Systems, methods, and media for identifying object characteristics based on fixation points

• US 9,501,710 B2
• Filed: 07/01/2013
• Issued: 11/22/2016
• Est. Priority Date: 06/29/2012
• Status: Active Grant

• Current Assignee: Arizona Board of Regents (University of Arizona)
• Original Assignee: Arizona Board of Regents (University of Arizona)
• Inventors: Karam, Lina, Dodge, Samuel
• Primary Examiner(s): Sherali, Ishrat I
• Application Number:
  US13/932,969
• Publication Number:
  US 20140016859A1
• Time in Patent Office: 1,240 Days
• Field of Search: 382/159, 382/305
• US Class Current: 1/1
• CPC Class Codes: G06V 10/44 Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

Systems, methods, and media for optical recognition are provided. In some embodiments, systems for optical recognition are provided, the systems comprising: at least one hardware processor that: identifies a plurality of fixation points in optically detected data; identifies features of the plurality of fixation points; and identifies one or more characteristics of an object represented in the optically detected data. In some embodiments, methods for optical recognition are provided, the methods comprising: identifying a plurality of fixation points in optically detected data using a hardware processor; identifying features of the plurality of fixation points using the hardware processor; and identifying one or more characteristics of an object represented in the optically detected data using the hardware processor. In some embodiments, non-transitory computer-readable media containing computer-executable instructions that, when executed by a hardware processor, cause the processor to perform these methods for optical recognition are provided.

Claims

1. A system for optical recognition, comprising:

at least one hardware processor that;

computes a saliency map of an image;

identifies a plurality of image fixation points in the image based on the saliency map; and

identifies an object represented in the image by comparing a shape formed by the plurality of image fixation points to a shape formed by a plurality of model fixation points in a model of a known object.

2. The system of claim 1, wherein the image is a video frame.

3. The system of claim 1, wherein the plurality of image fixation points are identified using a visual attention model.

4. The system of claim 1, wherein the plurality of image fixation points are identified by finding salient features of the image.

5. The system of claim 1, wherein the plurality of image fixation points are identified by simulating a foveation process.

6. The system of claim 1, wherein the plurality of image fixation points are identified by blurring at least a portion of the image.

7. The system of claim 1, wherein the shape is characterized and compared using a geometric arrangement of the plurality of image fixation points.

8. The system of claim 1, wherein the shape is characterized and compared using statistical features of the plurality of image fixation points.

9. The system of claim 1, wherein the shape is characterized and compared using a spatio-temporal arrangement of the plurality of image fixation points.

10. The system of claim 1, wherein the identification of the object is performed by comparing the shape formed by the plurality of image fixation points to the shape formed by the plurality of the model fixation points in a database.

11. The system of claim 1, wherein the identification of the object is performed using at least one of a rule-based recognition method or a machine learning based recognition method.

12. The system of claim 1, wherein the object is inanimate.

13. The system of claim 1, wherein the object is animate.

40. The system of claim 1, wherein the hardware processor further identifies features of the plurality of fixation points.

43. The system of claim 1, wherein the image is optically detected data.

14. A method for optical recognition, comprising:

computing a saliency map of an image;

identifying a plurality of image fixation points in the image based on the saliency map using a hardware processor; and

identifying an object represented in the image by comparing a shape formed by the plurality of image fixation points to a shape formed by a plurality of model fixation points in a model of a known object using the hardware processor.

15. The method of claim 14, wherein the image is a video frame.

16. The method of claim 14, wherein the plurality of image fixation points are identified using a visual attention model.

17. The method of claim 14, wherein the plurality of image fixation points are identified by finding salient features of the image.

18. The method of claim 14, wherein the plurality of image fixation points are identified by simulating a foveation process.

19. The method of claim 14, wherein the plurality of image fixation points are identified by blurring at least a portion of the image.

20. The method of claim 14, wherein the shape is characterized and compared using a geometric arrangement of the plurality of image fixation points.

21. The method of claim 14, wherein the shape is characterized and compared using statistical features of the plurality of image fixation points.

22. The method of claim 14, wherein the shape is characterized and compared using a spatio-temporal arrangement of the plurality of image fixation points.

23. The method of claim 14, wherein the identification of the object is performed by comparing the shape formed by the plurality of image fixation points to the shape formed by the plurality of the model fixation points in a database.

24. The method of claim 14, wherein the identification of the object is performed using at least one of a rule-based recognition method or a machine learning based recognition method.

25. The method of claim 14, wherein the object is inanimate.

26. The method of claim 14, wherein the object is animate.

41. The method of claim 14, wherein the method further comprises identifying features of the plurality of fixation points.

44. The method of claim 14, wherein the image is optically detected data.

27. A non-transitory computer-readable medium containing computer-executable instructions that, when executed by a hardware processor, cause the processor to perform a method for optical recognition, the method comprising:

computing a saliency map of an image;

identifying a plurality of image fixation points in the image based on the saliency map; and

identifying an object represented in the image by comparing a shape formed by the plurality of image fixation points to a shape formed by a plurality of model fixation points in a model of a known object.

28. The non-transitory computer-readable medium of claim 27, wherein the image is a video frame.

29. The non-transitory computer-readable medium of claim 27, wherein the plurality of image fixation points are identified using a visual attention model.

30. The non-transitory computer-readable medium of claim 27, wherein the plurality of image fixation points are identified by finding salient features of the image.

31. The non-transitory computer-readable medium of claim 27, wherein the plurality of image fixation points are identified by simulating a foveation process.

32. The non-transitory computer-readable medium of claim 27, wherein the plurality of image fixation points are identified by blurring at least a portion of the image.

33. The non-transitory computer-readable medium of claim 27, wherein the shape is characterized and compared using a geometric arrangement of the plurality of image fixation points.

34. The non-transitory computer-readable medium of claim 27, wherein the shape is characterized and compared using statistical features of the plurality of image fixation points.

35. The non-transitory computer-readable medium of claim 27, wherein the shape is characterized and compared using a spatio-temporal arrangement of the image plurality of fixation points.

36. The non-transitory computer-readable medium of claim 27, wherein the identification of the object is performed by comparing the shape formed by the plurality of image fixation points to the shape formed by the plurality of the model fixation points in a database.

37. The non-transitory computer-readable medium of claim 27, wherein the identification of the object is performed using at least one of a rule-based recognition method or a machine learning based recognition method.

38. The non-transitory computer-readable medium of claim 27, wherein the object is inanimate.

39. The non-transitory computer-readable medium of claim 27, wherein the object is animate.

42. The non-transitory computer-readable medium of claim 27, wherein the method further comprises identifying features of the plurality of fixation points.

45. The non-transitory computer-readable medium of claim 27, wherein the image is optically detected data.