THREE-FRAME DIFFERENCE MOVING TARGET ACQUISITION SYSTEM AND METHOD FOR TARGET TRACK IDENTIFICATION

• US 20130094694A1
• Filed: 10/12/2011
• Published: 04/18/2013
• Est. Priority Date: 10/12/2011
• Status: Active Grant

• Current Assignee: Raytheon Company (Rtx Corporation)
• Original Assignee: Raytheon Company (Rtx Corporation)
• Inventors: Newman, Arthur M., Sun, Nicholas, Kaganovich, Vitaliy M., Douglas, Dave S.
• Granted Patent: US 8,611,600 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10032 Satellite or aerial image; Remote sensing
• CPC Class Codes: G06T 2207/30212 Military
• CPC Class Codes: G06T 7/254 involving subtraction of images

Abstract

Embodiments of a target-tracking system and method of determining an initial target track in a high-clutter environment are generally described herein. The target-tracking system may register image information of first and second warped images with image information of a reference image. Pixels of the warped images may be offset based on the outputs of the registration to align each warped images with the reference image. A three-frame difference calculation may be performed on the offset images and the reference image to generate a three-frame difference output image. Clutter suppression may be performed on the three-frame difference image to generate a clutter-suppressed output image for use in target-track identification. The clutter suppression may include performing a gradient operation on a background image to remove any gradient objects.

Claims

1. A three-frame difference moving target-tracking system for target track identification, the system configured to:

perform a three-frame difference calculation on a first offset image, a second offset image and a reference image to generate a three-frame difference output image; and

perform clutter suppression on the three-frame difference image to generate a clutter-suppressed output image.

2. The system of claim 1 further configured to:

perform a first registration to align image information of a first warped image with image information of the reference image;

perform a second registration to align image information of a second warped image with the image information of the reference image;

offset pixels of the first warped image based on the output of the first registration to align with corresponding pixels of the reference image to generate the first offset image; and

offset pixels of the second warped image based on the output of the second registration to align with corresponding pixels of the reference image to generate the second offset image.

3. The system of claim 2 wherein to perform the clutter suppression, the system is configured to:

calculate image background by subtracting the three-frame difference output image from the reference image and generate a background image;

perform a gradient operation on the background image to calculate a gradient of the background image; and

divide the three-frame difference output image by the gradient to generate the clutter-suppressed output image.

4. The system of claim 3 further configured to perform image processing on the clutter-suppressed output image to segment background pixels from non-background pixels to detect a moving target and identify a target track.

5. The system of claim 4 wherein the three-frame difference calculation comprises generating the three-frame difference output image based on the following equation:

[abs(D_C−

D_B)−

abs(D_A)]/2*sign(D_C) wherein;

D_A is a difference between corresponding pixels in the reference image and the second offset image,

D_B is a difference between corresponding pixels in the reference image and the first offset image, and

D_C is a difference between corresponding pixels in the first offset image and the second offset image.

6. The system of claim 3 further configured to:

delay a first frame of a selected group of frames by twice a predetermined time to generate a twice-delayed frame;

delay a second frame of the selected group of frames by the predetermined time to generate the reference frame;

warp, based on navigation data, a third frame of the selected group of frames to generate the first warped image; and

warp, based on navigation data, the twice-delayed frame to generate the second warped image.

7. The system of claim 6 wherein the frames are part of a sequence of frames generated by an imager, and

wherein the navigation data, comprises first and second navigation data streams and includes position information of the imager, directional information of the imager, and velocity information of the imager, when the imager is moving.

8. The system of claim 7 further configured to:

perform a transform on the first warped image to generate spatial frequency domain image information of the first warped image;

perform a transform on the second warped image to generate spatial frequency domain image information of the second warped image; and

perform a transform on the reference image to generate spatial frequency domain image information of the reference image,

wherein the first registration is performed to align the spatial frequency domain image information of the first warped image with the spatial frequency domain image information of a reference image, and

wherein the second registration is performed to align the spatial frequency domain image information of the second warped image with the spatial frequency domain image information of the reference image.

9. The system of claim 8 wherein the transforms comprise fast-Fourier transforms (FFTs),

wherein the first registration comprises a first correlation that is performed to correlate the spatial frequency domain image information of the first warped image with the spatial frequency domain image information of the reference image, and

wherein the second registration comprises performing a second correlation to correlate the spatial frequency domain image information of the second warped image with the spatial frequency domain image information of the reference image.

10. The system of claim 8 wherein the system is configured to delay, warp, transform, register, offset, perform a three-frame difference calculation and perform clutter suppression for a next set of three images for further use in target track identification.

11. The system of claim 8 wherein the system is further configured to:

select frames from a video stream generated from an imager based on frame-timing and target information, the target information including a length of a target, a velocity of the target, and

wherein the imager is configured to be located on a moving platform.

12. A method of target acquisition comprising:

registering image information of first and second warped images with image information of a reference image;

offsetting pixels of the warped images based on the outputs of the registering to align the warped images with the reference image;

performing a three-frame difference calculation on the offset images and the reference image to generate a three-frame difference output image; and

performing clutter suppression on the three-frame difference image to generate a clutter-suppressed output image for use in identifying a moving target.

13. The method of claim 12 further comprising:

delaying a first frame of a selected group of frames by twice a predetermined time to generate a twice-delayed frame;

delaying a second frame of the selected group of flames by the predetermined time to generate the reference frame;

warping, based on navigation data, a third frame of the selected group of frames to generate the first warped image; and

warping, based on navigation data, the twice-delayed frame to generate the second warped image.

14. The method of claim 13 wherein the frames are part of a sequence of frames generated by an imager, and

wherein the navigation data, comprises first and second navigation data streams and includes position information of the imager, directional information of the imager, and velocity information of the imager, when the imager is moving.

15. The method of claim 14 wherein the image information, of first and second warped images, comprises spatial frequency domain image information of first and second warped images, and

wherein the image information of the reference image comprises spatial frequency domain image information of the reference image.

16. The method of claim 14 wherein performing the three-frame difference calculation comprises generating the three-frame difference output image based on the following equation:

[abs(D_C−

D_B)−

abs(D_A)]/2*sign(D_C) wherein;

D_A is a difference between corresponding pixels in the reference age and the second offset image,

D_B is a difference between corresponding pixels in the reference image and the first offset image, and

D_C is a difference between corresponding pixels in the first offset image and the second offset image.

17. A moving platform for surveillance comprising:

an imager for generating a frames of video; and

a target-tracking system for acquiring and tracking a moving target based on the frames of video,

wherein the target-tracking system is configured to;

perform a first registration to align image information of a first warped image with image information of a reference image;

perform a second registration to align image information of a second warped image with the image information of the reference image;

offset pixels of the first warped image based on the output of the first registration to align with corresponding pixels of the reference image to generate a first offset image;

offset pixels of the second warped image based on the output of the second registration to align with corresponding pixels of the reference image to generate a second offset image;

perform a three-frame difference calculation on the first offset image, the second offset image and the reference image to generate a three-frame difference output image; and

perform clutter suppression on the three-frame difference image to generate a clutter-suppressed output image.

18. The moving platform of claim 17, wherein the moving platform is an unmanned aerial vehicle (UAV).

19. The moving platform of claim 18 wherein to perform the clutter suppression, the target-tracking system is configured to:

calculate image background by subtracting the three-frame difference output image from the reference image and generate a background image;

perform a gradient operation on the background image to calculate a gradient of the background image; and

divide the three-frame difference output image by the gradient to generate the clutter-suppressed output image.

20. The system of claim 19 wherein the target-tracking system is further configured to:

delay a first frame of a selected group of frames by twice a predetermined time to generate a twice-delayed frame;

delay a second frame of the selected group of frames by the predetermined time to generate the reference frame;

warp, based on navigation data, a third frame of the selected group of frames to generate the first warped image; and

warp, based on navigation data, the twice-delayed frame to generate the second warped image.