Granted Patent B2

US 7,991,226 · App. 11/871,740 · Granted Aug 2, 2011

Abstract

Claims (78)

1. An automated process for color-balancing a series of oblique images captured from one or more positions and from one or more orientations, comprising the steps of:

using a computer to perform the steps of:

selecting a series of color-balanced images to use as reference images;

on an oblique image by image basis performing the following steps:

locating a portion(s) of a reference image(s) that overlaps the oblique image;

creating multiple color-balancing transformations that approximately match a color distribution of the oblique image to a color distribution of the overlapping portion(s) of the reference image(s);

transforming pixels in the oblique image according to more than one of the multiple color-balancing transformations created for the oblique image; and

storing the transformed pixel value in the oblique image or a copy of the oblique image.

2. An automated process for color-balancing a series of oblique images captured from one or more positions and from one or more orientations, comprising the steps of:

using a computer to perform the steps of:

selecting a series of color-balanced nadir images to use as reference images; on an oblique image by image basis performing the following steps:

locating a portion(s) of a reference image(s) that overlaps the oblique image;

creating multiple color-balancing transformations that approximately match a color distribution of the oblique image to a color distribution of the overlapping portion(s) of the reference image(s); and

transforming pixels in the oblique image according to more than one of the color-balancing transformations created for that oblique image.

3. An automated process for color-balancing a series of oblique images captured from one or more positions and from one or more orientations, comprising the steps of:

using a computer to perform the steps of:

a. selecting a series of color-balanced images to use as reference images;

on an oblique image by oblique image basis, performing the steps of:

b. dividing the oblique image into a plurality of oblique image sections;

on a section by section basis performing the steps of:

c. locating a portion(s) of a reference image(s) that overlaps the oblique image section; and

d. creating a color-balancing transformation that approximately matches a color distribution of the oblique image section to a color distribution of the overlapping reference portion(s);

on a pixel by pixel basis for the oblique image, performing the steps of:

e. selecting the oblique image section(s) that apply to the pixel;

f. transforming the pixel by the color balancing transformation for that selected oblique image section yielding a transformed pixel value for each selected oblique image section;

g. blending the transformed pixel values into a single resulting pixel value; and

h. storing the resulting pixel value in the oblique image or a copy of the oblique image.

4. The automated process of claim 3 , wherein steps d, e, f, g and h are performed independently for multiple different pixel color values.

5. The automated process of claim 4 , wherein step g. is defined further as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

6. The automated process of claim 3 , wherein after performing steps a-d, the process further comprises the step of d1. storing the section color balancing transformations so that they can be applied at a later time.

7. The automated process of claim 6 , wherein steps a-d and d1 are performed at one time, and steps e-h are performed at a later time using the section color balancing transformations stored in step d1.

8. The automated process of claim 3 , wherein step g. is defined further as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

9. The automated process of claim 3 , wherein the oblique images are aerial oblique images and the reference images are aerial nadir images.

10. The automated process of claim 9 , wherein steps d, e, f, g and h are performed independently for multiple different pixel color values.

11. The automated process of claim 10 , wherein step g. is defined further as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

12. The automated process of claim 9 , wherein after performing steps a-d, the process further comprises the step of d1. storing the color balancing transformations so that they can be applied at a later time.

13. The automated process of claim 12 , wherein steps a-d and d1 are performed at one time, and steps e-h are performed at a later time using the color balancing transformations stored in step d1.

14. The automated process of claim 9 , wherein step g. is defined further as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

15. An image processing system, comprising:

a computer running image processing software to:

select a series of color-balanced images to use as reference images;

on an oblique image by image basis performing the following steps:

locate a portion(s) of a reference image(s) that overlaps the oblique image;

create multiple color-balancing transformations that approximately match a color distribution of the oblique image to a color distribution of overlapping portion(s) of the reference image(s);

transform pixels in the oblique image according to more than one of the multiple color-balancing transformations created for the oblique image; and

store the transformed pixel value in the oblique image or a copy of the oblique image.

16. An image processing system, comprising:

a computer running image processing software to:

select a series of color-balanced nadir images to use as reference images;

on an oblique image by image basis performing the following steps:

locate a portion(s) of a reference image(s) that overlaps the oblique image;

create multiple color-balancing transformations that approximately match a color distribution of the oblique image to a color distribution of overlapping portion(s) of the reference image(s); and

transform pixels in the oblique image according to more than one of the color-balancing transformations created for that oblique image.

17. An image processing system, comprising:

a computer running image processing software to:

a. select a series of color-balanced images to use as reference images;

on an oblique image by oblique image basis, performing the steps of:

b. divide the oblique image into a plurality of oblique image sections;

on a section by section basis performing the steps of:

c. locate a portion(s) of a reference image(s) that overlaps the oblique image section; and

d. create a color-balancing transformation that approximately matches a color distribution of the oblique image section to a color distribution of the overlapping reference portion(s);

on a pixel by pixel basis for the oblique image, performing the steps of:

e. select the oblique image section(s) that apply to the pixel;

f. transform the pixel by the color balancing transformation for that selected oblique image section yielding a transformed pixel value for each selected oblique image section;

g. blend the transformed pixel values into a single resulting pixel value; and

h. store the resulting pixel value in the oblique image or a copy of the oblique image.

18. The image processing system of claim 17 , wherein the computer performs steps d, e, f, g, and h independently for multiple different pixel color values.

19. The image processing system of claim 18 , wherein the computer performs step g. which is further defined as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

20. The image processing system of claim 17 , wherein, after the computer performs steps a-d, the computer performs the step of d1: storing the color balancing transformations for the sections so that they can be applied at a later time.

21. The image processing system of claim 20 , wherein the computer performs steps a-d and d1 at one time, and performs steps e-h at a later time using the color balancing transformations stored in step d1.

22. The image processing system of claim 17 , wherein the computer performs step g. which is defined as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

23. The image processing system of claim 17 , wherein the oblique images are aerial oblique images and the reference images are aerial nadir images.

24. The image processing system of claim 23 , wherein the computer performs steps d, e, f, g, and h independently for multiple different pixel color values.

25. The image processing system of claim 24 , wherein the computer performs step g. which is further defined as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

26. The image processing system of claim 23 , wherein the computer, after performing steps a-d,

d1. storing the color balancing transformations so that they can be applied at a later time.

27. The image processing system of claim 26 , wherein the computer performs steps a-d and d1 at one time, and perform steps e-h at a later time using the color balancing transformations stored in step d1.

28. The image processing system of claim 23 , wherein the computer performs step g. which is defined as blending the transformed pixel values into a single resulting pixel value using bi-linear interpolation.

Assignments (12)