CN103941245A - Method for Freeman-Durden target decomposition of condensed polarization interference data - Google Patents

Abstract

The invention provides a method for Freeman-Durden target decomposition of condensed polarization interference data. The method comprises the steps of first deducing two target vectors of pi/4 mode condensed polarization interference data at two ends of a base line, and synthesizing a cross-correlation measurement matrix Jint; applying the same method to deducing the target vectors according to a full-polarization scattering matrix of three scattering mechanisms, synthesizing cross-correlation matrix models Js, Jd and Jvol of the three scattering mechanisms, expressing the cross-correlation matrix models as weighting summation of the three scattering cross-correlation matrixes to obtain an underdetermined nonlinear equation set; turning the underdetermined nonlinear equation set into a determined nonlinear equation set capable of being solved through a numerical method; and finally applying a numerical solution method to obtain the scattering phase center height and power contribution of the three scattering mechanisms. According to the method, not only can the power contribution quantity of every scattering mechanism be obtained, but also the scattering phase center height of every scattering mechanism can be distinguished, and a novel technical means is provided for decomposition of polarization targets, ground-object recognition and height information extraction.

Description

Condense the Freeman-Durden target decomposition method of polarization interference data

Technical field

The present invention relates to radar remote sensing technology field, relate in particular to a kind of Freeman-Durden target decomposition method of condensing polarization interference data.

Background technology

Along with the fast development of Radar Technology, radar has been widely used in Aerospace Control in recent years, military detection, target detection, forestry remote sensing, the various fields such as environment supervision.In order to realize better radar remote sensing application and target detection, polarimetric radar is arisen at the historic moment and in the past flourish over 60 years.

Polarization target decomposition is an important application of polarimetric radar, and it is target classification, the basis of forestry remote sensing application.Target resolution theory can be divided into four large classes: the dichotomy based on Kennaugh matrix decomposes; Decomposition based on covariance matrix or coherent matrix eigenvalue; Decomposition based on scattering model (Freeman-Durden three-component decomposes, and Yamaguchi tetra-components decompose) and the Coherent decomposition based on scattering matrix.Wherein the Freeman-Durden three-component decomposition method based on model is most widely used.

Freeman-Durden decomposition can clearly pick out atural object and belong to which kind of scattering mechanism, but only rely on polarization data to carry out target and decompose the performance number that can only tell each scattering, can not tell the phase center elevation information of scattering mechanism, for example target decomposition is carried out in forest land, detecting single scattering and occupy an leading position, is source tree crown top layer or from earth's surface but can not distinguish single scattering on earth.When volume scattering and for example being detected and occupying an leading position, but can not determine that volume scattering is from trees or the higher earth's surface of roughness.

Summary of the invention

(1) technical matters that will solve

In view of above-mentioned technical matters, the invention provides a kind of Freeman-Durden target decomposition method of condensing polarization interference data.

(2) technical scheme

The Freeman-Durden target decomposition method that the present invention condenses polarization interference data comprises: steps A: the master image complete polarization scattering matrix S that obtains baseline one end ₁with the baseline other end from image complete polarization scattering matrix S ₂; Step B: at master image complete polarization scattering matrix S ₁with from image complete polarization scattering matrix S ₂in, extract and condense polarizing target vector respectively synthetic simple crosscorrelation observing matrix J _int, wherein:

$J_{\mathrm{int}}=\left[\begin{array}{cc}{J}_{11}& J_{12}\\ J_{21}& J_{22}\end{array}\right];$

Step C: set up respectively single scattering mechanism, rescattering mechanism and volume scattering mechanism and condensing cross-correlation matrix model J under polarization interference observation _s, J _dand J _vol, wherein:

$J_s=f_s\left[\begin{array}{cc}{\left|\mathrm{\β}\right|}^2& \mathrm{\β}\\ {\mathrm{\β}}^{*}& 1\end{array}\right],J_d=f_d\left[\begin{array}{cc}{\left|\mathrm{\α}\right|}^2& \mathrm{\α}\\ {\mathrm{\α}}^{*}& 1\end{array}\right],J_{\mathrm{vol}}=f_{\mathrm{vol}}\left[\begin{array}{cc}1& 0.5\\ 0.5& 1\end{array}\right]$

Wherein, ^*represent to get altogether to grip f _s, f _d, f _volbe respectively J _s, J _dand J _volcomplex weighting coefficient, f _s, β, f _d, α, f _volbe the unknown quantity of plural form; Step D: condensing cross-correlation matrix model J under polarization interference observation by single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed _s, J _d, and J _voland simple crosscorrelation observing matrix J _int, build and owe to determine Nonlinear System of Equations:

$\left\{\begin{array}{c}{J}_{11}=f_s{\left|\mathrm{\β}\right|}^2+f_d{\left|\mathrm{\α}\right|}^2+f_{\mathrm{vol}}\\ J_{12}=f_s{\mathrm{\β}}^{*}+f_d{\mathrm{\α}}^{*}+{0.5f}_{\mathrm{vol}}\\ J_{21}=f_s\mathrm{\β}+f_d\mathrm{\α}+{0.5f}_{\mathrm{vol}}\\ J_{22}=f_s+f_d+f_{\mathrm{vol}}\end{array}\right.$

Step e: by simple crosscorrelation observing matrix J _intin observed quantity J ₁₁, J ₁₂, J ₂₁, J ₂₂substitution owes to determine Nonlinear System of Equations, solves this and owes to determine 5 complex parameters in Nonlinear System of Equations α, β; And step F: according to solve 5 complex parameter f _s, f _d, f _vol, α, β, calculates the phase center height that single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed.

(3) beneficial effect

The present invention condenses in the Freeman-Durden target decomposition method of polarization interference data, to condensing polarization interference data, carry out Freeman-Durden decomposition, that the cross-correlation matrix that not only comprises polarization information but also comprise interference information is resolved into single scattering, the weighted sum of rescattering and volume scattering, use method of value solving to solve parameter in each scattering model and the weighting coefficient of three kinds of scatterings, the weighting coefficient of three kinds of scatterings that solve is plural number, amplitude represents the power contribution of this scattering mechanism, and phase place represents the phase center of this scattering.Vertical beam k in known earth's surface phase place and interference _zin situation, further obtain the vertical height that various scattering mechanisms occur, finally can obtain the power contribution of each scattering mechanism, can tell the height of each scattering mechanism again simultaneously.

Accompanying drawing explanation

Fig. 1 is for condensing the process flow diagram of the Freeman-Durden target decomposition method of polarization interference data according to the embodiment of the present invention;

Fig. 2 A is the master image of the full polarimetric SAR data of emulation, horizontal ordinate presentation video direction position, ordinate presentation video distance to;

Fig. 2 B be emulation full polarimetric SAR data from image, horizontal ordinate presentation video direction position, ordinate presentation video distance to;

Fig. 3 is the volume scattering power that estimates according to the bisectability theory of ripple, image horizontal ordinate represent orientation to, ordinate represent apart to.

Fig. 4 is for condensing polarization interference data decomposition result: the scattering phase place centre-height of three kinds of scattering mechanisms.Horizontal ordinate represents that orientation is to pixel, and ordinate represents height value, unit (m);

Fig. 5 is for condensing polarization interference data decomposition result: the amplitude contribution amount of three kinds of scattering mechanisms.Upper figure is three kinds of scattering power contributions of VV simple crosscorrelation passage, and figure below is three kinds of scattering power contributions of HH simple crosscorrelation passage, and horizontal ordinate represents that orientation is to pixel, and ordinate represents performance number, unit (dB).

Embodiment

For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in more detail.It should be noted that, in accompanying drawing or instructions description, similar or identical part is all used identical figure number.The implementation that does not illustrate in accompanying drawing or describe is form known to a person of ordinary skill in the art in affiliated technical field.In addition, although the demonstration of the parameter that comprises particular value can be provided herein, should be appreciated that, parameter is without definitely equaling corresponding value, but can in acceptable error margin or design constraint, be similar to corresponding value.The direction term of mentioning in embodiment, such as " on ", D score, 'fornt', 'back', " left side ", " right side " etc., be only the direction with reference to accompanying drawing.Therefore, the direction term of use is to be not used for limiting the scope of the invention for explanation.

Condensing polarization (Compact Polarimetry is called for short CP) is emerging a kind of polarization mode, by launching a kind of polarized wave, receives two-way orthogonally polarized wave, has also alleviated the burden of polar system in acquisition and full polarimetric SAR data approximate information.The advantage of Polarization technique is condensed in utilization, by condensing polarization data, does polarization interference (Compact Polarization SAR Inteferometry the is called for short C-PolInSAR) technology of condensing of interfere processing and also becomes study hotspot in recent years.π/4 pattern is condensed polarization and is referred to 45 °, one tunnel of transmitting line polarization wave, receives the mutually orthogonal H of two-way, V polarized wave.π/4 pattern of take is herein carried out the decomposition of Freeman-Durden target as being listed as to condensing polarization interference data.

In one exemplary embodiment of the present invention, provide a kind of Freeman-Durden target decomposition method of condensing polarization interference data.Fig. 1 is for condensing the process flow diagram of the Freeman-Durden target decomposition method of polarization interference data according to the embodiment of the present invention.Please refer to Fig. 1, the present embodiment Freeman-Durden target decomposition method comprises:

Steps A: the master image complete polarization scattering matrix S that obtains baseline one end ₁with the baseline other end from image complete polarization scattering matrix S ₂;

In this step, the method for obtaining satisfactory to both parties polarization scattering matrix is well known to those skilled in the art, and is not described in detail herein.

Step B: at master image complete polarization scattering matrix S ₁with from image complete polarization scattering matrix S ₂in, extract and condense polarizing target vector respectively synthetic simple crosscorrelation observing matrix J _int;

Wherein, this step B further comprises:

Sub-step B1, according to following formula, under the pattern of π/4 by master image complete polarization scattering matrix S ₁extract baseline one end and condense polarizing target vector

$\begin{array}{c}{\stackrel{\→}{k}}_1=S_1\·\frac{1}{\sqrt{2}}\left[\begin{array}{c}1\\ 1\end{array}\right]\\ =\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_1}& S_{{\mathrm{HV}}_1}\\ S_{{\mathrm{VH}}_1}& S_{{\mathrm{VV}}_1}\end{array}\right]\·\frac{1}{\sqrt{2}}\left[\begin{array}{c}1\\ 1\end{array}\right]\\ =\frac{1}{\sqrt{2}}\left[\begin{array}{c}{S}_{{\mathrm{HH}}_1}+S_{{\mathrm{HV}}_1}\\ S_{{\mathrm{VH}}_1}+S_{{\mathrm{VV}}_1}\end{array}\right]\end{array}---(1-1)$

Wherein, with respectively baseline one end master image complete polarization scattering matrix S ₁the polarization data of HH passage, HV passage, VH passage and VV passage;

Sub-step B2, according to following formula, under the pattern of π/4 by from image complete polarization scattering matrix S ₂extract the baseline other end and condense polarizing target vector

$\begin{array}{c}{\stackrel{\→}{k}}_2=S_2\·\frac{1}{\sqrt{2}}\left[\begin{array}{c}1\\ 1\end{array}\right]\\ =\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_2}& S_{{\mathrm{HV}}_2}\\ S_{{\mathrm{VH}}_2}& S_{{\mathrm{VV}}_2}\end{array}\right]\·\frac{1}{\sqrt{2}}\left[\begin{array}{c}1\\ 1\end{array}\right]\\ =\frac{1}{\sqrt{2}}\left[\begin{array}{c}{S}_{{\mathrm{HH}}_2}+S_{{\mathrm{HV}}_2}\\ S_{{\mathrm{VH}}_2}+S_{{\mathrm{VV}}_2}\end{array}\right]\end{array}---(1-2)$

Wherein, with respectively that the baseline other end is from image complete polarization scattering matrix S ₂the polarization data of HH passage, HV passage, VH passage and VV passage;

Sub-step B3, according to following formula, condenses polarizing target vector according to the baseline two ends of asking for obtain following simple crosscorrelation observing matrix J _int:

$J_{\mathrm{int}}={\stackrel{\→}{k}}_1{\stackrel{\→}{k}}_2^{*T}---\left(2\right)$

Wherein, symbol * T represents that matrix is carried out to conjugation goes to operation, $J_{\mathrm{int}}=\left[\begin{array}{cc}{J}_{11}& J_{12}\\ J_{21}& J_{22}\end{array}\right].$

Step C: set up respectively single scattering mechanism, rescattering mechanism and volume scattering mechanism and condensing cross-correlation matrix model under polarization interference observation:

J _s, J _dand J _volcomplex weighting coefficient be respectively complex parameter f _s, f _d, f _vol, wherein,

Wherein, single scattering mechanism condensing cross-correlation matrix model under polarization interference observation is:

$J_s=M_s\left[\begin{array}{cc}{\left|\mathrm{\β}\right|}^2& \mathrm{\β}\\ {\mathrm{\β}}^{*}& 1\end{array}\right]---\left(3\right)$

Wherein, M _sbe the unknown quantity of plural form with β, from physical significance, r _{υ 1}=S _vV1and R _h1=S _hH1, represent single scattering interferometric phase. for same polarization channel data phase differential, i.e. S in single scattering matrix _hH1with S _vV1phase differential.

Wherein, rescattering mechanism condensing cross-correlation matrix model under polarization interference observation is:

$J_d=M_d\left[\begin{array}{cc}{\left|\mathrm{\α}\right|}^2& \mathrm{\α}\\ {\mathrm{\α}}^{*}& 1\end{array}\right]---\left(4\right)$

Wherein, f _dbe the unknown quantity of plural form with α, from physical significance, for rescattering interferometric phase, for same polarization channel data phase differential in rescattering matrix.R _th1, R _gh1represent respectively in master image under horizontal polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface; R _tv1, R _gv1represent respectively under vertical polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface.

Wherein, volume scattering mechanism condensing cross-correlation matrix model under polarization interference observation is:

$J_{\mathrm{vol}}=M_{\mathrm{vol}}\left[\begin{array}{cc}1& 0.5\\ 0.5& 1\end{array}\right]---\left(5\right)$

Wherein, f _volfor the unknown quantity of plural form, from physical significance, for volume scattering interferometric phase.

1, introduce below single scattering mechanism in the derivation of condensing cross-correlation matrix model under polarization interference observation:

By master image single scattering complete polarization scattering matrix S _s1and from image single scattering complete polarization scattering matrix S _s2

$S_{s1}=\left[\begin{array}{cc}{R}_{h1}& 0\\ 0& R_{v1}\end{array}\right]$

(6)

$S_{s2}=\left[\begin{array}{cc}{R}_{h2}& 0\\ 0& R_{v2}\end{array}\right]$

And formula (1) show that the polarizing target vector of condensing at baseline two ends is:

${\stackrel{\→}{k}}_{s1}=\frac{1}{\sqrt{2}}{[R_{h1},R_{v1}]}^T$

(7)

${\stackrel{\→}{k}}_{s2}=\frac{1}{\sqrt{2}}{[R_{h2},R_{v2}]}^T$

Single scattering model cross-correlation matrix is

$\begin{array}{c}{J}_s={\stackrel{\→}{k}}_{s1}{\stackrel{\→}{k}}_{s2}^{*T}\\ =\frac{1}{2}\left[\begin{array}{cc}{R}_{h1}{R}_{h2}^{*T}& R_{h2}{R}_{v2}^{*T}\\ R_{v1}{R}_{h2}^{*T}& R_{v1}{R}_{v2}^{*T}\end{array}\right]\end{array}---\left(8\right)$

According to interference data feature, we suppose that master image is with identical from the amplitude of image equipolarization passage, and equipolarization phase difference between channels is that interferometric phase causes completely.

Obtain thus J _sfour elements be:

(9)

The cross-correlation matrix of finally deriving normalization single scattering model is

$J_s=M_s\left[\begin{array}{cc}{\left|\mathrm{\β}\right|}^2& \mathrm{\β}\\ {\mathrm{\β}}^{*}& 1\end{array}\right]---\left(10\right)$

Wherein

2, introduce below rescattering mechanism in the derivation of condensing cross-correlation matrix model under polarization interference observation:

Master image rescattering complete polarization scattering matrix S _d1and from image rescattering complete polarization scattering matrix S _d2for

$S_{d1}=\left[\begin{array}{cc}{e}^{{j2\mathrm{\γ}}_{h1}}{R}_{\mathrm{th}1}{R}_{\mathrm{gh}1}& 0\\ 0& e^{{j2\mathrm{\γ}}_{v1}}{R}_{\mathrm{tv}1}{R}_{\mathrm{gv}1}\end{array}\right]$

(11)

$S_{d2}=\left[\begin{array}{cc}{e}^{{j2\mathrm{\γ}}_{h2}}{R}_{\mathrm{th}2}{R}_{\mathrm{gh}1}& 0\\ 0& e^{{j2\mathrm{\γ}}_{v2}}{R}_{\mathrm{tv}2}{R}_{\mathrm{gv}2}\end{array}\right]$

R wherein _th1, R _gh1represent respectively in the master image of baseline one end under horizontal polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface; R _tv1, R _gv1represent respectively in master image under vertical polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface; γ _v1, γ _h1represent respectively in master image energy attenuation and the phase change in ripple transmitting procedure under horizontal polarization and vertical polarization.R _th2, R _gh2represent respectively the baseline other end from image under horizontal polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface; R _tv2, R _gv2represent respectively from image under vertical polarization vertically to object, and the Fresnel coefficient on horizontal earth's surface; γ _v2, γ _h2represent respectively from image energy attenuation and the phase change in ripple transmitting procedure under horizontal polarization and vertical polarization.According to actual physics model, we suppose R _th1=R _th2, R _gh1=R _gh2r _tv1=R _tv2, R _gv1=R _gv2.

By S _d1, S _d2and formula (1) draw baseline two ends condense polarizing target vector and

Finally by formula (2), draw J _d:

$J_d=M_d\left[\begin{array}{cc}{\left|\mathrm{\α}\right|}^2& \mathrm{\α}\\ {\mathrm{\α}}^{*}& 1\end{array}\right]---\left(12\right)$

Wherein for rescattering interferometric phase, for rescattering matrix S _d1in r _th1r _gh1, and r _tv1r _gv1two channel data phase differential.

3, introduce below volume scattering mechanism in the derivation of condensing cross-correlation matrix model under polarization interference observation:

Volume scattering complete polarization scattering matrix S in the master image of baseline one end _vol1, and baseline other end volume scattering complete polarization scattering matrix S from image _vol2

$S_{\mathrm{vol}1}=\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_1}& 0\\ 0& S_{{\mathrm{VV}}_1}\end{array}\right]$

(13)

$S_{\mathrm{vol}2}=\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_2}& 0\\ 0& S_{{\mathrm{VV}}_2}\end{array}\right]$

Around radar line of sight rotation θ, volume scattering matrix is modified to:

$S_1\left(\mathrm{\θ}\right)=\left[\begin{array}{cc}\cos \mathrm{\θ}& -\sin \mathrm{\θ}\\ \sin \mathrm{\θ}& \cos \mathrm{\θ}\end{array}\right]\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_1}& 0\\ 0& S_{{\mathrm{VV}}_1}\end{array}\right]\left[\begin{array}{cc}\cos \mathrm{\θ}& \sin \mathrm{\θ}\\ -\sin \mathrm{\θ}& \cos \mathrm{\θ}\end{array}\right]$

(14)

$S_2\left(\mathrm{\θ}\right)=\left[\begin{array}{cc}\cos \mathrm{\θ}& -\sin \mathrm{\θ}\\ \sin \mathrm{\θ}& \cos \mathrm{\θ}\end{array}\right]\left[\begin{array}{cc}{S}_{{\mathrm{HH}}_2}& 0\\ 0& S_{{\mathrm{VV}}_2}\end{array}\right]\left[\begin{array}{cc}\cos \mathrm{\θ}& \sin \mathrm{\θ}\\ -\sin \mathrm{\θ}& \cos \mathrm{\θ}\end{array}\right]$

Suppose that volume scattering is that superfine horizontal direction cylinder scatterer forms, now get $\left|S_{{\mathrm{HH}}_1}\right|=\left|S_{{\mathrm{HH}}_2}\right|=1,$

$S_1\left(\mathrm{\θ}\right)=\left[\begin{array}{cc}{\cos }^2\mathrm{\θ}& \cos \mathrm{\θ}\sin \mathrm{\θ}\\ \cos \mathrm{\θ}\sin \mathrm{\θ}& {\sin }^2\mathrm{\θ}\end{array}\right]$

(15)

be main, poor from image phase, i.e. volume scattering interferometric phase.

Formula (1) obtains baseline two ends and condenses polarizing target vector for:

${\stackrel{\→}{k}}_{\mathrm{vol}1}=\frac{1}{\sqrt{2}}{\left[\begin{array}{cc}{\cos }^2\mathrm{\θ}+\cos \mathrm{\θ}\sin \mathrm{\θ}& \cos \mathrm{\θ}\sin \mathrm{\θ}+{\sin }^2\mathrm{\θ}\end{array}\right]}^T$

(16)

By formula try to achieve preliminary volume scattering cross-correlation matrix J ', suppose that θ obedience is uniformly distributed (p (θ)=1/2 π) final body scattering cross-correlation matrix J _volmiddle element should be J ' _volelement carries out integration to θ and obtains, and integral formula is as follows:

$J_{{\mathrm{vol}}_{\mathrm{mn}}}={\∫}_0^{2\mathrm{\π}}\frac{1}{2\mathrm{\π}}{J}_{{\mathrm{vol}}_{\mathrm{mn}}}^{\′}\mathrm{d\θ}---\left(17\right)$

M, n=1,2 represent respectively matrix J ' _volrow and row.

Volume scattering model cross-correlation matrix J _volrepresent:

$J_{\mathrm{vol}}=M_{\mathrm{vol}}\left[\begin{array}{cc}1& 0.5\\ 0.5& 1\end{array}\right]---\left(18\right)$

Wherein

Step D, is condensing cross-correlation matrix model J under polarization interference observation by single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed _s, J _d, and J _volfoundation is condensed polarization interference target and is divided and solve an equation:

J _int＝aJ _s+bJ _s+cJ _vol (19)

Wherein a, b, c are respectively J _s, J _d, J _volthe real weighting coefficient of combination.

Redefine three parameter f _s, f _d, f _vol:

$\left\{\begin{array}{c}{f}_s={\mathrm{aM}}_s\\ f_d={\mathrm{bM}}_d\\ f_{\mathrm{vol}}={\mathrm{cM}}_{\mathrm{vol}}\end{array}\right.---\left(20\right)$

Be f _s, f _d, f _volbe respectively J _s, J _dand J _volcomplex weighting coefficient, its amplitude represents the power contribution of scattering mechanism, phase place represents the scattering elevation information of scattering mechanism.

Formula (19) is launched:

$J_{\mathrm{int}}=\left[\begin{array}{cc}{J}_{11}& J_{12}\\ J_{21}& J_{22}\end{array}\right]=\left[\begin{array}{cc}{f}_s{\left|\mathrm{\β}\right|}^2+f_d{\left|\mathrm{\α}\right|}^2+f_{\mathrm{vol}}& f_s{\mathrm{\β}}^{*}+f_d{\mathrm{\α}}^{*}+{0.5f}_{\mathrm{vol}}\\ f_s\mathrm{\β}+f_d\mathrm{\α}+{0.5f}_{\mathrm{vol}}& f_s+f_d+f_{\mathrm{vol}}\end{array}\right]---(21-1)$

Build system of equations:

$\left\{\begin{array}{c}{J}_{11}=f_s{\left|\mathrm{\β}\right|}^2+f_d{\left|\mathrm{\α}\right|}^2+f_{\mathrm{vol}}\\ J_{12}=f_s{\mathrm{\β}}^{*}+f_d{\mathrm{\α}}^{*}+{0.5f}_{\mathrm{vol}}\\ J_{21}=f_s\mathrm{\β}+f_d\mathrm{\α}+{0.5f}_{\mathrm{vol}}\\ J_{22}=f_s+f_d+f_{\mathrm{vol}}\end{array}\right.---(21-2)$

This system of equations comprises f _s, f _d, f _vol, α, five plural known variables of β, only have four known observed quantity J of plural number ₁₁, J ₁₂, J ₂₁, J ₂₂, these four observed quantities are to try to achieve matrix J in step B _intin four elements, this system of equations is for owing to determine Nonlinear System of Equations.

Step e, by simple crosscorrelation observing matrix J _intin observed quantity J ₁₁, J ₁₂, J ₂₁, J ₂₂substitution owes to determine Nonlinear System of Equations, solves this and owes to determine unknown quantity f in Nonlinear System of Equations _s, f _d, f _vol, α, β;

This step further comprises:

Sub-step E1, according in step B generate Stokes (Stokes) vector g ₁, g ₂, and then solve f _volamplitude F _vol;

Concrete solution procedure is as follows:

Son is E1a step by step, in steps A, obtains and condenses polarizing target vector that calculates baseline one end master image condenses polarized wave from coherence matrix C ₁, and the baseline other end is condensed polarized wave from coherence matrix C from image ₂, that is:

$C_1={\stackrel{\→}{k}}_1{\stackrel{\→}{k}}_1^{*T}$

(22)

$C_2={\stackrel{\→}{k}}_2{\stackrel{\→}{k}}_2^{*T}$

Son is E1b step by step, condenses the Stokes vector g of polarization data _ican be expressed as:

I=1,2 represent respectively the master and slave image in baseline two ends.C _{mn (1)}represent C ₁the capable n columns of the m value of matrix, C _{mn (2)}represent C ₂the capable n columns of the m value of matrix, represent to ask for real part, represent to ask for imaginary part.

Son is E1c step by step, introduces polarization m and characterizes wave polarization degree:

$m_i=\frac{\sqrt{g_{1\left(i\right)}^2+g_{2\left(i\right)}^2+g_{3\left(i\right)}^2}}{g_{0\left(i\right)}}---\left(24\right)$

I=1,2 represent respectively the master and slave image in baseline two ends.

According to formula (22), by three kinds of Stokes for scattering model (Stokes) vector representation, can be decomposed into condensing polarizing target echo:

Theoretical according to the bisectability of ripple, target scattering echo can be decomposed into perfact polarization component and depolarization component.Because volume scattering is described to the scattering process of completely random, polarization is 0, the depolarization component of corresponding ripple, and volume scattering power is approximately equal to the power of echo depolarization component.

Son is E1d step by step, according to formula (23), calculates respectively baseline two ends master image and polarization m from the reception echo of image ₁and m ₂, the depolarization component of two groups of echoes is made to geometric mean, obtain the power F of volume scattering component _vol:

$F_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|=\sqrt{(1-m_1)g_{0\left(1\right)}(1-m_2)g_{0\left(2\right)}}---\left(26\right)$

The volume scattering power estimating as shown in Figure 3.

Sub-step E2, according to the earth's surface phase solution f of forest periphery bare area _dphase place

In wood land, the mainly interaction between earth's surface and trunk from echo of even scattering, the approximate topographical surface that is positioned at of even scattering phase center, so we are approximately f by estimating earth's surface phase place _dphase place this routine mesorelief is without fluctuating, and earth's surface phase place can be estimated to draw by forest bare area around.

${\mathrm{\&gamma;}}_{\mathrm{HH}1-\mathrm{VV}1}=\frac{<({\mathrm{HH}}_1-{\mathrm{VV}}_1)\&CenterDot;{({\mathrm{HH}}_2-{\mathrm{VV}}_2)}^{*}>}{<{|{\mathrm{HH}}_1-{\mathrm{VV}}_1|}^2>\&CenterDot;<{|{\mathrm{HH}}_2-{\mathrm{VV}}_2|}^2>}---\left(28\right)$

Wherein, HH ₁, VV ₁, be respectively master image complete polarization scattering matrix S ₁hH and VV passage polarization data, HH ₂, VV ₂be respectively from image complete polarization scattering matrix S ₂hH and VV passage polarization data; Symbol <> represents on average.

Sub-step E3, by the volume scattering amplitude F solving _vol, and rescattering phase interference in substitution equation (20), by numerical solution Nonlinear System of Equations (20), solve f _s, f _damplitude, f _volphase place, α, β;

Solve f _volamplitude F _voland f _dphase place after, the phase and magnitude of multiple unknown quantity is counted to just surplus 8 real number unknown quantitys, 4 multiple observed quantities, also to there being 8 real known quantities, therefore adopt numerical method to carry out solving equation, and equation of constraint is:

F wherein _dfor rescattering weighting coefficient f _damplitude, for volume scattering weighting coefficient f _volphase place.

Step F, take earth's surface as reference point, and the phase center height of three kinds of scattering mechanisms is calculated by following formula.

Wherein, for vertical beam parameters; B _⊥for interfering the vertical component of baseline, λ is wavelength, R ₀for antenna is to atural object oblique distance, θ is radar incident angle, B _⊥, λ, R ₀be known quantity with θ.

Scattering phase place centre-height figure as shown in Figure 4.

According to solve 5 complex parameter f _s, f _d, f _vol, α, β, solves three kinds of scattering mechanisms respectively in the power contribution of HH simple crosscorrelation passage and VV simple crosscorrelation passage.

Three kinds of power contribution of HH passage are:

$\left\{\begin{array}{c}{H}_s=\left|f_s{\mathrm{\&beta;}}^2\right|\\ H_d=\left|f_d{\mathrm{\&alpha;}}^2\right|\\ H_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|\end{array}\right.---\left(31\right)$

Wherein, symbol || represent delivery value, H _srepresent HH passage single scattering power, H _drepresent HH passage rescattering power, H _volrepresent HH passage volume scattering power.

Three kinds of power contribution of VV passage are:

$\left\{\begin{array}{c}{V}_s=\left|f_s\right|\\ V_d=\left|f_d\right|\\ V_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|\end{array}\right.---\left(32\right)$

Wherein, V _srepresent VV passage single scattering power, V _drepresent VV passage rescattering power, V _volrepresent VV passage volume scattering power.

Because former data are larger, we only get in image data in a position line and implement decomposition experiment here, and the power contribution of decomposition as shown in Figure 5.

So far, by reference to the accompanying drawings the present embodiment be have been described in detail.According to above, describe, those skilled in the art tackle π/4 of the present invention pattern and condense the Freeman-Durden decomposition method of polarization interference target and had clearly understanding.

In sum, the present invention condenses the Freeman-Durden target decomposition method of polarization interference data.First the method derives two target vectors that π/4, baseline two ends pattern is condensed polarization data, synthetic simple crosscorrelation observing matrix J _int; Use same method according to the complete polarization scattering matrix derived object vector of three kinds of scattering mechanisms, the cross-correlation matrix model J of synthetic three kinds of scattering mechanisms _s, J _d, J _vol, the weighted sum that simple crosscorrelation observing matrix is expressed as to three kinds of scattering cross-correlation matrixs is J _int=f _sj _s+ f _dj _d+ f _volj _vol, three kinds of scattering mechanism weighting coefficients are plural number, and amplitude represents the power contribution of scattering mechanism, and phase place represents the scattering elevation information of scattering mechanism; J _int=f _sj _s+ f _dj _d+ f _volj _volbe one and owe to determine Nonlinear System of Equations, first according to the bisectability theory of ripple, solve the amplitude of volume scattering weight, i.e. f _volamplitude, then according to earth's surface phase information, solving rescattering phase place is f _dphase place.After two parametric solutions, owe to determine Nonlinear System of Equations and become and can use numerical method to solve the Nonlinear System of Equations of really shaping; Final power contribution and the scattering phase place centre-height of using method of value solving to try to achieve three kinds of scattering mechanisms.The power contribution amount that the present invention not only can obtain various scattering mechanisms can also pick out the scattering phase place centre-height of each scattering mechanism, is the decomposition of polarizing target, atural object identification and extract elevation information new technological means is provided.

Above-described specific embodiment; object of the present invention, technical scheme and beneficial effect are further described; institute is understood that; the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any modification of making, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (10)

1. a Freeman-Durden target decomposition method of condensing polarization interference data, is characterized in that, comprising:

Steps A: the master image complete polarization scattering matrix S that obtains baseline one end ₁with the baseline other end from image complete polarization scattering matrix S ₂;

Step B: at master image complete polarization scattering matrix S ₁with from image complete polarization scattering matrix S ₂in, extract and condense polarizing target vector respectively synthetic simple crosscorrelation observing matrix J _int, wherein:

$J_{\mathrm{int}}=\left[\begin{array}{cc}{J}_{11}& J_{12}\\ J_{21}& J_{22}\end{array}\right];$

Step C: set up respectively single scattering mechanism, rescattering mechanism and volume scattering mechanism and condensing cross-correlation matrix model J under polarization interference observation _s, J _dand J _vol, wherein:

$J_s=f_s\left[\begin{array}{cc}{\left|\mathrm{\&beta;}\right|}^2& \mathrm{\&beta;}\\ {\mathrm{\&beta;}}^{*}& 1\end{array}\right],J_d=f_d\left[\begin{array}{cc}{\left|\mathrm{\&alpha;}\right|}^2& \mathrm{\&alpha;}\\ {\mathrm{\&alpha;}}^{*}& 1\end{array}\right],J_{\mathrm{vol}}=f_{\mathrm{vol}}\left[\begin{array}{cc}1& 0.5\\ 0.5& 1\end{array}\right]$

Wherein, ^*represent to get altogether to grip f _s, f _d, f _volbe respectively J _s, J _dand J _volcomplex weighting coefficient, f _s, β, f _d, α, f _volbe the unknown quantity of plural form;

Step D: condensing cross-correlation matrix model J under polarization interference observation by single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed _s, J _d, and J _voland simple crosscorrelation observing matrix J _int, build and owe to determine Nonlinear System of Equations:

$\left\{\begin{array}{c}{J}_{11}=f_s{\left|\mathrm{\&beta;}\right|}^2+f_d{\left|\mathrm{\&alpha;}\right|}^2+f_{\mathrm{vol}}\\ J_{12}=f_s{\mathrm{\&beta;}}^{*}+f_d{\mathrm{\&alpha;}}^{*}+{0.5f}_{\mathrm{vol}}\\ J_{21}=f_s\mathrm{\&beta;}+f_d\mathrm{\&alpha;}+{0.5f}_{\mathrm{vol}}\\ J_{22}=f_s+f_d+f_{\mathrm{vol}}\end{array}\right.$

Step e: by simple crosscorrelation observing matrix J _intin observed quantity J ₁₁, J ₁₂, J ₂₁, J ₂₂substitution owes to determine Nonlinear System of Equations, solves this and owes to determine 5 complex parameters in Nonlinear System of Equations α, β; And

Step F: according to solve 5 complex parameter f _s, f _d, f _vol, α, β, calculates the phase center height that single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed.

2. Freeman-Durden target decomposition method according to claim 1, is characterized in that, described step e comprises:

Sub-step E1: according to condensing polarizing target vector generate Stokes vector g ₁, g ₂, and then solve f _volamplitude F _vol;

Sub-step E2: according to the earth's surface phase solution f of observation area periphery bare area _dphase place and

Sub-step E3: by the volume scattering amplitude F solving _voland rescattering phase interference substitution owes to determine Nonlinear System of Equations, adopts numerical value mode to solve f _s, f _damplitude, f _volphase place, α, β.

3. Freeman-Durden target decomposition method according to claim 2, is characterized in that, described sub-step E1 specifically comprises:

Son is E1a step by step: by condensing polarizing target vector that calculates baseline one end master image condenses polarized wave from coherence matrix C ₁, and the baseline other end is condensed polarized wave from coherence matrix C from image ₂:

$C_1={\stackrel{\&RightArrow;}{k}}_1{\stackrel{\&RightArrow;}{k}}_1^{*T}$

$C_2={\stackrel{\&RightArrow;}{k}}_2{\stackrel{\&RightArrow;}{k}}_2^{*T}$

Son is E1b step by step: the Stokes vector g that condenses polarization data _ibe expressed as:

Wherein, i=1,2 represent respectively the master and slave image in baseline two ends; C _{mn (1)}represent C ₁the capable n columns of the m value of matrix, C _{mn (2)}represent C ₂the capable n columns of the m value of matrix, represent to ask for real part, represent to ask for imaginary part;

Son is E1c step by step: introduce polarization m and characterize wave polarization degree:

$m_i=\frac{\sqrt{g_{1\left(i\right)}^2+g_{2\left(i\right)}^2+g_{3\left(i\right)}^2}}{g_{0\left(i\right)}}$

Son is E1d step by step: according to following formula, solve volume scattering component f _volpower F _vol:

$F_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|=\sqrt{(1-m_1)g_{0\left(1\right)}(1-m_2)g_{0\left(2\right)}}.$

4. Freeman-Durden target decomposition method according to claim 3, is characterized in that, in described sub-step E2, for f _dphase place

${\mathrm{\&gamma;}}_{\mathrm{HH}1-\mathrm{VV}1}=\frac{<({\mathrm{HH}}_1-{\mathrm{VV}}_1)\&CenterDot;{({\mathrm{HH}}_2-{\mathrm{VV}}_2)}^{*}>}{<{|{\mathrm{HH}}_1-{\mathrm{VV}}_1|}^2>\&CenterDot;<{|{\mathrm{HH}}_2-{\mathrm{VV}}_2|}^2>}$

Wherein, HH ₁, VV ₁, be respectively master image complete polarization scattering matrix S ₁hH and VV passage polarization data, HH ₂, VV ₂be respectively from image complete polarization scattering matrix S ₂hH and VV passage polarization data; Symbol <> represents on average.

5. the Freeman-Durden target decomposition method of stating according to claim 4, is characterized in that, in the process that described sub-step E3 employing numerical value mode solves, equation of constraint is:

6. Freeman-Durden target decomposition method according to claim 1, is characterized in that, in described step F, according to following formula, calculates the phase center height h that single scattering is machine-processed, rescattering is machine-processed and volume scattering is machine-processed _s, h _dand h _vol:

Wherein, for vertical beam parameters; B _⊥for interfering the vertical component of baseline, λ is wavelength, R ₀for antenna is to atural object oblique distance, θ is radar incident angle.

7. Freeman-Durden target decomposition method according to claim 1, is characterized in that, described step F also comprises:

According to solve 5 complex parameter f _s, f _d, f _vol, α, β, solves three kinds of scattering mechanisms respectively in the power contribution of HH simple crosscorrelation passage and VV simple crosscorrelation passage.

8. Freeman-Durden target decomposition method according to claim 7, is characterized in that, in described step F, according to following formula, calculates three kinds of scattering mechanisms respectively in the power contribution of HH simple crosscorrelation passage and VV simple crosscorrelation passage:

Three kinds of power contribution of HH passage are:

$\left\{\begin{array}{c}{H}_s=\left|f_s{\mathrm{\&beta;}}^2\right|\\ H_d=\left|f_d{\mathrm{\&alpha;}}^2\right|\\ H_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|\end{array}\right.,$

Wherein, symbol || represent delivery value, H _srepresent HH passage single scattering power, H _drepresent HH passage rescattering power, H _volrepresent HH passage volume scattering power;

Three kinds of power contribution of VV passage are:

$\left\{\begin{array}{c}{V}_s=\left|f_s\right|\\ V_d=\left|f_d\right|\\ V_{\mathrm{vol}}=\left|f_{\mathrm{vol}}\right|\end{array}\right.,$

Wherein, V _srepresent VV passage single scattering power, V _drepresent VV passage rescattering power, V _volrepresent VV passage volume scattering power.

9. according to the Freeman-Durden target decomposition method described in any one in claim 1 to 8, it is characterized in that, in described step D:

The described polarization interference target of condensing is divided and is solved an equation: J _int=f _sj _s+ f _dj _d+ f _volj _vol;

By single scattering mechanism, rescattering is machine-processed and volume scattering is machine-processed is condensing cross-correlation matrix model J under polarization interference observation _s, J _dand J _voldescribed in substitution, condense polarization interference target and divide and solve an equation, obtain:

$J_{\mathrm{int}}=\left[\begin{array}{cc}{J}_{11}& J_{12}\\ J_{21}& J_{22}\end{array}\right]=\left[\begin{array}{cc}{f}_s{\left|\mathrm{\&beta;}\right|}^2+f_d{\left|\mathrm{\&alpha;}\right|}^2+f_{\mathrm{vol}}& f_s{\mathrm{\&beta;}}^{*}+f_d{\mathrm{\&alpha;}}^{*}+{0.5f}_{\mathrm{vol}}\\ f_s\mathrm{\&beta;}+f_d\mathrm{\&alpha;}+0.5f_{\mathrm{vol}}& f_s+f_d+f_{\mathrm{vol}}\end{array}\right].$

10. according to the Freeman-Durden target decomposition method described in any one in claim 1 to 8, it is characterized in that, described step B comprises:

Sub-step B1, according to following formula, under the pattern of π/4 by master image complete polarization scattering matrix S ₁extract baseline one end and condense polarizing target vector

${\stackrel{\&RightArrow;}{k}}_1=\frac{1}{\sqrt{2}}\left[\begin{array}{c}{S}_{{\mathrm{HH}}_1}+S_{{\mathrm{HV}}_1}\\ S_{{\mathrm{HV}}_1}+S_{{\mathrm{VV}}_1}\end{array}\right]$

Wherein, with respectively baseline one end master image complete polarization scattering matrix S ₁the polarization data of HH passage, HV passage, VH passage and VV passage;

Sub-step B2, according to following formula, under the pattern of π/4 by from image complete polarization scattering matrix S ₂extract the baseline other end and condense polarizing target vector

${\stackrel{\&RightArrow;}{k}}_2=\frac{1}{\sqrt{2}}\left[\begin{array}{c}{S}_{{\mathrm{HH}}_2}+S_{{\mathrm{HV}}_2}\\ S_{{\mathrm{VH}}_2}+S_{{\mathrm{VV}}_2}\end{array}\right]$

Wherein, respectively that the baseline other end is from image complete polarization scattering matrix S ₂the polarization data of HH passage, HV passage, VH passage and VV passage;

Sub-step B3, according to following formula, condenses polarizing target vector according to the baseline two ends of asking for obtain following simple crosscorrelation observing matrix J _int:

$J_{\mathrm{int}}={\stackrel{\&RightArrow;}{k}}_1{\stackrel{\&RightArrow;}{k}}_2^{*T}$

Wherein, symbol * T represents that matrix is carried out to conjugation goes to operation.