Adjustment of sensor geometry model parameters using digital imagery co-registration process to reduce errors in digital imagery geolocation data

• US 5,995,681 A
• Filed: 06/03/1997
• Issued: 11/30/1999
• Est. Priority Date: 06/03/1997
• Status: Expired due to Term

• Current Assignee: Harris Corporation (L3Harris Technologies, Inc.)
• Original Assignee: Harris Corporation (L3Harris Technologies, Inc.)
• Inventors: Bell, David M., Needham, Jack M., Lee, Andrew J.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): PATEL, KANJIBHAI B
• Application Number:
  US08/885,371
• Time in Patent Office: 910 Days
• Field of Search: 382/293, 382/294, 382/295, 382/291, 382/109, 356/3.02, 356/3.06, 356/3.07, 356/3.08, 356/138, 356/139, 356/139.03, 356/139.04, 356/139.05, 356/139.07, 356/139.08, 701/2, 701/3, 701/4, 701/223
• US Class Current: 382/293
• CPC Class Codes: G01C 11/02 Picture taking arrangements specially adapted for photogrammetry or photographic surveying, e.g. controlling overlapping of pictures
• CPC Class Codes: G06T 3/12 Panospheric to cylindrical image transformations
• CPC Class Codes: G06T 7/32 using correlation-based methods

Abstract

A digital image processing system reduces errors in the parameters of a sensor geometry model, through which points in a captured digital image are geolocated to the surface of the earth by means of a `real time` co-registration mechanism that refines the geometry model associated with the working image in a matter of seconds. Using a co-registration mechanism such as that described in the U.S. Pat. No. 5,550,937, the system co-registers the reduced accuracy working digital image with a reference image, geographical spatial locations of respective pixels of which have been previously determined with a high degree of accuracy. The imagery co-registration operator adjusts the respective geometry models associated with its input images, in accordance with differences in cross-correlations of the respectively different spatial resolution versions of the two images, so as to bring the respective images into effective co-registration on image registration surface. Mutual registration on the image registration surface of the working and reference images reduces the parameter errors in the original working image'"'"'s sensor geometry model to the same error resolution of the reference image'"'"'s geometry model.

Claims

1. For use with a digital imagery processing system, in which a first digital image of an object is derived from a first image capture system, and wherein the position of a respective pixel of said first digital image, in terms of a spatial coordinate system for said object, is definable to a first spatial accuracy, a method of improving the spatial accuracy of the location of the position of said respective pixel of said first digital image, said method comprising the steps of:

(a) providing a first geometry model associated with said first image capture system that projects said first digital image onto a registration surface;

(b) providing a second digital image of said object that has been derived from a second image capture system, locations of respective pixels of which, in terms of said spatial coordinate system, are known to a second spatial accuracy, said second image capture system having a second geometry model that projects said second digital image onto said registration surface; and

(c) modifying parameters of said first geometry model by co-registering said first digital image with said second digital image on said registration surface in accordance with said first and second respective geometry models therefor, thereby causing locations of respective pixels of said first image, in terms of said spatial coordinate system, to be known to said second spatial accuracy.

2. A method according to claim 1, wherein said first and second digital images comprise first and second respective images of an area of the surface of the earth as captured by said first and second image capture systems, and wherein the positions of respective pixels of said first and second digital images are defined in terms of latitude, longitude and elevation.

3. A method according to claim 1, wherein step (c) comprises iteratively cross-correlating content-emphasized neighborhoods of pixels of respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, and adjusting said first and second respective geometry models in accordance with differences in cross-correlations of said respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, so as to bring said first and second digital images into effective co-registration on said registration surface.

4. A method according to claim 1, wherein said first and second image capture systems comprise diverse image capture systems.

5. A method according to claim 1, wherein step (b) comprises selecting said second digital image of said object, from a library of digital images containing said object, based upon contents of said first digital image of said object.

6. A method according to claim 5, wherein step (b) comprises selecting said second digital image of said object, from a library of digital images containing said object, based upon spatial coordinate information of said first digital image of said object.

7. For use with a digital imagery processing system, in which a first digital image of an area of the surface of the earth is captured by a first image capture system, and wherein spatial coordinates on the surface of the earth of a respective pixel of said first digital image are definable to a first degree of error, a method of reducing said first degree of error and thereby improving the accuracy of values of spatial coordinates of said respective pixel of said first digital image, said first image capture system a first geometry model associated therewith that projects said first digital image onto a registration surface, said method comprising the steps of:

(a) providing a second digital image of said area of the surface of the earth that has been derived from a second image capture system, locations of respective pixels of said second digital image having spatial coordinates on the surface of the earth to a second degree of error, less than said first degree of error, said second image capture system having a second geometry model that projects said second digital image onto said registration surface; and

(b) modifying parameters of said first geometry model by co-registering said first digital image with said second digital image on said registration surface in accordance with said first and second respective geometry models therefor, thereby reducing said first degree of error of spatial coordinates of respective pixels of said first image to said second degree of error.

8. A method according to claim 7, wherein step (b) comprises iteratively cross-correlating content-emphasized neighborhoods of pixels of respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, and adjusting said first and second respective geometry models in accordance with differences in cross-correlations of said respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, so as to bring said first and second digital images into effective co-registration on said registration surface.

9. A method according to claim 7, wherein spatial coordinates of pixels of said first and second digital images of said area of the earth are defined in terms of latitude, longitude and elevation.

10. A method according to claim 7, wherein said first and second image capture systems comprise diverse image capture systems.

11. A method according to claim 7, wherein step (a) comprises selecting said second digital image, from a library of digital images containing said area of the surface of the earth, based upon contents of said first digital image.

12. A method according to claim 11, wherein step (a) comprises selecting said second digital image, from a library of digital images containing said area of the surface of the earth, based upon spatial coordinate information of said first digital image.

13. For use with a digital imagery processing system, in which a first digital image of an area of the surface of the earth is captured by a first image capture system having a first, associated geometry model through which geographical coordinates on the surface of the earth of a respective pixel of said first digital image are definable to a first degree of error, a digital image processing arrangement for reducing said first degree of error and thereby improving the accuracy of values of geographical coordinates of said respective pixel of said first digital image comprising:

a second digital image of said area of the surface of the earth that has been derived from a second image capture system, locations of respective pixels of said second digital image having spatial coordinates on the surface of the earth to a second degree of error, less than said first degree of error, and a second geometry model through which said respective pixels of said second digital image are projectable onto a registration surface; and

a digital image processor which is coupled to receive first data representative of said first digital image and said first, associated geometry model, and second data representative of said second digital image and said second geometry model, and being operative to modify co-register said first digital image with said second digital image on said registration surface in accordance with said first and second respective geometry models therefor, thereby reducing said first degree of error of spatial coordinates of respective pixels of said first image to said second degree of error.

14. A digital image processing arrangement according to claim 13, wherein said digital image processor is operative to iteratively cross-correlate content-emphasized neighborhoods of pixels of respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, and to adjust said first and second respective geometry models in accordance with differences in cross-correlations of said respectively different spatial resolution versions of said first and second digital images as projected onto said registration surface, so as to bring said first and second digital images into effective co-registration on said registration surface.

15. A digital image processing arrangement according to claim 13, wherein spatial coordinates of pixels of said first and second digital images of said area of the earth are defined in terms of latitude, longitude and elevation.

16. A digital image processing arrangement according to claim 13, wherein said first and second image capture systems comprise diverse image capture systems.

17. A digital image processing system according to claim 13, wherein said second digital image is one that has been selected from a library of digital images containing said area of the surface of the earth, based upon contents of said first digital image.

18. A digital image processing system according to claim 17, wherein said second digital image is one that has been selected from a library of digital images containing said area of the surface of the earth, based upon spatial coordinate information of said first digital image.