Automated geospatial image mosaic generation
View Patent ↗Automatic generation of a mosaic comprising a plurality of geospatial images. An embodiment of the automatic mosaic generation may include automated source image selection that includes comparison of source images to base layer image to determine radiometric similar source images. Additionally, an embodiment of an automatic cutline generator may be provided to automatically determine a cutline when merging two images such that radiometric differences between the images along the cutline are reduced. In this regard, less perceivable outlines may be provided. Further still, an embodiment of a radiometric normalization module may be provided that may determine radiometric adjustments to source images to match certain properties of the base layer image. In some embodiments, when processing source images, the source images may be downsampled during a portion of the processing to reduce computational overhead. Additionally, some highly parallel computations may be performed by a GPU to further enhance performance.
1. A method for radiometric normalization of target images relative to a base image, comprising:
downsampling a first target image to produce a first low resolution target image;
calculating first target image metadata for the first low resolution target image and first base image metadata for a first portion of the base image corresponding to the first target image;
determining a first normalization function based on a comparison of the first target image metadata for the first low resolution target image and the first base image metadata for the first portion of the base image;
applying the first normalization function to the first target image to normalize the first target image relative to the base image;
downsampling a second target image to produce a second low resolution target image;
calculating second target image metadata for the second low resolution target image and second base image metadata for a second portion of the base image corresponding to the second target image;
determining a second normalization function based on a comparison of the second target image metadata for the second low resolution target image and the second base image metadata for the second portion of the base image; and
applying the second normalization function to the second target image to normalize the second target image relative to the base image;
wherein the first target image and the second target image are geospatial source images and the first portion of the base image and the second portion of the base image comprise different respective portions of a geospatial base layer image that are geographically concurrent with the respective first and second target images, and the base image comprises a global color balanced mosaic image.
2. The method of claim 1 , wherein the image metadata is independent of the spatial resolution of the image.
3. The method of claim 2 , wherein the image metadata comprises histogram data.
4. The method of claim 3 , wherein the calculating comprises calculation of a cumulative distribution function (CDF) for the low resolution target image and the base image.
5. The method of claim 4 , wherein the determining comprises comparing the CDF for the low resolution target image to the CDF for the base image to determine the normalization function.
6. The method of claim 5 , wherein the normalization function is tabulated into a lookup table.
7. The method of claim 3 , further comprising:
determining at least one anomalous pixel; and
masking at least one of the low resolution target image or the base image prior to remove the at least one anomalous pixel from the calculating.
8. The method of claim 7 , wherein the at least one anomalous pixel removed from the calculating comprises a pixel having a saturated pixel value.
9. The method of claim 8 , wherein the saturated pixel value removed from the calculating by way of the masking is identified as a pixel corresponding to water or a cloud.
10. The method of claim 1 , further comprising:
performing spectral flattening on at least one of the low resolution target image or the base image prior to the calculating.
11. The method of claim 10 , wherein the spectral flattening includes calculating an average intensity for a plurality of spectral channels of the image.
12. A system for radiometric normalization, comprising:
a source image database including at least a first and a second target image comprising geospatial source images of different geographic areas;
a base layer database including at least one base image comprising a global color balanced mosaic image having corresponding portions geographically concurrent with respective ones of the first target image and the second target image;
a downsampling module, executed by a processor of the system, for downsampling the first and second target images to produce corresponding first and second low resolution target images;
a radiometric normalization module, executed by a processor of the system, operable to normalize the first target image from the source image database relative to the at least one base image based on a normalization function that is at least partially based on a comparison of image metadata from the first low resolution target image and the geographically concurrent portion of the base layer image corresponding to the first target image and normalize the second target image from the source image database relative to the at least one base image based on a normalization function that is at least partially based on a comparison of image metadata from the second low resolution target image and the geographically concurrent portion of the base layer image corresponding to the second target image.
13. The system of claim 12 , wherein the image metadata is independent of the spatial resolution of the image.
14. The system of claim 13 , wherein the image metadata comprises histogram data.
15. The system of claim 14 , wherein the histogram data comprises a cumulative distribution function (CDF) for the low resolution target image and the base image.
16. The system of claim 12 , wherein the radiometric normalization module is operable to compare the CDF for the low resolution target image to the CDF for the base image to determine the normalization function.
17. The system of claim 16 , wherein the radiometric normalization module is operable to tabulate a lookup table based on the normalization function.
18. The system of claim 12 , wherein at least one of the target image or the base layer image comprises at least one anomalous pixel, and wherein the normalization module is operable to generate a pixel mask for masking at least one of the low resolution target image or the base image prior to remove the at least one anomalous pixel from the image metadata.
19. The system of claim 18 , wherein the at least one anomalous comprises a pixel having a saturated pixel value.
20. The system of claim 19 , wherein the saturated pixel value removed from the calculating by way of the masking is identified as a pixel corresponding to water or a cloud.
21. The system of claim 12 , wherein the radiometric normalization module is operable to perform spectral flattening on at least one of the low resolution target image or the base image, wherein the corresponding image metadata is determined using the spectrally flattened image data.
22. The system of claim 21 , wherein the radiometric normalization module is operable to calculating an average intensity for a plurality of spectral channels of the image for the spectral flattening.