SAR ATR treeline extended operating condition

• US 20090174595A1
• Filed: 03/09/2009
• Published: 07/09/2009
• Est. Priority Date: 09/22/2005
• Status: Active Grant

• Current Assignee: Albert Ezekiel, Nader Khatib
• Original Assignee: Albert Ezekiel, Nader Khatib
• Inventors: Khatib, Nader, Ezekiel, Albert
• Granted Patent: US 7,787,657 B2
• Time in Patent Office: Days
• US Class Current: 342/162
• CPC Class Codes: G01S 13/9027 Pattern recognition for feature extraction
• CPC Class Codes: G01S 13/9052 Spotlight mode
• CPC Class Codes: G01S 7/414 Discriminating targets with respect to background clutter

Abstract

A synthetic aperture radar acquires an image of one or more objects and identifies them as targets. The objects are located in the proximity of clutter within the image such of trees, or a tree line. The radar acquires a SAR image having pixels descriptive of the clutter and the object(s). Regions having object pixels are identified within the synthetic aperture image using an object identification (algorithm), where the object identification (algorithm) utilizes one or more historically known target characteristics and one or more measured characteristic to obtain an output. Boundaries are identified for the one or more objects within the output using an object isolation, such as, for example, a Watershed transform. Clutter pixels are identified external to the one or more objects. The clutter pixels are suppressed from the synthetic aperture image thereby generating a clutter reduced image containing the one or more objects. The objects are compared with known images of a probable target until a match is found, the match representing the target identification.

Claims

1-14. -14. (canceled)

15. A system for processing an image comprising:

first means for forming a density map of said image, wherein each pixel of said density map represents a density of bright pixels around a corresponding pixel in said image, and

second means for applying a Watershed transform to said density map to identify boundaries of one or more objects in said image.

16. The invention of claim 15 wherein said first means includes means for identifying a predetermined percentage of brightest pixels in said image.

17. The invention of claim 16 wherein said first means further includes means for computing a value for each pixel of said density map by determining a number of said brightest pixels located in a window around a corresponding pixel of said image.

18. The invention of claim 17 wherein said first means further includes means for applying a threshold to said density map to identify regions in said image containing one or more objects.

19. The invention of claim 18 wherein said system further includes means for splitting said regions to form said boundaries based on watersheds found by said Watershed transform.

20. The invention of claim 19 wherein said system further includes third means for suppressing selected pixels in said image located outside of a selected boundary to create a clutter reduced image.

21. The invention of claim 20 wherein said system further includes means for masking portions of said image to generate one or more masks and sorting said masks by a maximum density of each mask.

29. The invention of claim 19 wherein said system further includes fourth means for determining a number of objects in each said region or watershed.

30. The invention of claim 29 wherein said fourth means includes means for identifying density maxima in said region or watershed that fit one or more predetermined criteria.

31. The invention of claim 30 wherein said system further includes means for removing regions or watersheds containing no identified density maxima.

32. The invention of claim 19 wherein said system further includes means for combining adjacent watersheds having a peak to valley ratio less than a predetermined value.

33. The invention of claim 20 wherein said third means includes means for computing a cumulative density function of non-targeted regions of said image.

34. The invention of claim 32 wherein said third means further includes means for replacing pixels in said image located outside of said selected boundary and having a value above said cumulative density function with noise.

22. A method for processing an image for automatic target recognition comprising the steps of:

forming a density map of said image, wherein each pixel of said density map represents a density of bright pixels around a corresponding pixel in said image, and

applying a Watershed transform to said density map to identify boundaries of one or more objects in said image.

23. A method as described in claim 22 further including identifying a predetermined percentage of brightest pixels in said image.

24. A method as described in claim 23 wherein each pixel of said density map is computed by determining a number of said brightest pixels located in a window around a corresponding pixel of said image.

25. A method as described in claim 24 further including applying a threshold to said density map to identify regions in said image containing one or more objects.

26. A method as described in claim 25 further including separating objects in said regions and identifying boundaries of said objects based on watersheds found by said Watershed transform.

27. A method as described in claim 22 further including suppressing pixels in said image located outside of said boundary of a selected object to create a clutter reduced image.

28. A method as described in claim 22 further including masking portions of said image to generate one or more masks and sorting said masks by a maximum density of each mask.

35. A radar processing system comprising:

an analog to digital converter for converting a plurality of radar returns into a digital stream;

a computer for;

converting said digital stream into a radar image;

obtaining a density map of said image, wherein each pixel of said density map represents a density of bright pixels around a corresponding pixel in said image;

applying a Watershed transform to said density map to identify a boundary of each object in said image; and

replacing selected pixels in said image outside of the boundary of a selected object with noise to form a clutter reduced image; and

an automatic target recognition system adapted to receive said clutter reduced image and identify said selected object.