Apparatus, system and method for highlighting anomalous change in multi-pass synthetic aperture radar imagery

• US 10,535,127 B1
• Filed: 11/14/2017
• Issued: 01/14/2020
• Est. Priority Date: 01/11/2017
• Status: Active Grant

• Current Assignee: National Technology & Engineering Solutions Of Sandia LLC (Honeywell International Inc.)
• Original Assignee: National Technology & Engineering Solutions Of Sandia LLC (Honeywell International Inc.)
• Inventors: Simonson, Katherine M., Lizarraga, Ivan, Perkins, David Nikolaus
• Primary Examiner(s): Diallo, Mamadou L
• Application Number:
  US15/812,238
• Time in Patent Office: 791 Days
• CPC Class Codes: G01S 13/90 using synthetic aperture techniques , e.g. synthetic aperture radar [SAR] techniques
• CPC Class Codes: G01S 13/9023 combined with interferometric techniques
• CPC Class Codes: G01S 13/9027 Pattern recognition for feature extraction
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/10044 Radar image
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06T 7/337 involving reference images or patches

Abstract

Described herein are various technologies relating to constructing a statistically-normalized coherence (SNC) image. A plurality of synthetic aperture radar (SAR) images of a scene are generated based upon radar signals directed towards and reflected off of the scene, and a plurality of coherence change detection (CCD) images of the scene are generated based upon the SAR images. The CCD images are registered with one another, and mean and variance are computed on a pixel-wise basis. A new CCD image is subsequently received, and registered with the plurality of CCD images. The SNC image is generated based upon the computed mean and variance values for each pixel in the registered CCD images, and further based upon values of pixels in the new CCD image. The SNC image identifies locations in the scene where anomalous activity is represented in the new CCD image.

Claims

1. A system comprising:

a radar system, the radar system comprises a transmitter that is configured to emit radar signals and a receiver that is configured to detect radar signals;

at least one processor; and

memory that stores instructions that, when executed by the at least one processor, cause the at least one processor to perform acts comprising;

generating a plurality of synthetic aperture radar (SAR) images of a scene, the SAR images constructed based upon radar signals detected by the receiver of the radar system responsive to the radar signals being emitted by the transmitter towards the scene and reflecting off of the scene;

generating a plurality of coherency change detection (CCD) images of the scene based upon the plurality of SAR images, wherein each CCD image is constructed based upon a respective pair of SAR images in the plurality of SAR images;

receiving a new CCD image of the scene, the CCD image of the scene not included in the plurality of CCD images;

generating a statistically-normalized coherence (SNC) image of the scene based upon the plurality of CCD images and the CCD image, the SNC image comprises pixels that represent locations in the scene, wherein each pixel in the SNC image is assigned a respective value based upon values assigned to pixels in the plurality of CCD images that represent the same location in the scene as the pixel, and further wherein a first pixel in the pixels that represents a first location in the scene is assigned a first value that indicates that activity has occurred at the first location in the scene or that there is a lack of activity at the first location in the scene that is anomalous relative to normal activity patterns at the location in the scene represented in the plurality of CCD images; and

causing the SNC image to be displayed on a display of an analyst.

2. The system of claim 1, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the mean of values of pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the SNC image based upon the mean of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

3. The system of claim 1, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the variance of values of pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the SNC image based upon the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

4. The system of claim 1, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the mean of values of pixels in the plurality of CCD images that represent the first location in the scene;

computing the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the SNC image based upon the mean of the values of the pixels in the plurality of CCD images that represent the first location in the scene and the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

5. The system of claim 4, wherein the new CCD image comprises a second pixel that represents the first location in the scene, wherein the second pixel is assigned a second value, and wherein assigning the first value to the first pixel in the SNC image further comprises:

determining a difference between the mean value and the second value; and

assigning the first value to the first pixel in the SNC image based upon the difference between the mean value and the second value.

6. The system of claim 5, wherein assigning the first value to the first pixel in the SNC image further comprises:

dividing the difference by the square root of the variance to generate a result; and

assigning the first value to the first pixel in the SNC image based upon the result.

7. The system of claim 6, wherein the first pixel in the SNC image is depicted in a first color when the difference is positive, and wherein the first pixel in the SNC image is depicted in a second color when the difference is negative, the first color being different from the second color.

8. The system of claim 1, wherein the plurality of CCD images comprises at least six CCD images.

9. The system of claim 1, the acts further comprising:

prior to generating the SNC image, registering the plurality of CCD images to one another.

10. The system of claim 1, the acts further comprising:

prior to generating the SNC image, selecting the plurality of CCD images from a larger collection of CCD images, wherein the plurality of CCD images are selected from the larger collection of CCD images based upon times assigned to CCD images in the larger collection of CCD images.

11. A method that facilitates indicating to an analyst that an anomalous alteration has been identified at a location in a scene, the method comprising:

emitting, from a radar system transmitter, radar signals towards the scene as an aircraft that includes the radar system transmitter passes by the scene;

detecting, with a radar system receiver, the radar signals that have reflected off of the scene;

repeating the acts of emitting and detecting for a plurality of passes by the scene;

generating a plurality of synthetic aperture radar (SAR) images of the scene for the plurality of passes by the scene;

generating a plurality of coherence change detection (CCD) images of the scene based upon the plurality of SAR images of the scene, wherein each CCD image is generated based upon a respective pair of SAR images in the plurality of SAR images;

generating a statistically-normalized coherence (SNC) image based upon the plurality of CCD images and a new CCD image, the new CCD image based upon a new pair of SAR images, the SNC image comprises pixels that represent locations in the scene, wherein each pixel in the SNC image is assigned a respective value based upon values assigned to pixels in the plurality of CCD images that represent the same location in the scene as the pixel, and further wherein a first pixel in the SNC image is assigned a first value that indicates that an alteration has occurred at the first location in the scene that is anomalous relative to activity at the location represented in the plurality of CCD images; and

causing the SNC image to be displayed on a display of an analyst.

12. The method of claim 11, wherein the plurality of CCD images comprise at least six CCD images.

13. The method of claim 11, further comprising selecting the plurality of SAR images from a larger collection of SAR images prior to generating the plurality of CCD images, wherein the plurality of SAR images are selected from the larger collection of SAR image based upon timestamps assigned thereto.

14. The method of claim 11, further comprising registering the plurality of CCD images to one another, wherein the plurality of CCD images are registered to one another prior to generating the SNC image.

15. The method of claim 11, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the mean of values of pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the SNC image based upon the mean of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

16. The method of claim 11, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the variance of values of pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the SNC image based upon the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

17. The method of claim 11, wherein generating the SNC image comprises:

assigning the first value to the first pixel in the SNC image, wherein assigning the first value to the first pixel in the SNC image comprises;

computing the mean of values of pixels in the plurality of CCD images that represent the first location in the scene;

computing the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene; and

assigning the first value to the first pixel in the image based upon the mean of the values of the pixels in the plurality of CCD images that represent the first location in the scene and the variance of the values of the pixels in the plurality of CCD images that represent the first location in the scene.

18. The method of claim 17, wherein the new CCD image comprises a second pixel that represents the first location in the scene, wherein the second pixel is assigned a second value, and wherein assigning the first value to the first pixel in the SNC image further comprises:

determining a difference between the mean value and the second value; and

assigning the first value to the first pixel in the SNC image based upon the difference between the mean value and the second value.

19. The method of claim 18, wherein assigning the first value to the first pixel in the SNC image further comprises:

dividing the difference by the square root of the variance to generate a result; and

assigning the first value to the first pixel in the SNC image based upon the result.

20. A computer-readable storage medium comprising instructions that, when executed by a processor, cause the processor to perform acts comprising:

receiving a plurality of coherence change detection (CCD) images of a scene, each CCD image in the plurality of CCD images is based upon a respective pair of synthetic aperture radar (SAR) images of the scene;

registering the plurality of CCD images to generate a CCD image stack, wherein pixel (i,j) in each CCD image in the CCD image stack has a respective value and represents the same location in the scene, and further wherein the CCD image stack comprises at least six CCD images;

computing a mean of values of pixels (i,j) of the plurality of CCD images in the CCD image stack;

computing a variance of values of the pixels (i,j) of the plurality of CCD images in the CCD image stack;

subsequent to computing the mean and the variance, receiving a new CCD image;

registering the new CCD image to the plurality of CCD images to align the new CCD image with the plurality of CCD images; and

generating a statistically-normalized coherence (SNC) image responsive to registering the new CCD image with the plurality of CCD images, wherein pixel (i,j) in the SNC image is assigned a value that is based upon the mean and the variance, wherein the value is indicative of activity at the location in the scene represented in the new CCD image that is anomalous relative to activity at the location in the scene represented in the plurality of CCD images.