Methods for object-based identification, sorting and ranking of target detections and apparatuses thereof

• US 8,897,489 B2
• Filed: 01/28/2011
• Issued: 11/25/2014
• Est. Priority Date: 01/29/2010
• Status: Active Grant

• Current Assignee: Geospatial Technology Associates LLC
• Original Assignee: Rochester Institute of Technology
• Inventors: Basener, William
• Primary Examiner(s): Bitar, Nancy
• Application Number:
  US13/015,894
• Publication Number:
  US 20120070033A1
• Time in Patent Office: 1,397 Days
• Field of Search: None
• US Class Current: 382/103
• CPC Class Codes: G06V 10/25 Determination of region of interest [ROI] or a volume of interest [VOI]
• CPC Class Codes: G06V 10/267 by performing operations on regions, e.g. growing, shrinking or watersheds

Abstract

A method, non-transitory computer readable medium, and apparatus that provides object-based identification, sorting and ranking of target detections includes determining a target detection score for each pixel in each of one or more images for each of one or more targets. A region around one or more of the pixels with the determined detection scores which are higher than the determined detection scores for the remaining pixels in each of the one or more of images is identified. An object based score for each of the identified regions in each of the one or more images is determined. The one or more identified regions with the determined object based score for each region is provided.

Claims

1. A method for identification, sorting and ranking detections of one or more targets, the method comprising:

determining with a target detection processing apparatus a target detection score for each pixel of a spectral image for one or more targets by obtaining with the target detection processing apparatus a signature for one or more of the targets for the image, and applying with the target detection processing apparatus the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by its statistical score;

identifying with the target detection processing apparatus a region around one or more of the pixels with the determined detection scores which are higher than a first score in said image;

determining with the target detection processing apparatus an object-based spectral identification score for each of the identified regions in said image; and

providing with the target detection processing apparatus the one or more identified regions with the determined object-based score for each region.

2. The method as set forth in claim 1 wherein the determining with a target detection processing apparatus a target detection score for each pixel further comprising applying with the target detection processing apparatus a statistical target detection filter to each pixel in said image for each of the one or more targets to determine the detection score for each of the pixels.

6. The method as set forth in claim 1 further comprising obtaining with the target detection processing apparatus the image.

7. The method as set forth in claim 6 wherein the obtaining the image further comprises obtaining with the target detection processing apparatus one or more hyperspectral images.

8. The method as set forth in claim 1 further comprising providing with the target detection processing apparatus geographic location information for each the one or more identified regions in said image.

9. The method as set forth in claim 1 further comprising ranking with the target detection processing apparatus each of the identified regions based on the object-based score for one or more of the targets, wherein the providing further comprises providing the ranking for each of the one or more identified regions for one or more of the targets.

28. The method as set forth in claim 1 wherein the identifying further comprises:

masking out from each of the regions with the target detection processing apparatus the one or more of the pixels in the image which have the highest determined statistical detection scores in the image;

selecting with the target detection processing apparatus one or of the pixels which are end members for each of the regions; and

identifying with the target detection processing apparatus a convex hull of the selected end members as a boundary for the one or more regions for the image.

29. The method as set forth in claim 1 wherein the determining the object based score further comprises:

unmixing with the target detection processing apparatus the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing with the target detection processing apparatus target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.

3. A method for identification, sorting and ranking detections of one or more targets, the method comprising:

determining with a target detection processing apparatus a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying with the target detection processing apparatus a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining with the target detection processing apparatus an object-based score for each of the identified regions in each of the one or more images; and

providing with the target detection processing apparatus the one or more identified regions with the determined object-based score for each region, wherein the determining a target detection score further comprises;

obtaining with the target detection processing apparatus a signature for one or more of the targets for each of the one or more images; and

applying with the target detection processing apparatus the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by a number of standard deviations from a mean for each of the one or more images to determine the statistical deviation score.

4. A method for identification, sorting and ranking detections of one or more targets, the method comprising:

determining with a target detection processing apparatus a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying with the target detection processing apparatus a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining with the target detection processing apparatus an object-based score for each of the identified regions in each of the one or more images; and

providing with the target detection processing apparatus the one or more identified regions with the determined object-based score for each region, wherein the identifying further comprises;

masking out from each of the regions with the target detection processing apparatus the one or more of the pixels in each of the one or more images which have the highest determined statistical detection scores in each of the one or more images;

selecting with the target detection processing apparatus one or of the pixels which are end members for each of the regions; and

identifying with the target detection processing apparatus a convex hull of the selected end members as a boundary for the one or more regions for each of the one or more of images.

5. A method for identification, sorting and ranking detections of one or more targets, the method comprising:

determining with a target detection processing apparatus a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying with the target detection processing apparatus a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining with the target detection processing apparatus an object-based score for each of the identified regions in each of the one or more images; and

providing with the target detection processing apparatus the one or more identified regions with the determined object-based score for each region, wherein the determining the object based score further comprises;

unmixing with the target detection processing apparatus the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing with the target detection processing apparatus target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.

10. A non-transitory computer readable medium having stored thereon instructions for identification, sorting and ranking detections of one or more targets comprising machine executable code which when executed by at least one processor, causes the processor to perform steps comprising:

determining a target detection score for each pixel in a spectral image for each of one or more targets by obtaining a signature for one or more of the targets for the image, and applying the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by its statistical score;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in said image;

determining an object-based spectral identification score for each of the identified regions in said image; and

providing the one or more identified regions with the determined object-based spectral identification score for each region.

11. The medium as set forth in claim 10 wherein the determining a target detection score for each pixel further comprising applying a statistical target detection filter to each pixel in said image for each of the one or more targets to determine the detection score for each of the pixels.

15. The medium as set forth in claim 10 further comprising obtaining with the target detection processing apparatus the image.

16. The medium as set forth in claim 15 wherein the obtaining the image further comprises obtaining one or more hyperspectral images.

17. The medium as set forth in claim 10 further comprising providing geographic location information for each the one or more identified regions in said image.

18. The medium as set forth in claim 10 further comprising ranking each of the identified regions based on the object-based score for one or more of the targets, wherein the providing further comprises providing the ranking for each of the one or more identified regions for one or more of the targets.

30. The medium as set forth in claim 10 wherein the identifying further comprises:

masking out from each of the regions the one or more of the pixels in the image which have the highest determined statistical detection scores in the image;

selecting one or of the pixels which are end members for each of the regions; and

identifying a convex hull of the selected end members as a boundary for the one or more regions for each of the one or more of images.

31. The medium as set forth in claim 10 wherein the determining the object based score further comprises:

unmixing the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.

12. A non-transitory computer readable medium having stored thereon instructions for identification, sorting and ranking detections of one or more targets comprising machine executable code which when executed by at least one processor, causes the processor to perform steps comprising:

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the determining a target detection score further comprises;

obtaining a signature for one or more of the targets for each of the one or more images; and

applying the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by a number of standard deviations from a mean for each of the one or more images to determine the statistical deviation score.

13. A non-transitory computer readable medium having stored thereon instructions for identification, sorting and ranking detections of one or more targets comprising machine executable code which when executed by at least one processor, causes the processor to perform steps comprising:

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the identifying further comprises;

masking out from each of the regions the one or more of the pixels in each of the one or more images which have the highest determined statistical detection scores in each of the one or more images;

selecting one or of the pixels which are end members for each of the regions; and

identifying a convex hull of the selected end members as a boundary for the one or more regions for each of the one or more of images.

14. A non-transitory computer readable medium having stored thereon instructions for identification, sorting and ranking detections of one or more targets comprising machine executable code which when executed by at least one processor, causes the processor to perform steps comprising:

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the determining the object based score further comprises;

unmixing the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.

19. A target detection processing apparatus comprising:

one or more processors;

a memory coupled to the one or more processors which are configured to execute programmed instructions stored in the memory comprising;

determining a target detection score for each pixel in a spectral image for each of one or more targets by obtaining a signature for one or more of the targets for the image, and applying the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by its statistical score;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in said image;

determining an object-based spectral identification score for each of the identified regions in said image; and

providing the one or more identified regions with the determined object-based spectral identification score for each region.

20. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the determining the target detection score for each pixel further comprising applying a statistical target detection filter to each pixel in said image for each of the one or more targets to determine the detection score for each of the pixels.

24. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory further comprising obtaining with the target detection processing apparatus the image.

25. The apparatus as set forth in claim 24 wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the obtaining the image further comprising obtaining one or more hyperspectral images.

26. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory further comprising providing geographic location information for each the one or more identified regions in said image.

27. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory further comprising ranking each of the identified regions based on the object-based score for one or more of the targets, wherein the providing further comprises providing the ranking for each of the one or more identified regions for one or more of the targets.

32. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the identifying further comprising:

masking out from each of the regions the one or more of the pixels in the image which have the highest determined statistical detection scores in the image;

selecting one or of the pixels which are end members for each of the regions; and

identifying a convex hull of the selected end members as a boundary for the one or more regions for the image.

33. The apparatus as set forth in claim 19 wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the determining the object based score further comprising:

unmixing the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.

21. A target detection processing apparatus comprising:

one or more processors;

a memory coupled to the one or more processors which are configured to execute programmed instructions stored in the memory comprising;

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the determining a target detection score further comprising;

obtaining a signature for one or more of the targets for each of the one or more images; and

applying the statistical target detection filter using the one or more obtained signatures to rank each of the pixels by a number of standard deviations from a mean for each of the one or more images to determine the statistical deviation score.

22. A target detection processing apparatus comprising:

one or more processors;

a memory coupled to the one or more processors which are configured to execute programmed instructions stored in the memory comprising;

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the identifying further comprising;

masking out from each of the regions the one or more of the pixels in each of the one or more images which have the highest determined statistical detection scores in each of the one or more images;

selecting one or of the pixels which are end members for each of the regions; and

identifying a convex hull of the selected end members as a boundary for the one or more regions for each of the one or more of images.

23. A target detection processing apparatus comprising:

one or more processors;

a memory coupled to the one or more processors which are configured to execute programmed instructions stored in the memory comprising;

determining a target detection score for each pixel in each of one or more images for each of one or more targets;

identifying a region around one or more of the pixels with the determined detection scores which are higher than a first score in each of the one or more of images;

determining an object-based score for each of the identified regions in each of the one or more images; and

providing the one or more identified regions with the determined object-based score for each region, wherein the one or more processors is further configured to execute programmed instructions stored in the memory for the determining the object based score further comprising;

unmixing the pixel with highest determined statistical detection score in each of the identified regions by finding one or more abundances to determine pixel spectra; and

comparing target spectra for each of the one or more targets with the determined pixel spectra to determine the object based score for each of the identified regions.