Systems and methods for multi-factor image recognition

• US 9,760,809 B2
• Filed: 10/20/2015
• Issued: 09/12/2017
• Est. Priority Date: 10/20/2014
• Status: Active Grant

• Current Assignee: BAE Systems Information and Electronic Systems Integration Incorporated (BAE Systems Plc)
• Original Assignee: BAE Systems Information and Electronic Systems Integration Incorporated (BAE Systems Plc)
• Inventors: Mercier, Michael N.
• Primary Examiner(s): Kholdebarin, Iman K
• Application Number:
  US14/887,863
• Publication Number:
  US 20160140426A1
• Time in Patent Office: 693 Days
• Field of Search: None
• 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: G06V 10/768 using context analysis, e.g. recognition aided by known co-occurring patterns
• CPC Class Codes: G06V 20/30 in albums, collections or shared content, e.g. social network photos or video
• CPC Class Codes: G06V 2201/10 Recognition assisted with metadata

Abstract

A mechanism for image recognition based on multiple factors is described. A method, system and computer-readable medium for multi-factor image recognition includes using environmental contextual attributes to create likelihood tiers in an image recognition database such that irrelevant entries are excluded from the search. The mechanism described here limits and sorts by contextual likelihood the number of entries to be searched, increasing both the speed and accuracy of the image recognition process.

Claims

1. A method for performing multi-factor image recognition comprising:

categorizing images in a database based on one or more environmental contextual attributes of the images;

identifying one or more environmental contextual attributes associated with an object in an observed image;

creating a plurality of likelihood tiers from the categorized images in the database based on the presence or absence of the one or more identified environmental contextual attributes; and

performing image recognition for the object using images in a first likelihood tier having the greatest number of the one or more environmental contextual attributes among the plurality of likelihood tiers.

2. The method of claim 1, further comprising:

identifying an image that matches the object from the first likelihood tier.

3. The method of claim 1, further comprising:

in response to identifying no matches for the object among the images in the first likelihood tier, performing image recognition for the object using images in a second likelihood tier having the next greatest number of the one or more environmental contextual attributes; and

identifying an image that matches the object from the second likelihood tier.

4. The method of claim 1, further comprising:

receiving a spectral profile indicating spectral or color-palate properties of an image.

5. The method of claim 4, further comprising:

associating at least one spectral profile with at least one image in the database; and

comparing at least one spectral observation with the at least one spectral profile.

6. The method of claim 4, further comprising:

associating at least one spectral profile with at least one image in the database; and

generating at least one contextual attribute for said the at least one image based on the at least one spectral profile.

7. A system for performing multi-factor image recognition comprising:

a memory; and

one or more processors in communication with the memory, the one or more processors configured to;

categorize images in a database based on one or more environmental contextual attributes;

identify one or more environmental contextual attributes associated with an object in an observed image;

create a plurality of likelihood tiers from the categorized images in the database based on the presence or absence of the one or more identified environmental contextual attributes; and

perform image recognition for the object using images in a first likelihood tier having the greatest number of the one or more environmental contextual attributes among the plurality of likelihood tiers.

8. The system of claim 7, wherein the one or more processors is a field-programmable gate array.

9. The system of claim 7, wherein the one or more processors is a microprocessor.

10. The system of claim 7, wherein the one or more processors are further configured to:

receive a spectral profile indicating spectral or color-palate properties of an image.

11. The system of claim 10, wherein the one or more processors are further configured to:

associate at least one spectral profile with at least one image in the database; and

compare at least one spectral observation with the at least one spectral profile.

12. The system of claim 10, wherein the one or more processors are further configured to:

associate at least one spectral profile with at least one image in the database; and

generate at least one contextual attribute for the at least one image in the database based on the at least one spectral profile.

13. The system of claim 7 wherein the one or more processors are further configured to:

identify an image that matches the object from the first likelihood tier.

14. The system of claim 7 wherein the one or more processors are further configured to:

in response to identifying no matches for the object among the images in the first likelihood tier, perform image recognition for the object using images in a second likelihood tier having the next greatest number of the one or more environmental contextual attributes; and

identify an image that matches the object from the second likelihood tier.

15. A non-transitory computer readable medium storing instructions executable by a processing device, the instructions implementing a method for performing multi-factor image recognition, execution of the instructions causes the processing device to:

identify one or more environmental contextual attributes associated with an object in an observed image;

categorize images in a database based on the one or more environmental contextual attributes;

create a plurality of likelihood tiers from the categorized images in the database based on the presence or absence of the one or more identified environmental contextual attributes; and

perform image recognition for the object using images in a first likelihood tier having the greatest number of the one or more environmental contextual attributes among the plurality of likelihood tiers.

16. The medium of claim 15, wherein execution of the instructions further causes the processing device to:

identify an image that matches the object from the first likelihood tier.

17. The medium of claim 15, wherein execution of the instructions further causes the processing device to:

in response to identifying no matches for the object among the images in the first likelihood tier, perform image recognition for the object using images in a second likelihood tier having the next greatest number of the one or more environmental contextual attributes; and

identify an image that matches the object from the second likelihood tier.

18. The medium of claim 15, wherein execution of the instructions further causes the processing device to:

receive a spectral profile indicating spectral or color-palate properties of an image.

19. The medium of claim 18, wherein execution of the instructions further causes the processing device to:

associate at least one spectral profile with at least one image in the database; and

compare at least one spectral observation with the at least one spectral profile.

20. The medium of claim 18, wherein execution of the instructions further causes the processing device to:

associate at least one spectral profile with at least one image in the database; and

generate at least one contextual attribute for said the at least one image based on the at least one spectral profile.