SYSTEMS AND METHODS FOR ESTIMATING THE GEOGRAPHIC LOCATION AT WHICH IMAGE DATA WAS CAPTURED

• US 20140064624A1
• Filed: 05/12/2012
• Published: 03/06/2014
• Est. Priority Date: 05/11/2011
• Status: Active Grant

• Current Assignee: University of Florida Research Foundation, Incorporated (State University System of Florida)
• Original Assignee: University of Florida Research Foundation, Incorporated (State University System of Florida)
• Inventors: Dixon, Warren, Kim, Hakjae
• Granted Patent: US 9,501,699 B2
• Time in Patent Office: Days
• US Class Current: 382/201
• CPC Class Codes: G06T 7/80 Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
• CPC Class Codes: G06V 10/757 Matching configurations of points or features
• CPC Class Codes: G06V 20/13 Satellite images

Abstract

In some embodiments, a system and method for estimating the geographical location at which image data was captured with a camera identifies matching feature points between the captured images, estimates a pose of the camera during the image capture from the feature points, performs geometric reconstruction of a scene in the images using the estimated pose of the camera to obtain a reconstructed scene, and compares the reconstructed scene to overhead images of known geographical origin to identify potential matches

Claims

1. A method for estimating the geographical location at which image data was captured with a camera, the method comprising:

identifying matching feature points between the captured images;

estimating a pose of the camera during the image capture from the feature points;

performing geometric reconstruction of a scene in the images using the estimated pose of the camera to obtain a reconstructed scene; and

comparing the reconstructed scene to overhead images of known geographical origin to identify potential matches.

2. The method of claim 1, wherein identifying matching feature points comprises identifying matching feature points between ground-level images.

3. The method of claim 1, wherein identifying matching feature points comprises identifying matching feature points between the captured images using scale invariant feature transformation (SIFT).

4. The method of claim 1, wherein estimating a pose of the camera comprises estimating rotation and translation of the camera.

5. The method of claim 1, wherein estimating a power of the camera comprises performing pose estimation by gridding unit sphere (PEGUS).

6. The method of claim 1, wherein performing geometric reconstruction comprises performing geometric reconstruction using both the estimated pose of the camera and a known length in the scene.

7. The method of claim 1, wherein performing geometric reconstruction comprises performing geometric reconstruction using both the estimated pose of the camera and a known velocity of the camera.

8. The method of claim 1, wherein performing geometric reconstruction converts the a ground-level perspective of the images to an overhead perspective.

9. The method of claim 1, wherein the reconstructed scene identifies geometries of objects in the scene and white spaces between the camera and the objects that are substantially free of vertically-oriented objects.

10. The method of claim 1, wherein comparing the reconstructed scene to overhead images comprises comparing the reconstructed scene to satellite images.

11. The method of claim 1, wherein comparing the reconstructed scene to overhead images comprises comparing the reconstructed scene to occupancy-grid maps (OGMs) generated from satellite images.

12. The method of claim 1, further comprising categorizing white spaces between the camera and the objects that are substantially free of vertically-oriented objects in both the reconstructed scene and the overhead images to increase the speed with which potential matches are identified.

13. A system for estimating the geographical location at which image data was captured with a camera, the system comprising:

a processing device; and

memory storing a geolocation estimator comprising logic configured to;

identify matching feature points between the captured images,

estimate a pose of the camera during the image capture from the feature points,

perform geometric reconstruction of a scene in the images using the estimated pose of the camera to obtain a reconstructed scene, and

compare the reconstructed scene to overhead images of known geographical origin to identify potential matches.

14. The system of claim 13, wherein the logic configured to identify matching feature points comprises logic configured to identify matching feature points between the captured images using scale invariant feature transformation (SIFT).

15. The system of claim 13, wherein the logic configured to estimate a pose of the camera comprises logic configured to estimate rotation and translation of the camera by performing pose estimation by gridding unit sphere (PEGUS).

16. The system of claim 13, wherein the logic configured to perform geometric reconstruction comprises logic configured to perform geometric reconstruction using both the estimated pose of the camera and a known length in the scene.

17. The system of claim 13, wherein the logic configured to perform geometric reconstruction comprises logic configured to perform geometric reconstruction using both the estimated pose of the camera and a known velocity of the camera.

18. The system of claim 13, wherein the logic configured to perform geometric reconstruction converts the a ground-level perspective of the images to an overhead perspective.

19. The system of claim 13, wherein the reconstructed scene identifies geometries of objects in the scene and white spaces between the camera and the objects that are substantially free of vertically-oriented objects.

20. The system of claim 13, wherein the logic configured to compare the reconstructed scene to overhead images comprises logic configured to compare the reconstructed scene to occupancy-grid maps (OGMs) generated from satellite images.

21. The system of claim 13, further comprising logic configured to categorize white spaces between the camera and the objects that are substantially free of vertically-oriented objects in both the reconstructed scene and the overhead images to increase the speed with which potential matches are identified.

22. A non-transitory computer-readable medium that stores a geolocation estimator for estimating the geographical location at which image data was captured with a camera, the computer-readable medium comprising:

logic configured to identify matching feature points between the captured images;

logic configured to estimate a pose of the camera during the image capture from the feature points;

logic configured to perform geometric reconstruction of a scene in the images using the estimated pose of the camera to obtain a reconstructed scene; and

logic configured to compare the reconstructed scene to overhead images of known geographical origin to identify potential matches.

23. The computer-readable medium of claim 22, wherein the logic configured to identify matching feature points is configured to identify matching feature points between the captured images using scale invariant feature transformation (SIFT).

24. The computer-readable medium of claim 22, wherein the logic configured to estimate a pose of the camera is configured to estimate rotation and translation of the camera by performing pose estimation by gridding unit sphere (PEGUS).

25. The computer-readable medium of claim 22, wherein the logic configured to perform geometric reconstruction is configured to perform geometric reconstruction using both the estimated pose of the camera and a known length in the scene.

26. The computer-readable medium of claim 22, wherein the logic configured to perform geometric reconstruction is configured to perform geometric reconstruction using both the estimated pose of the camera and a known velocity of the camera.

27. The computer-readable medium of claim 22, wherein the logic configured to perform geometric reconstruction converts the a ground-level perspective of the images to an overhead perspective.

28. The computer-readable medium of claim 22, wherein the reconstructed scene identifies geometries of objects in the scene and white spaces between the camera and the objects that are substantially free of vertically-oriented objects.

29. The computer-readable medium of claim 22, wherein the logic configured to compare the reconstructed scene to overhead images is configured to compare the reconstructed scene to occupancy-grid maps (OGMs) generated from satellite images.

30. The computer-readable medium of claim 22, further comprising logic configured to categorize white spaces between the camera and the objects that are substantially free of vertically-oriented objects in both the reconstructed scene and the overhead images to increase the speed with which potential matches are identified.