An automated ship detection technique includes accessing data associated with an image of a portion of Earth. The data includes reflectance values. A first portion of pixels within the image are masked with a cloud and land mask based on spectral flatness of the reflectance values associated with th
An automated ship detection technique includes accessing data associated with an image of a portion of Earth. The data includes reflectance values. A first portion of pixels within the image are masked with a cloud and land mask based on spectral flatness of the reflectance values associated with the pixels. A given pixel selected from the first portion of pixels is unmasked when a threshold number of localized pixels surrounding the given pixel are not masked by the cloud and land mask. A spatial variability image is generated based on spatial derivatives of the reflectance values of the pixels which remain unmasked by the cloud and land mask. The spatial variability image is thresholded to identify one or more regions within the image as possible ship detection regions.
대표청구항▼
1. A computer-readable storage medium that provides instructions that, if executed by a machine, will cause the machine to perform operations comprising: accessing data associated with an image of a portion of Earth, the data including reflectance values;masking a first portion of pixels within the
1. A computer-readable storage medium that provides instructions that, if executed by a machine, will cause the machine to perform operations comprising: accessing data associated with an image of a portion of Earth, the data including reflectance values;masking a first portion of pixels within the image with a cloud and land mask based on spectral flatness of the reflectance values associated with the pixels, wherein the reflectance values comprise a near infrared (NIR) reflectance value and a visible reflectance value for each of the pixels of the image and wherein the spectral flatness of the reflectance values is determined based on a ratio of the NIR reflectance value to the visible reflectance value for each of the pixels;unmasking a given pixel selected from the first portion of pixels when a threshold number of localized pixels surrounding the given pixel are not masked by the cloud and land mask;generating a spatial variability image based on spatial derivatives of the reflectance values of the pixels which remain unmasked by the cloud and land mask; andthresholding the spatial variability image to identify one or more regions within the image as possible ship detection regions. 2. The computer-readable storage medium of claim 1, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: rendering the image of the portion of Earth to a screen; andvisually marking the possible ship detection regions on the image. 3. The computer-readable storage medium of claim 1, wherein: the NIR reflectance value comprises a first reflectance value of light having a first wavelength between approximately 750 nm and approximately 950 nm, andthe visible reflectance value comprises a second reflectance value of light having a second wavelength between approximately 580 nm and approximately 700 nm. 4. The computer-readable storage medium of claim 1, wherein the spatial derivative is calculated based on the NIR reflectance values without using the visible reflectance values. 5. The computer-readable storage medium of claim 1, wherein unmasking the given pixel selected from the first portion of pixels when the threshold number of localized pixels surrounding the given pixel are not masked by the cloud and land mask, further comprises: analyzing a plurality of linear pixel arrays extending out from the given pixel in different directions on the image to a threshold distance; andif a threshold number of the linear pixel arrays include at least one pixel not masked by the cloud and land mask, then unmasking the given pixel. 6. The computer-readable storage medium of claim 5, wherein the plurality of linear pixel arrays extending out from the given pixel in the different directions on the image comprises four pixel arrays extending out vertically or horizontally from the given pixel and four pixels arrays extending out diagonally from the given pixel. 7. The computer-readable storage medium of claim 1, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: extending boundaries associated with the cloud and land mask by: grouping the pixels into first pixel groups; andmasking all pixels within a given first pixel group with the cloud and land mask, if any of the pixels within the given first pixel group were previously masked by the cloud and land mask. 8. The computer-readable storage medium of claim 7, wherein extending the boundaries associated with the cloud and land mask further comprises: re-grouping the pixels into second pixel groups offset from the first pixel groups; andmasking all pixels within a given second pixel group with the cloud and land mask, if any of the pixels within the given second pixel group were previously masked by the cloud and land mask. 9. The computer-readable storage medium of claim 1, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: preventing single or double pixel clusters that otherwise would be deemed to be a possible ship detection from being designated with the possible ship detection. 10. The computer-readable storage medium of claim 1, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: determining whether to execute the instructions associated with the masking, the unmasking, the generating, and the thresholding based upon whether the portion of the Earth is determined to be during night at a time of capturing the image, whether the portion of the Earth is determined to be within a glint region, or whether the portion of the Earth is determined to be land. 11. A computer automated method for detecting ships in water, the method comprising: accessing data associated with an image of a portion of Earth, the data including reflectance values, wherein the reflectance values comprise a near infrared (NIR) reflectance value and a visible reflectance value for each of the pixels of the image;categorizing pixels within the image into first or second categories based on spectral flatness of the reflectance values associated with the pixels, wherein categorizing the pixels within the image into the first or second categories comprises categorizing each of the pixels based on a ratio of the NIR reflectance value to the visible reflectance value for each of the pixels;re-categorizing a given pixel into the second category when a threshold number of localized pixels surrounding the given pixel, which was previously categorized into the first category, are categorized into the second category;calculating spatial derivatives on the reflectance values of the pixels categorized into the second category; andmarking a region of the image as a possible ship detection when the spatial derivatives associated with the pixels in the region have an absolute value greater than a threshold value. 12. The computer automated method of claim 11, wherein marking the region as a possible ship detection comprises: rendering the image of the portion of Earth to a screen; andvisually marking the region on the image. 13. The computer automated method of claim 11, wherein the NIR reflectance value comprises a first reflectance value of light having a first wavelength between approximately 750 nm and approximately 950 nm, andwherein the visible reflectance value comprises a second reflectance value of light having a second wavelength between approximately 580 nm and approximately 700 nm. 14. The computer automated method of claim 11, wherein the spatial derivatives are calculated based on the NIR reflectance values without using the visible reflectance values to calculate the spatial derivatives. 15. The computer automated method of claim 11, wherein the first category comprises pixels deemed to be images of clouds or land and the second category comprises pixels deemed to be images of non-clouds and non-land. 16. The computer automated method of claim 11, wherein the threshold value comprises a first threshold value, the method further comprising: marking a given region as a possible ship detection when the spatial derivatives associated with the pixels in the given region of the image have a value greater than the first threshold value; andmarking the given region as a possible ship detection when the spatial derivatives associated with the pixels in the given region of the image have a value less than a second threshold value. 17. The computer automated method of claim 11 wherein re-categorizing the given pixel into the second category, further comprises: analyzing a plurality of linear pixel arrays extending out from the given pixel in different directions on the image to a threshold distance; andif a threshold number of the linear pixel arrays includes at least one pixel categorized into the second category, then re-categorizing the given pixel into the second category. 18. The computer automated method of claim 17, wherein the plurality of linear pixel arrays extending out from the given pixel in the different directions on the image comprises four pixel arrays extending out vertically or horizontally from the given pixel and four pixels arrays extending out diagonally from the given pixel. 19. The computer automated method of claim 11, further comprising extending boundaries associated with the pixels categorized into the first category by: grouping the pixels into first pixel groups; andif any of the pixels within a given first pixel group is categorized into the first category, then categorizing all pixels within the given first pixel group into the first category. 20. The computer automated method of claim 19, wherein extending the boundaries associated with the pixels categorized into the first category further comprises: re-grouping the pixels into second pixel groups offset from the first pixel groups; andif any of the pixels within a given second pixel group is categorized into the first category, then categorizing all pixels within the given second pixel group into the first category. 21. The computer automated method of claim 11, further comprising: preventing single or double pixel clusters that otherwise would be deemed to be a possible ship detection from being designated with the possible ship detection. 22. The computer automated method of claim 11, further comprising determining whether to employ the computer automated method for detecting ships based upon whether the portion of the Earth is determined to be during night at a time of capturing the image, whether the portion of the Earth is determined to be within a glint region, or whether the portion of the Earth is determined to be land. 23. A computer-readable storage medium that provides instructions that, if executed by a machine, will cause the machine to perform operations comprising: accessing data associated with an image of a portion of Earth, the data including reflectance values, wherein the reflectance values comprise a near infrared (NIR) reflectance value and a visible reflectance value for each of the pixels of the image;categorizing pixels within the image into first or second categories based on spectral flatness of the reflectance values associated with the pixels, wherein categorizing the pixels within the image into the first or second categories comprises categorizing each of the pixels based on a ratio of the NIR reflectance value to the visible reflectance value for each of the pixels;re-categorizing a given pixel into the second category when a threshold number of localized pixels surrounding the given pixel, which was previously categorized into the first category, are categorized into the second category;calculating spatial derivatives on the reflectance values of the pixels categorized into the second category; andmarking a region of the image as a possible ship detection when the spatial derivatives associated with the pixels in the region have an absolute value greater than a threshold value. 24. The computer-readable storage medium of claim 23, wherein marking the region as the possible ship detection comprises: rendering the image of the portion of Earth to a screen; andvisually marking the region on the image. 25. The computer-readable storage medium of claim 23, wherein the threshold value comprises a first threshold value, the computer-readable storage medium further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: marking a given region of the image as the possible ship detection when the spatial derivatives associated with the pixels in the given region have a value greater than the first threshold value; andmarking the given region of the image as the possible ship detection when the spatial derivatives associated with the pixels in the given region have a value less than a second threshold value. 26. The computer-readable storage medium of claim 23 wherein re-categorizing the given pixel into the second category, comprises: analyzing a plurality of linear pixel arrays extending out from the given pixel in different directions on the image to a threshold distance; andif a threshold number of the linear pixel arrays includes at least one pixel categorized into the second category, then re-categorizing the given pixel into the second category. 27. The computer-readable storage medium of claim 23, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: extending boundaries associated with the pixels categorized into the first category by: grouping the pixels into first pixel groups; andif any of the pixels within a given first pixel group is categorized into the first category, then categorizing all pixels within the given first pixel group into the first category. 28. The computer-readable storage medium of claim 23, further providing instructions that, if executed by the computer, will cause the computer to perform further operations, comprising: preventing single or double pixel clusters that otherwise would be deemed to be a possible ship detection from being designated with the possible ship detection.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (8)
Snider, Robin Terry; McGee, Jeffrey Dykes; Perry, Michael Dale, Apparatus and methods for use in flash detection.
Novick, Arnold W.; Janik, Michael F.; McHale, Thomas J.; Rozenfeld, Ilya; McPhillips, Kenneth J.; Short, John R.; Casalegno, James W., Autonomous sonar system and method.
Angell, Robert Lee; Cosby, David Wayne; Friedlander, Robert R.; Kraemer, James R., Measuring a cohort'S velocity, acceleration and direction using digital video.
Arambel, Pablo; Newman, Arthur M.; Martinez, Mario, Method and system for processing a sequence of images to identify, track, and/or target an object on a body of water.
Kraft, Adam Wiggen; Babenko, Boris Aleksandrovich; Avtanski, Alexander Bogdanov; Sammons, Daniel Michael; Lin, Jasper; Lohn, Jason D., Remote determination of containers in geographical region.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.