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 having a particular person found in at least two separate collections of different users by using a communication network, the method comprising:

a) identifying the particular person that is labeled with a name in a first collection of a first user and providing features for the particular person;

b) using a controller to identify a second separate image collection of a second user that potentially has images of the particular person and authorizing the second collection to be available for searching;

c) identifying features and names corresponding to faces in images from the second collection;

d) receiving a social relationship between the first user and the second user; and

e) using the features and the name from the first collection and features and names from the second collection and the social relationship between the first and second users to produce a probability that images from the second collection contain the particular person and using the probability to identify images containing the particular person by searching the second collection.

2. The method of claim 1 further including providing identified images in the second collection on a display for viewing by a user to determine if they contain the particular person.

3. A method of identifying images having a particular person found in at least two separate collections of different users by using a communication network, the method comprising:

a) identifying the particular person that is labeled with a name in a first collection of a first user;

b) using a controller to identify a second separate image collection of a second user that potentially has images of the particular person and authorizing the second collection to be available for searching;

c) identifying names corresponding to faces in images from the second collection;

d) providing a social relationship between the first and second users; and

e) using the name from the first collection, the names from the second collection and the social relationship between the users of the image collections to produce a probability that images from the second collection contain the particular person and using the probability to identify images containing the particular person by searching the second collection.

4. The method of claim 3 further including providing identified images in the second collection on a display for viewing by a user to determine if they contain the particular person.

5. The method of claim 3, wherein the name from the first collection is different from one of the names in the second collection and either the name in the first collection or one of the names in the second collection indicate a social relationship.