US 2018/0003815 A1Published application

Method and system for detecting man-made objects using polarimetric, synthetic aperture radar imagery

AssigneeU.S. Army Research Laboratory

InventorsKENNETH IRVING RANNEY, David Chun Wong, Tuan That Ton

Life of this patent

Estimating the term…

Loading patent…

Abstract

A system and method for locating a man-made object comprising a transmitter and receiver combination or transceiver configured to emit mixtures of polarizations comprising HH, VV, VH and or HV polarization images, at least one processor configured to form co-polarimetric and cross-polarimetric images, to select a pixel under test and analyze the surrounding pixels by performing spatial averaging using the cross polarimetric image, and to replace the pixel under test and the pixels adjacent thereto with an average pixel value calculated from the pixel under test and pixels adjacent thereto; the at least one processor configured to diminish background effects to produce clearer co-polarimetric and cross-polarimetric images and to locate the left-right point of symmetry indicative of a man-made object by comparing each pixel under test in the cross-polarimetric image to pixels in the vicinity to locate an intensity differential in excess of 3 dB.

Claims18

Dependent claims

A system for determining the location of a man-made object based upon symmetry of the object comprising:

at least one of a transmitter and receiver combination or transceiver for emitting and receiving mixtures of polarizations and using receiving antennas with a specific polarization to thereby collect images from radar pulses, the receiver-transmitter mixtures of polarizations comprising horizontal-horizontal polarimetric images, vertical-vertical polarimetric images, vertical-horizontal polarimetric images and horizontal-vertical polarimetric images,
at least one processor, the at least one processor configured to combine the horizontal-horizontal polarimetric images and vertical-vertical polarimetric images to form co-polarimetric images and operate on one or both of the vertical-horizontal polarimetric images and horizontal-vertical polarimetric images to form cross-polarimetric images;
the at least one processor configured to process the co-polarimetric and cross-polarimetric images individually;
each of the co-polarimetric and cross-polarimetric images comprising a plurality of incrementally selected pixels under test, the at least one processor configured to select a pixel under test and analyze the surrounding pixels to determine whether a manmade object is present;
the at least one processor configured to perform spatial averaging using the cross polarimetric image by replacing the pixel under test and the pixels adjacent to the pixel under test with an average pixel value calculated from the pixel under test and pixels adjacent thereto;
using the co-polarimetric image, the at least one processor configured to determine the intensity of the background of the pixel under test and the surrounding pixels in order to diminish the effect of background to produce clearer co-polarimetric and cross-polarimetric images;
the at least one processor is configured to locate the left-right point of symmetry indicative of a man-made object by comparing each pixel under test in the cross-polarimetric image to pixels in the vicinity and if the intensity of the pixel under test differs by at least 3 dB, the pixel under test is determined to be a candidate pixel for locating a cross-range coordinate determinative of a point of symmetry indicating a man-made object.

4 more elements

A method for determining the location of a man-made object comprising the following steps, not necessarily in order;

inputting image data comprising four co-registered polarimetric SAR images of a common scene;
the four co-registered polarimetric images comprising horizontal-horizontal, horizontal-vertical, vertical-vertical and vertical-horizontal polarimetric images, the inputted image data comprising pixel values representing the polarimetric SAR images, a location in each of the four co-registered SAR images having a corresponding location in the other three co-registered SAR images;
each of the four co-registered SAR images being inputted into at least one processor, the at least one processor being configured to calculate a statistic indicating the likelihood that a manmade object is present by selecting a pixel under test and:
(i) using at least one processor, spatial averaging at a plurality of pixel locations in the vicinity of the pixel under test using the equation:
I
filter

(
x
,
y
)
=
∑
i
=
0
N
p

I

(
x
,
y
-
⌊
N
p
/
2
⌋
+
i
)
where I(x,y) denotes the image from either co-polarimetric or cross polarimetric radar data at pixel (x,y), where x and y are coordinates, Np is the number of pixels used for spatial averaging, and └Np/2┘ denotes the largest integer less than or equal to Np/2, where Np could be equal to zero for the co-polarimetric image;
(ii) capturing a background average by determining the number of cross-range cells (M) on either side of a pixel under test in the cross-pol image to calculate a background average for pixels on either side of the point of left-right symmetry and then specifying m, the number of “guard” cells on either side of the pixel under test to be “skipped” before calculating this background average, where m could be equal to zero;
(iii) for each pixel in the filtered cross-pol image, calculating the quantity:
I
cross


denominator

(
x
,
y
)
=
∑
i
=
m
+
1
M

I
filter
,

cross

(
x
-
i
,
y
)
+
∑
i
=
m
+
1
M

I
filter
,

cross

(
x
+
i
,
y
)
where, Ifilter, cross is the filtered cross-pol image and Icross denominator is used to determine the Icross contrast in the following equation;
(iv) for each pixel in the cross-pol image calculating the quantity:
I
cross


contrast

(
x
,
y
)
=
I
filter
,

cross

(
x
,
y
)
I
cross


denominator

(
x
,
y
)
(v) calculating a polarimetric manmade object detector output statistic TPMOD using the equation:
T
PMOD

(
x
,
y
)
=
I
filter
,

co

(
x
,
y
)
I
cross


contrast

(
x
,
y
)
,
where Ifilter, co(x,y) denotes the filtered co-pol image;
(vi) processing the image (two-dimensional array) of polarimetric manmade object output detector values to determine if the object under investigation is man-made.

162 more elements

A system for determining the location of a man-made object comprising:

at least one input configured to input image data comprising four co-registered SAR images of a common scene, each scene comprising a plurality of pixel values, the pixel values representing a radar cross section of the same region in each of the four basis polarizations;
at least one processor, each of the four co-registered SAR images being inputted into the at least one processor, the at least one processor being configured to calculate a statistical likelihood that a manmade object is present;
the at least one processor configured to perform spatial averaging at a plurality of pixel locations according to the equation:
I
filter

(
x
,
y
)
=
∑
i
=
0
N
p

I

(
x
,
y
-
⌊
N
p
/
2
⌋
+
i
)
where I(x,y) denotes the image from either co-polarimetric or cross polarimetric radar data at pixel (x,y), Np is the number of pixels used for spatial averaging, and └Np/2┘ denotes the largest integer less than or equal to Np/2, and Np possibly equal to zero for the co-polarimetric image;
the at least one processor configured to determine M, the number of cross-range cells on either side of a pixel under test in the cross-pol image in order to capture a background average for pixels on either side of the point of left-right symmetry;
the at least one processor configured to determine the number (m) of “guard” cells on either side of the pixel under test to be “skipped” before calculating this background average, where m may be equal to zero;
the at least one processor configured to calculate for each pixel in the filtered cross-polarimetric image the quantity:
I
cross


denominator

(
x
,
y
)
=
∑
i
=
m
+
1
M

I
filter
,

cross

(
x
-
i
,
y
)
+
∑
i
=
m
+
1
M

I
filter
,

cross

(
x
+
i
,
y
)
where, Ifilter, cross is the filtered cross-polarimetric image;
the at least one processor configured to, for each pixel in the cross-pol image, calculate the quantity:
I
cross


contrast

(
x
,
y
)
=
I
filter
,

cross

(
x
,
y
)
I
cross


denominator

(
x
,
y
)
;
the at least one processor configured to calculate a polarimetric manmade object detector output statistic using the equation:
T
PMOD

(
x
,
y
)
=
I
filter
,

co

(
x
,
y
)
I
cross


contrast

(
x
,
y
)
,
where Ifilter, co(x,y) denotes the filtered co-polarimetric image;
and
the at least one processor configured to create a two-dimensional array of values to determine if the object under investigation is man-made.

164 more elements

Dependencies are inferred from the claim text.

Facts

Application: 14/563,270Filed 8 Dec 2014
Published: 4 Jan 2018
Term ends: Not estimatedThe source records don’t support an estimate.

Classification6 codes

G01S 13/887
Radar or analogous systems specially adapted for specific applications: for detection of concealed objects, e.g. contraband or weapons
G01S 7/2923
Extracting wanted echo-signals: based on data belonging to a number of consecutive radar periods
G01S 13/9029
SAR image post-processing techniques: specially adapted for moving target detection within a single SAR image or within multiple SAR images taken…
G01S 13/904
Using synthetic aperture techniques , e.g. synthetic aperture radar [SAR] techniques: sAR modes