Method and system for automatically locating static occlusions

• US 9,390,329 B2
• Filed: 04/25/2014
• Issued: 07/12/2016
• Est. Priority Date: 04/25/2014
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Shreve, Matthew Adam, Bernal, Edgar A, Li, Qun, Loce, Robert P.
• Primary Examiner(s): Lu, Tom Y
• Application Number:
  US14/262,456
• Publication Number:
  US 20150310274A1
• Time in Patent Office: 809 Days
• Field of Search: 382/103
• US Class Current: 1/1
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30236 Traffic on road, railway or crossing
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/77 using statistical methods

Abstract

This disclosure provides a method and system to locate/detect static occlusions associated with an image captured scene including a tracked object. According to an exemplary method, static occlusions are automatically located by monitoring the motion of single or multiple objects in a scene over time and with the use of an associated accumulator array.

Claims

1. A computer-implemented method of automatically identifying pixels corresponding to an occlusion in a scene captured with an image capturing device, where the occlusion obstructs a view of an object, in whole or part, being tracked within the scene, the method comprising:

a) identifying one or more foreground objects within the scene from at least one sequence of image frames including digital representations of the scene;

b) updating an accumulator array at pixel locations of the identified one or more foreground objects within the plurality of image frames;

c) determining a trajectory of the identified one or more foreground objects;

d) detecting pixels in an area of the updated accumulator array not associated with the foreground objects and located along the trajectory; and

e) associating the detected pixels as belonging to an occlusion.

2. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein prior to step a) the accumulator is initialized such that the accumulator array is equivalent in size to the image frames.

3. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein the accumulator array is a foreground or motion pixel accumulator array, wherein each index in the accumulator array corresponds to a pixel location and the value of the accumulator array at the given index is the number of times each pixel was observed to be a foreground or motion pixel.

4. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein the occlusion is static, and the image capturing device is a fixed single image capturing device.

5. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein the image capturing device is one of a video camera, infrared camera, thermal camera, and satellite imaging camera.

6. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein there are two or more occlusions identified in the scene.

7. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein the occlusion is one of a tree, bush, lamp post, sign, fixture, booth, vehicle, image capturing device, or a blind spot.

8. The computer-implemented method of automatically identifying an occlusion according to claim 1, wherein steps c) and d) include creating an occlusion mask from the accumulator array.

9. The computer-implemented method of automatically identifying an occlusion according to claim 8, wherein connected component analysis (CCA) is applied to the occlusion mask on a row-by-row basis after image axes have been adjusted to align with a general motion direction of the objects in the scene.

10. The computer-implemented method of automatically identifying an occlusion according to claim 8, wherein step c) includes determining a vector to represent a dominant direction of the identified objects in motion.

11. The computer-implemented method of automatically identifying an occlusion according to claim 8, wherein step d) includes detecting geometric patterns in the accumulator array to identify the pixels corresponding to the occlusion.

12. An image processing system for automatically locating pixels corresponding to an occlusion in a scene, the system comprising:

a processor; and

,

a non-transitory computer readable memory storing instructions that are executable by the processor to;

a) receive a digital image of a scene captured with an image capturing device;

b) identify one or more foreground objects within the scene from at least one sequence of image frames including digital representations of the scene;

c) update an accumulator array with a location of the identified one or more foreground objects within the plurality of image frames;

d) determine a trajectory of the identified one or more foreground objects;

e) detect pixels in an area of the updated accumulator array not associated with the foreground objects and located along the trajectory; and

associate the detected pixels as belonging to an occlusion.

13. The image processing system according to claim 12, wherein prior to step b) the accumulator is initialized such that the accumulator array is equivalent in size to a size of the image frames.

14. The image processing system according to claim 12, wherein the accumulator array is a foreground pixel accumulator array, the identified one or more foreground objects are associated with foreground pixels of the image frames where foreground objects are identified, and the accumulator array is updated with foreground information.

15. The image processing system according to claim 12, wherein the occlusion is static, and the image capturing device is a fixed single image capturing device.

16. The image processing system according to claim 12, wherein the image capturing device is one of a video camera, infrared camera, thermal camera, and satellite imaging camera.

17. The image processing system according to claim 12, wherein there are two or more occlusions identified in the scene.

18. The image processing system according to claim 12, wherein the occlusion is one of a tree, bush, lamp post, sign, fixture, booth, vehicle, image capturing device blind spot, and a person.

19. The image processing system according to claim 12, wherein steps d) and e) include creating an occlusion mask from the accumulator array.

20. The image processing system according to claim 19, wherein connected component analysis (CCA) is applied to the occlusion mask on a row-by-row basis after image axes have been adjusted to align with a general motion direction of the objects in the scene.

21. The image processing system according to claim 19, wherein step d) includes determining a vector to represent a dominant direction of the identified objects in motion.

22. The image processing system according to claim 19, wherein step e) includes detecting geometric patterns in the accumulator array to identify the occlusion.

23. A computer program product comprising:

a non-transitory computer-usable data carrier storing instructions that, when executed by a computer, cause the computer to perform a method of automatically locating an occlusion in a scene captured with an image capturing device, the method comprising;

a) identifying one or more foreground objects within the scene from at least one sequence of image frames including digital representations of the scene;

b) updating an accumulator array with a location of the identified one or more foreground objects within the plurality of image frames;

c) determining a trajectory of the identified one or more foreground objects; and

d) identifying pixels in an area of the accumulator array not associated with foreground objects and along the trajectory and associating the pixels as corresponding to an occlusion.

24. The computer program product according to claim 23, wherein prior to step a) the accumulator is initialized such that the accumulator array is equivalent in size to a size of the image frames.

25. The computer program product according to claim 23, wherein steps c) and d) include creating an occlusion mask from the accumulator array.