Advanced cloud cover assessment for panchromatic images
View Patent ↗Cloud cover assessment system and method provides for automatically determining whether a target digital image acquired from remote sensing platforms is substantially cloud-free. The target image is acquired and compared to a corresponding known cloud-free image from a cloud-free database, using an optimized feature matching process. A feature matching statistic is computed between pixels in the target image and pixels in the cloud-free image and each value is converted to a feature matching probability. Features in the target image that match features in the cloud-free image exhibit a high value of feature matching probability, and are considered unlikely to be obscured by clouds, and may be designated for inclusion in the cloud-free database.
1. A computer-implemented method, comprising:
receiving a panchromatic image;
comparing the panchromatic image to a second image from a database of images, the second image having cloud-free regions;
matching a first feature from the panchromatic image to a second feature from the second image, wherein the second feature is located in one of the cloud-free regions of the second image; and
identifying a cloudy region in the panchromatic image based on the matching of the first feature to the second feature.
2. The method according to claim 1 , wherein receiving the panchromatic image comprises receiving an image from a single spectral channel covering a broad wavelength range including at least a majority of the visible spectrum.
3. The method according to claim 2 , wherein the broad wavelength range is about 400 nm to about 900 nm.
4. The method according to claim 1 , wherein the panchromatic image has a resolution of at least 250 m.
5. The method according to claim 1 , wherein the panchromatic image has a resolution of at least 15 m.
6. The method according to claim 1 , wherein the database of images comprises panchromatic images having a broad wavelength range including at least a majority of the visible spectrum.
7. The method according to claim 6 , wherein the broad wavelength range is about 400 nm to about 900 nm.
8. The method according to claim 1 , wherein the database of images comprises images having resolutions lower than the resolution of the panchromatic image.
9. The method according to claim 1 , wherein the database of images comprises images having resolutions of at least 250 m.
10. The method according to claim 1 , wherein the database of images comprises images having resolutions of at least 15 m.
11. The method according to claim 1 , further comprising automatically converting the panchromatic image to a top-of-atmosphere reflectance image.
12. The method according to claim 1 , wherein the second image comprises a panchromatic image.
13. The method according to claim 1 , wherein the identifying operation comprises:
automatically determining a cloud presence probability value;
calculating a confidence measure based on the cloud presence probability value; and
determining if the confidence measure falls within a predetermined range that indicates a cloudy region in the panchromatic image.
14. The method according to claim 1 , wherein the matching operation comprises using a normalized cross-correlation metric.
15. The method according to claim 1 , further comprising converting the panchromatic image and the second image to a substantially similar brightness scale before matching the first feature from the panchromatic image to the second feature from the second image.
16. The method according to claim 1 , wherein the first feature is a pixel in the panchromatic image and the second feature is a pixel in the second image.
17. A computer-implemented method, comprising:
receiving a panchromatic image;
identifying an area of interest in the image;
automatically creating a first data set from the area of interest;
receiving a second data set from a database, wherein the second data set includes data regarding one or more cloud-free regions of an image;
comparing the first data set to the second data set to create a third data set; and
identifying cloudy data points in the panchromatic image based on the third data set.
18. The method according to claim 17 , further comprising adjusting the brightness scale of the panchromatic image before identifying an area of interest.
19. The method according to claim 17 , wherein the comparing operation includes comparing a single pixel from the first data set to a single pixel in the second data set.
20. The method according to claim 17 , wherein receiving the panchromatic image comprises receiving an image from a single spectral channel covering a broad wavelength range including at least a majority of the visible spectrum.
21. The method according to claim 20 , wherein the broad wavelength range is about 400 nm to about 900 nm.
22. The method according to claim 17 , wherein the database of images comprises panchromatic images having a broad wavelength range including at least a majority of the visible spectrum.
23. The method according to claim 22 , wherein the broad wavelength range is about 400 nm to about 900 nm.
24. The method according to claim 17 , further comprising automatically converting the panchromatic image to a top-of-atmosphere reflectance image.
25. The method according to claim 17 , wherein the identifying operation comprises:
automatically determining a cloud presence probability value;
calculating a confidence measure based on the cloud presence probability value; and
determining if the confidence measure falls within a predetermined range that indicates a cloudy area in the panchromatic image.
26. The method according to claim 17 , wherein the comparing operation comprises using a normalized cross-correlation metric.
27. A computer-implemented method, comprising:
receiving a first panchromatic image having a first set of pixels, wherein the first panchromatic image has a resolution of at least 15 m;
receiving a second panchromatic image from a database having a second set of pixels, wherein the second panchromatic image has a resolution of at least 15 m, wherein the second set of pixels is located in a cloud-free region of the second panchromatic image;
cross correlating the first set of pixels with the second set of pixels to create a cross-correlated set of pixels; and
identifying a cloudy region in the first panchromatic image from the cross-correlated set of pixels.
28. The method according to claim 27 , wherein the identifying operation comprises:
automatically determining a cloud presence probability value for the cross-correlated set of pixels;
calculating a confidence measure based on the cloud presence probability value; and
determining if the confidence measure falls within a predetermined range that indicates a cloudy region in the first panchromatic image.
29. The method according to claim 27 , wherein the cross correlating operation comprises using a normalized cross-correlation metric.
30. The method according to claim 29 , wherein using the normalized cross-correlation metric comprises:
defining a search window in the first panchromatic image;
defining a reference window in the second panchromatic image;
computing a correlation value by calculating
C
(
m
,
n
)
=
(
R
(
m
,
n
)
-
μ
R
)
×
(
S
(
m
,
n
)
-
μ
S
)
(
R
(
m
,
n
)
-
μ
R
)
2
×
(
S
(
m
,
n
)
-
μ
S
)
2
where C (m, n) is the correlation value at location (m,n), R (m,n) is a reference window value at location (m,n), S (m,n) is a search window value at location (m,n), λ R is a mean of the reference window, and λ S is a mean of the search window, the correlation value being used to identify a cloudy region in the first panchromatic image.
31. The method according to claim 27 , wherein the first set of pixels and the second set of pixels are cross-correlated on a pixel-by-pixel basis.
32. The method according to claim 27 , further comprising automatically converting the panchromatic image to a top-of-atmosphere reflectance image.
33. The method according to claim 27 , further comprising converting the first panchromatic image and the second panchromatic image to a substantially similar brightness scale.