Data volume reduction for imaging radar polarimetry

• US 4,829,303 A
• Filed: 05/18/1988
• Issued: 05/09/1989
• Est. Priority Date: 05/18/1988
• Status: Expired due to Fees

• Current Assignee: California Institute of Technology
• Original Assignee: United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration
• Inventors: Held, Daniel N., Dubois, Pascale C., Zebker, Howard A., van Zul, Jakob J., Norikane, Lynne
• Primary Examiner(s): Tarcza, Thomas H.
• Assistant Examiner(s): Sotomayor, John B.
• Application Number:
  US07/195,225
• Time in Patent Office: 356 Days
• Field of Search: 342/25, 342/99, 342/145, 342/188, 342/189, 342/190, 342/192, 342/195
• US Class Current: 342/25.00A
• CPC Class Codes: G01S 13/9027 Pattern recognition for feature extraction
• CPC Class Codes: G01S 13/9076 Polarimetric features in SAR
• CPC Class Codes: G01S 7/024 using polarisation effects in waveguides H01P1/165; for aerials H01Q, e.g. H01Q15/22, H01Q15/24, H01Q19/195

Abstract

Two alternative methods are disclosed for digital reduction of synthetic aperture multipolarized radar data using scattering matrices, or using Stokes matrices, of four consecutive along-track pixels to produce "averaged" data for generating a synthetic polarization image.

Claims

1. A method for reduction of digital synthetic aperture multipolarized radar data while still allowing full polarimetric utility of said data, comprised of averaging and storing matrices corresponding to four along-track consecutive pixels, and using a set of average matrices thus stored to synthesize an image by computing a power signal for each average matrix corresponding to a four-look pixel after first choosing the transmit and received polarization vectors, thereby reducing the data set and number of operations required to synthesize an image for displays.

2. A method as defined in claim 1 wherein the process of reducing said data set for one image by averging matrices corresponding to four along-track consecutive pixels comprises computing the average scattering matrix corresponding to each of said along-track consecutive pixels of an image, and wherein synthesis of an image from said reduced data set comprises first choosing transmit and receive polarization vectors, and then computing for each image pixel a synthesized signal from the equation

3. A method as defined in claim 2 wherein the four scattering matrices to be averaged are given by ##EQU10## wherein the power in each element of the averaged matrix is computed as follows:

space="preserve" listing-type="equation">|S.sub.xx |=sqrt(A.sub.xx A.sub.xx *+B.sub.xx B.sub.xx *+C.sub.xx C.sub.xx *+D.sub.xx D.sub.xx)

space="preserve" listing-type="equation">|S.sub.xy |=sqrt(A.sub.xy A.sub.xy *+B.sub.xy B.sub.xy *+C.sub.xy C.sub.xy *+D.sub.xy D.sub.xy)

space="preserve" listing-type="equation">|S.sub.yx |=sqrt(A.sub.yx A.sub.yx *+B.sub.yx B.sub.yx *+C.sub.yx C.sub.yx *+D.sub.yx D.sub.yx *)

space="preserve" listing-type="equation">|S.sub.yy |=sqrt(A.sub.yy A.sub.yy *+B.sub.yy B.sub.yy *+C.sub.yy C.sub.yy *+D.sub.yy D.sub.yy *)

and the phase of S_xx is set to zero while the phases at other terms S_xy, S_yx, and S_yy are preserved as follows;

space="preserve" listing-type="equation">Arg(S.sub.xx)=0

space="preserve" listing-type="equation">Arg(S.sub.xy)=Arg(A.sub.xy A.sub.xx *+B.sub.xy B.sub.xx *+C.sub.xy C.sub.xx *+D.sub.xy D.sub.xx *)

space="preserve" listing-type="equation">Arg(S.sub.yx)=Arg(A.sub.yx A.sub.xx *+B.sub.yx B.sub.xx *+C.sub.yx C.sub.xx *+D.sub.yx D.sub.xx *)

space="preserve" listing-type="equation">Arg(S.sub.yy)=Arg(A.sub.yy A.sub.xx *+B.sub.yy B.sub.xx *+C.sub.yy C.sub.xx *+D.sub.yy D.sub.xx *).

4. A method as defined in claim 3 wherein each pixel has coded information said coded information includes four amplitudes, one for each of S_xx, X_yx and S_yy, and three phases, one for each of S_xy, S_yx and S_yy in 8 bytes as follows:

2bytes for amplitude stored as mantissa and exponent for the first amplitude S_xx,

3 bytes for the three remaining amplitudes S_xy, S_yx and S_yy,

3 bytes for the three phases,

and the phase of S_xx is set to zero so it need not be stored.

5. A method as defined in claim 1 wherein the process of reducing said data set for one image by averaging matrices corresponding to four along-track consecutive pixels, comprises transforming each scattering matrix of four-look pixels into a Stokes matrix, and adding the four resulting Stokes matrices to produce a Stokes matrix corresponding to each four-look pixel, and wherein synthesis of an image from said reduced data set comprises first choosing transmit and receive polarization vectors, transforming said chosen vectors into Stokes vectors of the transmitting and receiving vectors, and then computing for each time pixel a synthesized power signal from the equation