Multiply adaptive spatial spectral exploitation

• US 8,670,628 B2
• Filed: 08/16/2011
• Issued: 03/11/2014
• Est. Priority Date: 08/16/2011
• Status: Active Grant

• Current Assignee: Raytheon Company (Rtx Corporation)
• Original Assignee: Raytheon Company (Rtx Corporation)
• Inventors: Robinson, Ian S., Flanders, Bradley A., Frans, Eric P.
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): GILLIARD, DELOMIA L
• Application Number:
  US13/210,449
• Publication Number:
  US 20130044963A1
• Time in Patent Office: 938 Days
• Field of Search: None
• US Class Current: 382/250
• CPC Class Codes: G06T 2207/10032 Satellite or aerial image; Remote sensing
• CPC Class Codes: G06T 2207/10036 Multispectral image; Hyperspectral image
• CPC Class Codes: G06T 7/10 Segmentation; Edge detection motion-based segmentation G06T7/215
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06V 20/13 Satellite images
• CPC Class Codes: G06V 20/194 using hyperspectral data, i.e. more or other wavelengths than RGB

Abstract

A method of filtering hyperspectral image data associated with a hyperspectral image to produce a detection image data having a plurality of pixels, where the detection image data is associated with the degree to which a target may be present in a pixel. The method also includes adaptively processing the detection image data to determine a background variation in the plurality of pixels. The method additionally includes establishing a plurality of spatial filters for the detection image data, where each of the plurality of spatial filters are associated with energy being received at different locations on each of the plurality of pixels, and where the outputs of the plurality of spatial filters are weighted by the variation in background. The method further includes applying each of the plurality of spatial filers to the detection image data, such that each of the plurality of pixels are associated with a selected one of the plurality of spatial filters.

Claims

1. A method of filtering data associated with a hyperspectral image, comprising:

receiving detection image data having a plurality of pixels, the detection image data associated with the hyperspectral image;

processing the detection image data to determine a background in the plurality of pixels;

establishing a plurality of spatial filters for the detection image data, each of the plurality of spatial filters associated with energy being received at different locations on each of the plurality of pixels and assumes a different target phasing in which the energy is centered at different portions of the plurality of pixels, the plurality of spatial filters being weighted by the background; and

applying each of the plurality of spatial filers to the detection image data, such that each of the plurality of pixels are associated with a selected one of the plurality of spatial filters.

2. The method of claim 1, further comprising applying a detection filter configured to assign a value to each pixel of the hyperspectral image to generate the detection image.

3. The method of claim 1, wherein processing the detection image to determine the background comprises overlaying the detection image data with an offset copy of the detection image data.

4. The method of claim 3, wherein processing the detection image to determine the background further comprises calculating a spatial covariance matrix for the detection image.

5. The method of claim 3, wherein processing the detection image to determine the background further comprises calculating a spatial correlation matrix for the detection image.

6. The method of claim 1, wherein the selected one of plurality of spatial filters associated with each of the plurality of pixels is one of the plurality of spatial filters having a largest ratio of signal to clutter associated with one or more of the plurality of pixels in a location around each of the plurality of pixels.

7. The method of claim 1, further comprising determining signal peaks associated with the detection image.

8. The method of claim 7, wherein determining signal peaks comprises comparing one or more of the plurality of pixels with one or more surrounding pixels to determine if the one or more of the plurality of pixels has a greater value than the one or more surrounding pixels.

9. The method of claim 7, further comprising forming an inner region associated with one or more of the plurality of pixels including the signal peaks, and an outer region associated with one or more of the plurality of pixels surrounding the one or more of the plurality of pixels including the signal peaks.

10. The method of claim 9, further comprising:

adjusting a sum of values of the pixels in the outer region and a sum of pixels in the inner region for a relative numbers of pixels in the outer region and the inner region; and

subtracting the sum of pixels in the outer region from the sum of pixels in the inner region after said adjusting.

11. An article of manufacture comprising a physical nontransitory computer readable medium containing computer executable instructions for filtering data associated with a hyperspectral image which, when executed on a processor, carry out the functions of:

receiving detection image data having a plurality of pixels, the detection image data associated with the hyperspectral image;

processing the detection image data to determine a background in the plurality of pixels;

establishing a plurality of spatial filters for the detection image data, each of the plurality of spatial filters associated with energy being received at different locations on each of the plurality of pixels and assumes a different target phasing in which the energy is centered at different portions of the plurality of pixels, the plurality of spatial filters being weighted by the background; and

applying each of the plurality of spatial filers to the detection image data, such that each of the plurality of pixels are associated with a selected one of the plurality of spatial filters.

12. The article of manufacture of claim 11, wherein the function of processing the detection image to determine the background comprises overlaying the detection image data with an offset copy of the detection image data.

13. The article of manufacture of claim 11, wherein the selected one of plurality of spatial filters associated with each of the plurality of pixels is one of the plurality of spatial filters having a largest ratio of signal to clutter associated with one or more of the plurality of pixels in a location around each of the plurality of pixels.

14. A system for filtering data associated with a hyperspectral image comprising:

one or more processors configured to;

receive detection image data having a plurality of pixels, the detection image data associated with the hyperspectral image;

process the detection image data on the one or more processors to determine a background in the plurality of pixels;

establish, using the one or more processors, a plurality of spatial filters for the detection image data, each of the plurality of spatial filters associated with energy being received at different locations on each of the plurality of pixels and assumes a different target phasing in which the energy is centered at different portions of the plurality of pixels, the plurality of spatial filters being weighted by the background; and

apply each of the plurality of spatial filers to the detection image data, with the one or more processors, such that each of the plurality of pixels are associated with a selected one of the plurality of spatial filters.