Suppression of background clutter in video imagery
A method of suppressing background clutter in video imagery of a scene including a foreground target and background clutter includes applying successive images of the video imagery to a clutter suppression process, thereby producing video imagery of a derived scene in which the background clutter is suppressed. This process includes processing the current image to remove a fixed pattern noise associated with the camera, and spatially re-registering the current image to a fixed geolocation or known clutter pattern of the scene in a reference image of the successive images. The process also includes subjecting independent copies of the current image to two infinite impulse response filters having different time constants to produce filtered, time-delayed images, and differencing the filtered images to reduce the background clutter, thereby producing a clutter-suppressed version of the current image in which the background clutter is suppressed with respect to the foreground target.
1. An apparatus for suppressing background clutter with respect to a foreground target in video imagery, the apparatus comprising:
a processor configured to at least:
receive video imagery from a camera onboard a moving platform, the video imagery being of a scene including a foreground target and background clutter, the foreground target moving with respect to the background clutter;
apply successive images of successive frames of the video imagery to a clutter suppression process and thereby produce video imagery of a derived scene in which the background clutter is suppressed with respect to the foreground target, the clutter suppression process including for each image when a current image of the successive images:
process the current image of the video imagery to remove a fixed pattern nose associated with the camera;
spatially re-register the current image to a fixed geolocation or known clutter pattern of the scene in a reference image of the successive images; thereafter,
subject independent copies of the current image to two infinite impulse response (IIR) filters having different time constants to produce filtered images that are time-delayed with respect to each other;
difference the filtered images to reduce the background clutter, thereby produce a clutter-suppressed version of the current image in which the background clutter is suppressed with respect to the foreground target; and
render the video imagery of the derived scene including clutter-suppressed versions of the successive images.
2. The apparatus of claim 1 , wherein process the current image includes the processor further configured to determine the fixed pattern noise for removal from the current image, and determine the fixed pattern noise includes the processor further configured to at least:
subject a number of the successive images up to and including the current image to an IIR low-pass filter to produce IIR-filtered images in which the background clutter is smoothed into regionally-localized mean values with the fixed pattern noise being unaltered, the IIR-filtered images including an IIR-filtered version of the current image;
remove the regionally-localized mean values from the IIR-filtered images and the current image utilizing a spatial convolution operation;
remove the DC component from the IIR-filtered version of the current image to produce a DC-removed, IIR-filtered version of the current image; and
combine the IIR-filtered images and the current image with the regionally-localized mean values so removed, and thereby produce a scaling factor; and
normalize the DC-removed, IIR-filtered version of the current image using the scaling factor to produce an unbiased estimation of the fixed pattern noise that at least partially matches the fixed pattern noise in the current image.
3. The apparatus of claim 1 , wherein process the current image includes the processor further configured to perform a pixel-wise subtraction of the fixed pattern noise from the current image and thereby produce a processed image that contains the foreground target and background clutter with reduced or without the fixed pattern noise.
4. The apparatus of claim 1 , wherein the clutter suppression process further includes the processor further configured to apply a bad pixel map to the current image to remove invalid image data from any bad pixels therein, the bad pixel map being applied as the independent copies of the current image are subjected to the two IIR filters.
5. The apparatus of claim 1 , wherein the successive images are applied to the clutter suppression process at regular intervals and for each of which the processor is further configured to initialize the clutter suppression process with a particular geolocation of the scene, thereby establishing the reference image and the fixed geolocation or known clutter pattern to which the current image is spatially re-registered.
6. The apparatus of claim 5 , wherein the clutter suppression process is one of two parallel clutter suppression processes that are identical but alternately initialized at different times, and apply the successive images to the clutter suppression process includes the processor further configured to separately and independently apply the successive images to each of the two clutter suppression processes, and
wherein the processor is further configured to compose the video imagery of the derived scene by alternately switching between the two clutter suppression processes.
7. The apparatus of claim 6 , wherein the successive images are applied to the two clutter suppression processes at respective regular intervals of a number of frames of the video imagery that have an offset of a fraction of the number of frames, and wherein when either of the two clutter suppression processes initializes, compose the clutter-suppressed version of the successive images includes a switch to the other of the two clutter suppression processes in which the two IIR filters have achieved a steady state.
8. A method of suppressing background clutter with respect to a foreground target in video imagery, the method comprising:
receiving video imagery from a camera onboard a moving platform, the video imagery being of a scene including a foreground target and background clutter, the foreground target moving with respect to the background clutter;
applying successive images of successive frames of the video imagery to a clutter suppression process and thereby producing video imagery of a derived scene in which the background clutter is suppressed with respect to the foreground target, the clutter suppression process including for each image when a current image of the successive images:
processing the current image of the video imagery to remove a fixed pattern noise associated with the camera;
spatially re-registering the current image to a fixed geolocation or known clutter pattern of the scene in a reference image of the successive images; thereafter,
subjecting independent copies of the current image to two infinite impulse response (IIR) filters having different time constants to produce filtered images that are time-delayed with respect to each other;
differencing the filtered images to reduce the background clutter, thereby producing a clutter-suppressed version of the current image in which the background clutter is suppressed with respect to the foreground target; and
rendering the video imagery of the derived scene including clutter-suppressed versions of the successive images.
9. The method of claim 8 , wherein processing the current image includes determining the fixed pattern noise for removal from the current image, and determining the fixed pattern noise includes:
subjecting a number of the successive images up to and including the current image to an IIR low-pass filter to produce IIR-filtered images in which the background clutter is smoothed into regionally-localized mean values with the fixed pattern noise being unaltered, the IIR-filtered images including an IIR-filtered version of the current image;
removing the regionally-localized mean values from the IIR-filtered images and the current image utilizing a spatial convolution operation;
removing the DC component from the IIR-filtered version of the current image to produce a DC-removed, IIR-filtered version of the current image; and
combining the IIR-filtered images and the current image with the regionally-localized mean values so removed, and thereby producing a scaling factor; and
normalizing the DC-removed, IIR-filtered version of the current image using the scaling factor to produce an unbiased estimation of the fixed pattern noise that at least partially matches the fixed pattern noise in the current image.
10. The method of claim 8 , wherein processing the current image includes performing a pixel-wise subtraction of the fixed pattern noise from the current image and thereby producing a processed image that contains the foreground target and background clutter with reduced or without the fixed pattern noise.
11. The method of claim 8 , wherein the clutter suppression process further includes applying a bad pixel map to the current image to remove invalid image data from any bad pixels therein, the bad pixel map being applied as the independent copies of the current image are subjected to the two IIR filters.
12. The method of claim 8 , wherein the successive images are applied to the clutter suppression process at regular intervals and for each of which the method further comprises initializing the clutter suppression process with a particular geolocation of the scene, thereby establishing the reference image and the fixed geolocation or known clutter pattern to which the current image is spatially re-registered.
13. The method of claim 12 , wherein the clutter suppression process is one of two parallel clutter suppression processes that are identical but alternately initialized at different times, and applying the successive images to the clutter suppression process includes separately and independently applying the successive images to each of the two clutter suppression processes, and
wherein the method further comprises composing the video imagery of the derived scene by alternately switching between the two clutter suppression processes.
14. The method of claim 13 , wherein the successive images are applied to the two clutter suppression processes at respective regular intervals of a number of frames of the video imagery that have an offset of a fraction of the number of frames, and
wherein when either of the two clutter suppression processes initializes, composing the clutter-suppressed version of the successive images switches to the other of the two clutter suppression processes in which the two IIR filters have achieved a steady state.
15. A non-transitory computer-readable storage medium for suppressing background clutter with respect to a foreground target in video imagery, the non-transitory computer-readable storage medium having computer-readable program code portions stored therein that in response to execution by a processor, cause an apparatus to at least:
receive video imagery from a camera onboard a moving platform, the video imagery being of a scene including a foreground target and background clutter, the foreground target moving with respect to the background clutter;
apply successive images of successive frames of the video imagery to a clutter suppression process and thereby produce video imagery of a derived scene in which the background cutter is suppressed with respect to the foreground target, the clutter suppression process including for each image when a current image of the successive images:
process the current image of the video imagery to remove a fixed pattern noise associated with the camera;
spatially re-register the current image to a fixed geolocation or known clutter pattern of the scene in a reference image of the successive images; thereafter,
subject independent copies of the current image to two infinite impulse response (IIR) filters having different time constants to produce filtered images that are time-delayed with respect to each other;
difference the filtered images to reduce the background clutter, thereby produce a clutter-suppressed version of the current image in which the background clutter is suppressed with respect to the foreground target; and
render the video imagery of the derived scene including clutter-suppressed versions of the successive images.
16. The non-transitory computer-readable storage medium of claim 15 , wherein process the current image includes the apparatus further caused to determine the fixed pattern noise for removal from the current image, and determine the fixed pattern noise includes the apparatus further caused to at least:
subject a number of the successive images up to and including the current image to an IIR low-pass filter to produce IIR-filtered images in which the background clutter is smoothed into regionally-localized mean values with the fixed pattern noise being unaltered, the IIR-filtered images including an IIR-filtered version of the current image;
remove the regionally-localized mean values from the IIR-filtered images and the current image utilizing a spatial convolution operation;
remove the DC component from the IIR-filtered version of the current image to produce a DC-removed, IIR-filtered version of the current image; and
combine the IIR-filtered images and the current image with the regionally-localized mean values so removed, and thereby produce a scaling factor; and
normalize the DC-removed, IIR-filtered version of the current image using the scaling factor to produce an unbiased estimation of the fixed pattern noise that at least partially matches the fixed pattern noise in the current image.
17. The non-transitory computer-readable storage medium of claim 15 , wherein process the current image includes the apparatus further caused to perform a pixel-wise subtraction of the fixed pattern noise from the current image and thereby produce a processed image that contains the foreground target and background clutter with reduced or without the fixed pattern noise.
18. The non-transitory computer-readable storage medium of claim 15 , wherein the clutter suppression process further includes the apparatus further caused to apply a bad pixel map to the current image to remove invalid image data from any bad pixels therein, the bad pixel map being applied as the independent copies of the current image are subjected to the two IIR filters.
19. The non-transitory computer-readable storage medium of claim 15 , wherein the successive images are applied to the clutter suppression process at regular intervals and for each of which the apparatus is further caused to initialize the clutter suppression process with a particular geolocation of the scene, thereby establishing the reference image and the fixed geolocation or known clutter pattern to which the current image is spatially re-registered.
20. The non-transitory computer-readable storage medium of claim 19 , wherein the clutter suppression process is one of two parallel clutter suppression processes that are identical but alternately initialized at different times, and apply the successive images to the clutter suppression process includes the apparatus further caused to separately and independently apply the successive images to each of the two clutter suppression processes, and wherein the non-transitory computer-readable storage medium has computer-readable program code portions stored therein that in response to execution by the processor, cause the apparatus to further compose the video imagery of the derived scene by alternately switching between the two clutter suppression processes.
21. The non-transitory computer-readable storage medium of claim 20 , wherein the successive images are applied to the two clutter suppression processes at respective regular intervals of a number of frames of the video imagery that have an offset of a fraction of the number of frames, and wherein when either of the two clutter suppression processes initializes, compose the clutter-suppressed version of the successive images includes a switch to the other of the two clutter suppression processes in which the two IIR filters have achieved a steady state.