CN104020471A - Partitioning processing-based SAR real-time imaging method and system thereof - Google Patents

Abstract

The invention relates to a partitioning processing-based SAR real-time imaging method and a system thereof. A distance echo pulse compression module is used for carrying out pulse compression on each received distance echo datum; an azimuth dimension partitioning processing module is used for carrying out range migration correction and Doppler parameter and kinematic parameter estimation on azimuth subaperture data; a full-aperture parameter estimation module is used for carrying out conjoint analysis and estimation on each azimuth subaperture parameter to obtain full-aperture Doppler parameters and kinematic errors; an envelope compensation module is used for carrying out envelop compensation on the azimuth subaperture data; and a distance dimension partitioning processing module is used for carrying out phase compensation, azimuth pulse compression, multi-look processing and quantization output on data which has undergone envelope compensation. Then, partitioning processing-based SAR real-time imaging processing is finished. The method has low requirements on system memory, can be adopted to effectively solve the problem that range variance leads to poor image quality, and is especially suitable for a SAR real-time imaging system which has high requirements on real-time performance.

Description

A kind of SAR real time imagery method and system based on piecemeal processing

Technical field

The present invention relates to imaging radar field, relate in particular to a kind of SAR real time imagery method and system based on piecemeal processing.

Background technology

Synthetic-aperture radar (Synthetic Aperture Radar, SAR) is a kind of high-resolution imaging radar, and it is using respectively pulse compression technique and synthetic aperture principle apart from peacekeeping azimuth dimension, thereby obtains two-dimentional high-definition picture.SAR is widely used in the fields such as the meticulous mapping of aviation, environmental monitoring, Disaster Assessment, can obtain in time the high-resolution image information of paying close attention to target area.Along with the application of SAR is more and more extensive, the resolution to its imaging and real-time have also proposed very high requirement.

General SAR real time imagery method is based on full aperture processing, detailed process is: first the each of reception carried out to pulse compression apart from dimensional signal, until finish receiving after the data of whole aperture, entirety is carried out range migration correction, aperture parameters estimation, motion error extraction again, then carry out motion compensation, azimuth dimension pulse compression and look processing more, finally quantizing output.Under the condition of identical fabric width, the raising of imaging resolution will cause that echo data amount sharply increases, data processing time increases, and apart from space-variant, on image quality, impact also will become large simultaneously.In the time that imaging resolution is had relatively high expectations, the general SAR real time imagery method requirement system based on full aperture processing has great internal memory, cannot handle the impact apart from space-variant well, and real-time is also difficult to ensure simultaneously.

Summary of the invention

Technical matters to be solved by this invention is to provide a kind of SAR real time imagery method and system based on piecemeal processing, solve High Resolution SAR imaging to Installed System Memory require high, cause image quality variation and real-time to be difficult to the problem ensureing apart from space-variant.

The technical scheme that the present invention solves the problems of the technologies described above is as follows: a kind of SAR real time imagery method based on piecemeal processing, comprises the steps:

Step 1:SAR launches linear FM signal, receives each apart from go forward side by side horizontal pulse compression of echo data;

Step 2: when the distance echo data after pulse compression reaches the data volume in sub-aperture, an orientation, the distance echo data after paired pulses compression carries out range migration correction, Doppler parameter and kinematic parameter and estimates;

Step 3: in the time that the echo data receiving reaches the data volume of a full aperture, Doppler parameter and kinematic parameter to aperture, each side seat carry out Conjoint Analysis estimation, obtain full aperture Doppler parameter and kinematic error;

Step 4: sub-aperture, each orientation data are carried out respectively to envelope cancellation according to full aperture Doppler parameter and kinematic error;

Step 5: the full aperture data after envelope cancellation are carried out to distance dimension piecemeal, carry out phase compensation, orientation pulse pressure, look to process and quantize more and export apart from piece each.

The invention has the beneficial effects as follows: this method can complete the processing of SAR real time imagery, compared with the general formation method based on full aperture processing, this method greatly reduces the requirement of Installed System Memory, can effectively solve the problem that causes picture quality variation apart from space-variant, be particularly useful for the SAR Real Time Image System higher to requirement of real-time.

On the basis of technique scheme, the present invention can also do following improvement.

Further, in step 1, the signal of SAR transmitting is linear FM signal, and its expression formula is

$s\left(t_r\right)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \left\{\mathrm{j\π}{K}_r{t_r}^2\right\}$

Wherein, t _rfor the fast time, representative is apart from dimension time, T _pthe indicating impulse duration, K _rthe frequency modulation rate that represents chirp pulse signal, the impulse response of its corresponding matched filter is

$h\left(t_r\right)=s^{*}(-t_r)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \{-\mathrm{j\π}{K}_r{t_r}^2\}$

Apart from echo pulse pressure module, each receiving passed through to its matched filter apart from echo data, obtain apart from the data after pulse pressure, realize apart from high-resolution.

Further, the concrete steps of step 2 are:

Step 2.1: in advance full aperture is divided into sub-aperture, M orientation by azimuth dimension according to the processing power of hardware platform;

Step 2.2: in the time that the echo data receiving reaches the data volume in sub-aperture, an orientation, according to the processing power of hardware platform, the data in aperture, the party seat are carried out to the processing of distance dimension piecemeal, obtain several apart from piece;

Step 2.3: carry out range migration correction to each apart from piece, estimate the doppler centroid f in sub-aperture, orientation _dcwith doppler frequency rate f _dr;

Step 2.4: estimate the speed v of carrier aircraft along course-and-bearing according to each Doppler parameter apart from piece in aperture, the party seat _subacceleration a with the normal plane inner rays direction perpendicular to course line _sub.

Adopt the beneficial effect of above-mentioned further scheme: full aperture is divided into sub-aperture, several orientation in advance, needn't wait for and finishing receiving after the data of whole aperture, entirety is carried out data processing again, in the time that the data that receive reach the data volume in a sub-aperture and process, improve the real-time of system deal with data, reduce the requirement to Installed System Memory, and at the basis in sub-aperture, each orientation enterprising row distance dimension piecemeal, obtain less distance piece, migration correction to data and the estimation of Doppler parameter are all based on carrying out apart from piece, estimate the speed v of carrier aircraft along course-and-bearing by each Doppler parameter apart from piece in aperture, the party seat again _subacceleration a with the normal plane inner rays direction perpendicular to course line _sub, greatly improved the processing speed of system.

Further, the process of carrying out range migration correction in described step 2.3 is: sub-orientation aperture data are carried out to azimuth dimension Fourier transform and distance dimension Fourier transform, be multiplied by respectively range migration correction factor H in two-dimensional frequency ₁with second pulse compressibility factor H ₂, then carry out distance dimension inverse Fourier transform and azimuth dimension inverse Fourier transform, range migration correction factor H ₁with second pulse compressibility factor H ₂expression formula as shown in the formula

$\begin{array}{cc}{H}_1=\exp \left(j2\mathrm{\π}\frac{R_i}{C}{\left(\frac{f_d}{f_{\mathrm{dM}}}\right)}^2{f}_r\right)& H_2=\exp (-\mathrm{j\π}\frac{{2\mathrm{\λR}}_i{(f_d/f_{\mathrm{dM}})}^2{f_r}^2}{C^2{\left(\sqrt{1-{(f_d/f_{\mathrm{dM}})}^2}\right)}^3})\end{array}$

Wherein, R _irepresent i the line-of-sight distance corresponding apart from piece, C represents the light velocity, equals 3 × 10 ⁸meter per second, for echo Doppler, v is carrier aircraft actual measurement speed, and λ is carrier wavelength, and θ is actual measurement radar angle of squint, for being positioned at the echo Doppler of carrier aircraft dead ahead point target, f _rfor distance dimension frequency coordinate;

Wherein, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance.

Adopt the beneficial effect of above-mentioned further scheme: each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance, calibration is apart from the space-variant positive inaccuracy of moving of adjusting the distance effectively.

Further, the specific implementation of step 3:

Step 3.1: by the doppler centroid f in each sub-aperture, orientation _dccalculate the doppler centroid f of full aperture _dc0, f _dc0=mean (f _dc);

Step 3.2: the speed v according to carrier aircraft along course-and-bearing _{sub_all}acceleration a with the normal plane inner rays direction perpendicular to course line _{sub_all}calculating has the vector acceleration a apart from the directions of rays of space-variant _{r_var}vector acceleration a with the directions of rays without apart from space-variant _{r_fix}, computing formula as shown in the formula

$a_{R\_\mathrm{var}}=-\frac{v_{\mathrm{sub}\_\mathrm{all}}^2}{R_B}$

Wherein, R _bfor the vertical range in target and course line,

By cubic spline interpolation algorithm by the vector acceleration a of the normal plane inner rays direction perpendicular to course line _{sub_all}be interpolated into the length of counting of full aperture, obtain the vector acceleration a without the directions of rays apart from space-variant _{r_fix};

Step 3.3: calculate the normal plane inner rays direction resultant acceleration vector a of carrier aircraft perpendicular to course line _{r_sum}=a _{r_var}+ a _{r_fix};

Step 3.4: the carrier aircraft practical flight speed and the v that calculate full aperture ₀with radar angle of squint θ ₀,

$v_0=\sqrt{v_{\mathrm{LOS}}^2+v_{\mathrm{azi}}^2}=\sqrt{{\left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{2}\right)}^2+v_{\mathrm{azi}}^2}$

$v_{\mathrm{azi}}=\mathrm{mean}\left(v_{\mathrm{azi}\_\mathrm{sum}}\right)=\mathrm{mean}\left(\sqrt{{-a}_{R\_\mathrm{sum}}{R}_s}\right)$

${\mathrm{\θ}}_0=\arcsin \left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{{2v}_0}\right)$

Wherein, λ is carrier wavelength, f _dc0for the doppler centroid of full aperture, v _azicarrier aircraft is closed average velocity scalar, a along course-and-bearing _{r_sum}for carrier aircraft is perpendicular to the normal plane inner rays direction resultant acceleration vector in course line, R _sfor line-of-sight distance corresponding to full aperture data center;

Step 3.5: according to airborne practical flight speed v ₀with radar angle of squint θ ₀calculate the doppler frequency rate f of full aperture _dr0,

$f_{\mathrm{dr}0}=\frac{{2v}_0^2\cos {\mathrm{\θ}}_0}{{\mathrm{\λR}}_s}$

Wherein λ is carrier wavelength, R _sfor the line-of-sight distance of full aperture data center;

Step 3.6: calculate the directions of rays kinematic error R having apart from space-variant _{err_var}with the directions of rays kinematic error R without apart from space-variant _{err_fix};

R _{err_var}＝∫∫a _{R_var}

a _{err_fix}＝a _{R_fix}-mean(a _{R_fix})

R _{err_fix}＝∫∫a _{err_fix}

Wherein, a _{r_var}for thering is the vector acceleration apart from the directions of rays of space-variant, a _{r_fix}for the vector acceleration of the directions of rays without apart from space-variant, its a _{err_fix}for a _{r_fix}and the bias vector between acceleration average.

Adopt the beneficial effect of above-mentioned further scheme: obtain after full aperture data, carry out Conjoint Analysis estimation according to the Doppler parameter in aperture, each side seat and kinematic parameter, obtain full aperture Doppler parameter and kinematic error, instead of directly full aperture data are directly carried out to the estimation of Doppler parameter and kinematic error, reduce the requirement to Installed System Memory.

Further, the specific implementation of step 4 is:

Step 4.1: sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, obtain several apart from piece;

Step 4.2: carry out envelope cancellation to each apart from piece, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance,

Wherein, each while carrying out envelope cancellation apart from piece, advanced row distance dimension Fourier transform, then be multiplied by envelope correction factor H apart from frequency domain ₃, then carry out distance dimension inverse Fourier transform,

Wherein, $H_3=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_{\mathrm{Bi}}}\·R_{\mathrm{err}\_\mathrm{var}})\right)$

Wherein, R _{err_fix}without the directions of rays kinematic error apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant, R _sfor line-of-sight distance corresponding to full aperture data center, R _bifor line-of-sight distance corresponding to current distance piece.

Adopt the friendship effect of above-mentioned further scheme: carry out respectively envelope cancellation for sub-aperture, M orientation data, first sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, carry out envelope cancellation to each apart from piece again, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance, can solve the impact apart from mutability.

Further, the specific implementation of step 5 is:

Step 5.1: in advance the data of full aperture are divided into N apart from piece by distance dimension according to the processing power of hardware platform;

Step 5.2: go Doppler center, phase compensation, orientation pulse pressure, look processing more apart from piece each, obtain full aperture data x after treatment _ori;

Step 5.3: by the full aperture data x handling _oricarry out entirety and quantize, obtain a width SAR image.

Adopt the beneficial effect of above-mentioned further scheme: data are carried out after envelope cancellation, go Doppler center, phase compensation, orientation pulse pressure, look processing more, this writes and processes is according to distance dimension piecemeal by full aperture data, obtain some apart from piece, carry out above-mentioned processing to each apart from piece again, finally data after treatment being carried out to entirety quantizes, finally obtain SAR image, the method of piecemeal processing has improved the processing speed of system and the real-time of data processing greatly, has reduced the requirement to Installed System Memory.

Further, in step 5.2 to each apart from piece go Doppler center, phase compensation, orientation pulse pressure, look more process and entirety quantize process as follows:

A) eachly remove being embodied as of Doppler center apart from piece, each distance bar data apart from piece, are multiplied by respectively Doppler center factor H in time domain ₄, H ₄=exp (j2 π f _dc0t _a),

Wherein, f _dc0for the doppler centroid of full aperture, t _afor the slow time, represent the azimuth dimension time;

B) eachly carry out the each distance bar data apart from piece of being embodied as of phase compensation apart from piece, be multiplied by respectively phase compensating factor H in time domain ₅,

$H_5=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_i}\·R_{\mathrm{err}\_\mathrm{var}})\right),$

Wherein, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece, R _{err_fix}for the directions of rays kinematic error without apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant;

C) eachly carry out being embodied as of orientation pulse pressure apart from piece, each distance bar data apart from piece, carry out azimuth dimension Fourier transform in territory, orientation, are multiplied by orientation pulse compression match filter H ₆, then carry out azimuth dimension inverse Fourier transform,

$H_6=\exp \left(\mathrm{j\π}\frac{{f_a}^2}{(R_s/R_i)\·f_{\mathrm{dr}0}}\right),$

Wherein, f _afor orientation frequency coordinate, f _dr0for the doppler frequency rate of full aperture, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece;

D) eachly carry out looking being embodied as of processing apart from piece, each distance bar data apart from piece, are averaging every the data of azimuth dimension L point, carry out L to look after processing more, and azimuth dimension is counted and become M/L point from M point.

Further, in step 5.3 by the full aperture data x handling _orithe formula that carries out entirety quantification is as follows,

$x=\frac{x_{\mathrm{ori}}}{\mathrm{mean}\left(x_{\mathrm{ori}}\right)}\·p$

Wherein, x _orirepresent the data before quantizing, x represents the image after quantification, and p is quantization parameter.

Another technical scheme that the present invention solves the problems of the technologies described above is as follows: a kind of SAR Real Time Image System based on piecemeal processing, comprises apart from echo pulse pressure module, azimuth dimension piecemeal processing module, full aperture parameter estimation module, envelope cancellation module and distance dimension piecemeal processing module;

Described apart from echo pulse pressure module, it is for carrying out pulse compression by each receiving apart from echo data;

Described azimuth dimension piecemeal processing module, it carries out range migration correction, Doppler parameter and kinematic parameter for aperture, the other side seat data and estimates;

Described full aperture parameter estimation module, it,, for the parameter in aperture, each side seat is carried out to Conjoint Analysis estimation, obtains full aperture Doppler parameter and kinematic error;

Described envelope cancellation module, it carries out envelope cancellation for aperture, the other side seat data;

Described distance dimension piecemeal processing module, it carries out phase compensation, orientation pulse pressure, looks to process and quantize more and export for the data to after envelope cancellation.

Brief description of the drawings

Fig. 1 is a kind of SAR Real Time Image System block diagram based on piecemeal processing of the present invention;

Fig. 2 is a kind of SAR real time imagery method flow diagram based on piecemeal processing of the present invention.

In accompanying drawing, the list of parts of each label representative is as follows:

1, apart from echo pulse pressure module, 2, azimuth dimension piecemeal processing module, 3, full aperture parameter estimation module, 4, envelope cancellation module, 5, distance dimension piecemeal processing module.

Embodiment

Below in conjunction with accompanying drawing, principle of the present invention and feature are described, example, only for explaining the present invention, is not intended to limit scope of the present invention.

As shown in Figure 1, a kind of SAR Real Time Image System based on piecemeal processing, comprises apart from echo pulse pressure module 1, azimuth dimension piecemeal processing module 2, full aperture parameter estimation module 3, envelope cancellation module 4 and distance dimension piecemeal processing module 5;

Described apart from echo pulse pressure module 1, it is for carrying out pulse compression by each receiving apart from echo data;

Described azimuth dimension piecemeal processing module 2, it carries out range migration correction, Doppler parameter and kinematic parameter for aperture, the other side seat data and estimates;

Described full aperture parameter estimation module 3, it,, for the parameter in aperture, each side seat is carried out to Conjoint Analysis estimation, obtains full aperture Doppler parameter and kinematic error;

Described envelope cancellation module 4, it carries out envelope cancellation for aperture, the other side seat data;

Described distance dimension piecemeal processing module 5, it carries out phase compensation, orientation pulse pressure, looks to process and quantize more and export for the data to after envelope cancellation.

As shown in Figure 2, a kind of SAR real time imagery method based on piecemeal processing, comprises the steps:

Step 1:SAR launches linear FM signal, receives each apart from go forward side by side horizontal pulse compression of echo data;

Step 2: when the distance echo data after pulse compression reaches the data volume in sub-aperture, an orientation, the distance echo data after paired pulses compression carries out range migration correction, Doppler parameter and kinematic parameter and estimates;

Step 3: in the time that the echo data receiving reaches the data volume of a full aperture, Doppler parameter and kinematic parameter to aperture, each side seat carry out Conjoint Analysis estimation, obtain full aperture Doppler parameter and kinematic error;

Step 4: sub-aperture, each orientation data are carried out respectively to envelope cancellation according to full aperture Doppler parameter and kinematic error;

Step 5: the full aperture data after envelope cancellation are carried out to distance dimension piecemeal, carry out phase compensation, orientation pulse pressure, look to process and quantize more and export apart from piece each.

The first step, carries out pulse compression apart from echo pulse pressure module to receiving data

The signal of SAR transmitting is linear FM signal, and its expression formula is

$s\left(t_r\right)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \left\{\mathrm{j\π}{K}_r{t_r}^2\right\}---\left(1\right)$

Wherein, t _rrepresent the fast time, representative is apart from dimension time, T _pthe indicating impulse duration, K _rrepresent the frequency modulation rate of chirp pulse signal.The impulse response of its corresponding matched filter is

$h\left(t_r\right)=s^{*}(-t_r)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \{-\mathrm{j\π}{K}_r{t_r}^2\}---\left(2\right)$

Apart from echo pulse pressure module, each receiving passed through to its matched filter apart from echo data, obtain apart from the data after pulse pressure, realize apart from high-resolution.

Second step, aperture, azimuth dimension piecemeal processing module the other side seat data are carried out range migration correction, Doppler parameter and kinematic parameter and are estimated.

According to the processing power of hardware platform, full aperture is divided into M sub-aperture in advance by azimuth dimension, in the time that reception data reach the data volume in a sub-aperture, just can carries out azimuth dimension piecemeal and process.The distance piece that aperture, azimuth dimension piecemeal processing module the other side seat data are divided carries out range migration correction, and estimating Doppler centre frequency and doppler frequency rate, the airborne movement velocity of estimation and acceleration.

The process of range migration correction is: sub-orientation aperture data are carried out to azimuth dimension Fourier transform and distance dimension Fourier transform, be multiplied by respectively range migration correction factor H in two-dimensional frequency ₁with second pulse compressibility factor H ₂, then carry out distance dimension inverse Fourier transform and azimuth dimension inverse Fourier transform.When algorithm is realized, according to the processing power of hardware platform, first sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, carry out range migration correction to each apart from piece again, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance, calibration is apart from the space-variant positive inaccuracy of moving of adjusting the distance effectively.

The range migration correction factor and second pulse compressibility factor are respectively

$\begin{array}{c}{H}_1=\exp \left(j2\mathrm{\π}\frac{R_i}{C}{\left(\frac{f_d}{f_{\mathrm{dM}}}\right)}^2{f}_r\right)\end{array}---\left(3\right)$

$\begin{array}{cc}& H_2=\exp (-\mathrm{j\π}\frac{{2\mathrm{\λR}}_i{(f_d/f_{\mathrm{dM}})}^2{f_r}^2}{C^2{\left(\sqrt{1-{(f_d/f_{\mathrm{dM}})}^2}\right)}^3})\end{array}----\left(4\right)$

Wherein, R _irepresent i the line-of-sight distance corresponding apart from piece, C represents the light velocity, equals 3 × 10 ⁸meter per second, for echo Doppler, v is actual measurement carrier aircraft speed, and λ is carrier wavelength, and θ is actual measurement radar angle of squint, for being positioned at the echo Doppler (being maximum echo Doppler) of carrier aircraft dead ahead point target, f _rfor distance dimension frequency coordinate.

Estimation of Doppler central frequency adopts related function method.When algorithm is realized, the data of the sub-aperture of alternate orientation center section are estimated doppler centroid.Be located at while thering is no Doppler's off-centring, echo in orientation to power spectrum be S ₀(f), it is identical with antenna radiation pattern, and with zero-frequency symmetry, the related function that power spectrum is corresponding is R ₀(τ), be real function.In the time having Doppler shift, power spectrum S _b(f) be S ₀(f-f _dc), its related function becomes:

$R_b\left(\mathrm{\τ}\right)=e^{j2\mathrm{\π}{f}_{\mathrm{dc}}\mathrm{\τ}}{R}_0\left(\mathrm{\τ}\right)---\left(5\right)$

Because orientation echo is discrete sampling, so R _b(τ)=R _b(kT _a), the sample frequency of signal is PRF, T _a=1/PRF, makes k=1, obtains multifrequency and strangles centre frequency and be

$f_{\mathrm{dc}}=\frac{\mathrm{PRF}}{2\mathrm{\π}}\arg \left\{R_b\left(T_b\right)\right\}---\left(6\right)$

Wherein, arg{} is for asking phase angle function.

Doppler frequency rate estimates to adopt image shift method (MD algorithm).The full aperture time is divided into not overlapping two parts aperture by MD algorithm, utilizes quadratic phase to have different functional forms in two parts aperture, front and back.Constant, a component of degree n n and quadratic component can be resolved in every part aperture, and wherein constant is identical with quadratic component, and a component of degree n n makes the translation of two parts aperture image.MD algorithm, by estimating translational movement between two parts aperture image, is estimated the quadratic term coefficient in whole aperture.When algorithm is realized, according to the processing power of hardware platform, first sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, then carry out doppler frequency rate estimation to each apart from piece.

If echo sequence is s (t _a), wherein t _aits corresponding time period is [T, T].By s (t _a) do as down conversion

$s\left(t_a\right)=[s\left(t_a\right)\·\exp \{-\mathrm{j\π}{f}_{\mathrm{dr}}{t}_a^2\left\}\right]\·\exp \left\{\mathrm{j\π}{f}_{\mathrm{dr}}{t}_a^2\right\}=s^{\′}\left(t_a\right)\·\exp \left\{\mathrm{j\π}{f}_{\mathrm{dr}}{t}_a^2\right\}---\left(7\right)$

Echoed signal is divided into sub-aperture, left and right echoed signal

$s_L\left(t_a\right)=s(t_a-\frac{T}{2})=s^{\′}(t_a-\frac{T}{2})\·\exp \left\{\mathrm{j\π}{f}_{\mathrm{dr}}{(t_a-\frac{T}{2})}^2\right\}---\left(8\right)$

$s_R\left(t_a\right)=s(t_a+\frac{T}{2})=s^{\′}(t_a+\frac{T}{2})\·\exp \left\{\mathrm{j\π}{f}_{\mathrm{dr}}{(t_a+\frac{T}{2})}^2\right\}---\left(9\right)$

Upper two formulas are launched to arrange,

$s_L\left(t_a\right)=[s^{\′}(t_a-\frac{T}{2})\·\exp \{\mathrm{j\π}{f}_{\mathrm{dr}}({t_a}^2+\frac{T^2}{4})\left\}\right]\·\exp \{-\mathrm{j\π}{f}_{\mathrm{dr}}{\mathrm{Tt}}_a\}---\left(10\right)$

$s_R\left(t_a\right)=[s^{\′}(t_a+\frac{T}{2})\·\exp \{\mathrm{j\π}{f}_{\mathrm{dr}}({t_a}^2+\frac{T^2}{4})\left\}\right]\·\exp \left\{\mathrm{j\π}{f}_{\mathrm{dr}}{\mathrm{Tt}}_a\right\}---\left(11\right)$

Because front second half section s'(t _a) doppler spectral basic identical, in above two formula brackets the doppler spectral of represented signal also identical, establish it and be upper two formula doppler spectrals can be write as

So the Doppler's difference between the doppler spectral of second half section is in the past

Δf _a＝f _drT (14)

In the time completing above-mentioned processing procedure, conventionally adopt the method for digital signal processing.The sample frequency of signal is PRF, and the total duration of signal is 2T, and sampling number is N=2TPRF, and the sampling number of front and back half segment signal is N/2=TPRF, can obtain T=N/ (2PRF), so formula (13) is write discrete form is

${\mathrm{\Δf}}_a=f_{\mathrm{dr}}\frac{N}{2\mathrm{PRF}}---\left(15\right)$

If front and back half segment signal fast fourier transform is counted as N _fFT.Frequency spectrum to the sub-aperture signal in left and right is made Amplitude correlation, and now the peak value of related function is positioned at Δ N place, and Doppler's difference is

${\mathrm{\Δf}}_a=\frac{\mathrm{PRF}}{N_{\mathrm{FFT}}}\·\mathrm{\ΔN}---\left(16\right)$

Simultaneous formula (14) and formula (15), can obtain doppler frequency rate and be

$f_{\mathrm{dr}}=\frac{{2\mathrm{PRF}}^2}{N\·N_{\mathrm{FFt}}}\mathrm{\ΔN}---\left(17\right)$

Work as N _fFTwhile equaling the sampling number N/2 of front and back half segment signal, formula (15), can be reduced to

$f_{\mathrm{dr}}=\frac{{4\mathrm{PRF}}^2}{N^2}\mathrm{\ΔN}---\left(18\right)$

The estimation of airborne movement velocity and acceleration is to utilize linear relationship approximate between itself and doppler frequency rate to obtain.

If the Desired Track of carrier aircraft is expressed as [vt _a, 0,0], kinematic error component is [Δ x (t _a), Δ y (t _a), Δ z (t _a)], actual flight path can be expressed as [Vt _a+ Δ x (t _a), Δ y (t _a), Δ z (t _a)], radar downwards angle of visibility is β, n impact point P _ncoordinate be [x _n, y _n, z _n], P _nwith the vertical range in course line be R _b, radar physical location and impact point p _ndistance R (t _a) be

$R\left(t_a\right)=\sqrt{{({\mathrm{vt}}_a+\mathrm{\Δx}\left(t_a\right)-x_n)}^2+{(\mathrm{\Δy}\left(t_a\right)-y_n)}^2+{(\mathrm{\Δz}\left(t_a\right)-z_n)}^2}---\left(19\right)$

Utilize relation y _n=R _bsin β and z _n=R _bcos β, and ignore high-order term, formula (19) can be rewritten as

$R\left(t_a\right)=\sqrt{{({\mathrm{vt}}_a+\mathrm{\Δx}\left(t_a\right)-x_n)}^2+R_B^2+\mathrm{\Δy}{\left(t_a\right)}^2+\mathrm{\Δz}{\left(t_a\right)}^2-2\mathrm{\Δy}\left(t_a\right)R_B\sin \mathrm{\β}-2\mathrm{\Δz}\left(t_a\right)R_B\cos \mathrm{\β}}$

$\≈\sqrt{{({\mathrm{vt}}_a+\mathrm{\Δx}\left(t_a\right)-x_n)}^2+R_B^2+2\mathrm{\Δy}\left(t_a\right)R_B\sin \mathrm{\β}-2\mathrm{\Δz}\left(t_a\right)R_B\cos \mathrm{\β}}---\left(20\right)$

Make s=(vt _a+ Δ x (t _a)-x _n) ²-2 Δ y (t _a) R _bsin β-2 Δ z (t _a) R _bcos β, R (t _a) become R (s), R (s) is done to single order Maclaurin expansion, then become s again t _aexpression formula, and ignore high-order term, have

$R\left(t_a\right)\≈R_B+\frac{{({\mathrm{vt}}_a+\mathrm{\Δx}\left(t_a\right)-x_n)}^2}{{2R}_B}-\mathrm{\Δy}\left(t_a\right)\sin \mathrm{\β}-\mathrm{\Δz}\left(t_a\right)\cos \mathrm{\β}---\left(21\right)$

Consider that this point target phase of echo is

Above formula is carried out to differentiate can be obtained Doppler frequency and be

Wherein, v (t _a) represent that carrier aircraft is along the speed component of course-and-bearing, v _r(t _a) represent perpendicular to the normal plane inner rays in course line to speed component.

Above formula is carried out to differentiate again can be obtained doppler frequency rate and be

$f_{\mathrm{drn}}\left(t_a\right)=\frac{d}{{\mathrm{dt}}_a}{f}_{\mathrm{dn}}\left(t_a\right)=-\frac{{2v}^2\left(t_a\right)}{{\mathrm{\λR}}_B}-\frac{2[{\mathrm{vt}}_a+\mathrm{\Δx}\left(t_a\right)]a\left(t_a\right)}{{\mathrm{\λR}}_B}-\frac{2}{\mathrm{\λ}}{a}_R\left(t_a\right)---\left(24\right)$

Wherein, a (t _a) represent that carrier aircraft is along the component of acceleration of course-and-bearing, a _r(t _a) represent the component of acceleration perpendicular to the normal plane inner rays direction in course line.Because carrier aircraft machinery inertia is large, slow along the velocity variations in course line, therefore a (t _a) very little, can ignore, so above formula is write as

$f_{\mathrm{drn}}\left(t_a\right)=\frac{{2v}^2\left(t_a\right)}{{\mathrm{\λR}}_B}-\frac{2}{\mathrm{\λ}}{a}_R\left(t_a\right)---\left(25\right)$

Make v _sub=v (t _a), a _sub=a _r(t _a), above formula is organized into following form

$-\frac{{\mathrm{\λR}}_B}{2}{f}_{\mathrm{drn}}\left(t_a\right)=a_{\mathrm{sub}}\·R_B+v_{\mathrm{sub}}^2---\left(26\right)$

Write above formula as y _n=ax _nthe form of+b, wherein, x _n=R _b, a=a _sub, utilize least square method to solve, the speed v of carrier aircraft along course-and-bearing can be tried to achieve in sub-footpath, each orientation _subacceleration a with the normal plane inner rays direction perpendicular to course line _sub.

So far completed the range migration correction of sub-aperture, orientation data, and completed its doppler centroid f _dc, doppler frequency rate f _dr, carrier aircraft is along the speed v of course-and-bearing _subacceleration a with the normal plane inner rays direction perpendicular to course line _subestimation.

The 3rd step, the parameter that full aperture parameter estimation module is carried out sub-aperture, orientation to reception data is processed, and obtains full aperture Doppler parameter and kinematic error

The kinematic error of carrier aircraft is divided into and has apart from the directions of rays kinematic error of space-variant with without the directions of rays kinematic error apart from space-variant.Formula (21) is arranged, and ignore high-order term, can obtain

$R\left(t_a\right)\≈(R_B+\frac{{({\mathrm{vt}}_a-x_n)}^2}{{2R}_B})+\frac{({\mathrm{vt}}_a-x_n)}{R_B}\mathrm{\Δx}\left(t_a\right)-(\mathrm{\Δy}\left(t_a\right)\sin \mathrm{\β}+\mathrm{\Δz}\left(t_a\right)\cos \mathrm{\β})---\left(27\right)$

Above formula result can be divided into three, and wherein, Section 1 is the distance expression formula of target and radar under positive side-looking Desired Track; Section 2 is the directions of rays kinematic error having apart from space-variant; Section 3 is without the directions of rays kinematic error apart from space-variant.Make kinematic error R _errfor

$R_{\mathrm{err}}=R_{\mathrm{err}\_\mathrm{var}}+R_{\mathrm{err}\_\mathrm{fix}}$

$=\frac{({\mathrm{vt}}_a-x_n)}{R_B}\mathrm{\Δx}\left(t_a\right)+[-(\mathrm{\Δy}\left(t_a\right)\sin \mathrm{\β}+\mathrm{\Δz}\left(t_a\right)\cos \mathrm{\β})]---\left(28\right)$

Wherein, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant, R _{err_fix}for the directions of rays kinematic error without apart from space-variant.Kinematic error R _errcan solve by the method for estimating Doppler parameter.In the time receiving full aperture data, can obtain one group of length by step is above that the carrier aircraft of sub-aperture number M is along the speed v of course-and-bearing _{sub_all}=[v _sub1, v _sub2... v _subi... v _subM] (wherein, v _subibe the speed along course-and-bearing corresponding to sub-aperture, i orientation) and one group of acceleration a that length is the normal plane inner rays direction perpendicular to course line of sub-aperture number M _{sub_all}=[a _sub1, a _sub2... a _subi... a _subM] (wherein, a _subibe the acceleration of the normal plane inner rays direction perpendicular to course line corresponding to sub-aperture, i orientation).

There is the directions of rays kinematic error R apart from space-variant _{err_var}method for solving as follows:

Doppler frequency wherein θ is radar angle of squint, v _lOSfor along directions of rays speed.The differentiate about the time is carried out in above formula two ends, can obtain in the case of not considering to exist perpendicular to the acceleration of the normal plane inner rays direction in course line, can be obtained by formula (25) simultaneous above two about f _drexpression formula, can obtain the v of this formula estimates above the carrier aircraft that the obtains speed v along course-and-bearing _{sub_all}, by cubic spline interpolation algorithm by v _{sub_all}be interpolated into the vector v of the length of counting of full aperture _all, the bias vector between itself and speed average is

v _err＝v _all-mean(v _all) (29)

Wherein mean () is the function of averaging.Have apart from the vector acceleration of the directions of rays of space-variant and be

$a_{R\_\mathrm{var}}=-\frac{v_{\mathrm{err}}^2}{R_B}---\left(30\right)$

So there is the directions of rays kinematic error R apart from space-variant _{err_va}r can pass through a _{r_var}quadratic integral obtain

R _{err_fix}＝∫∫a _{err_fix} (31)。

Without the directions of rays kinematic error R apart from space-variant _{err_fix}method for solving as follows:

By cubic spline interpolation algorithm, the vector acceleration a of the normal plane inner rays direction perpendicular to course line is interpolated into the vectorial a of the length of counting of full aperture _{r_fix}, the bias vector between itself and acceleration average is

a _{err_fix}＝a _{R_fix}-mean(a _{R_fix}) (32)

So without the directions of rays kinematic error R apart from space-variant _{err_fix}can pass through a _{err_fix}quadratic integral obtain

R _{err_fix}＝∫∫a _{err_fix}(33)

The directions of rays resultant acceleration vector that can be obtained carrier aircraft by formula (30) is

$a_{R\_\mathrm{sum}}=-\frac{v_{\mathrm{azi}\_\mathrm{sum}}^2}{R_s}---\left(34\right)$

Wherein, a _{r_sum}=a _{r_var}+ a _{r_fix}, v _{azi_sum}represent the resultant velocity vector of the course-and-bearing of carrier aircraft, so the average velocity scalar that closes of the course-and-bearing of carrier aircraft is

$v_{\mathrm{azi}}=\mathrm{mean}\left(v_{\mathrm{azi}\_\mathrm{sum}}\right)=\mathrm{mean}\left(\sqrt{-a_{R\_\mathrm{sum}}{R}_s}\right)---\left(35\right)$

The doppler centroid of full aperture is

$f_{\mathrm{dc}0}=\mathrm{mean}\left({\hat{f}}_{\mathrm{dc}}\right)---\left(36\right)$

Wherein, for the doppler centroid in aperture, each side seat.

The practical flight speed of carrier aircraft is

$v_0=\sqrt{v_{\mathrm{LOS}}^2+v_{\mathrm{azi}}^2}=\sqrt{{\left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{2}\right)}^2+v_{\mathrm{azi}}^2}---\left(37\right)$

The angle of squint of radar is

${\mathrm{\θ}}_0=\arcsin \left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{{2v}_0}\right)---\left(38\right)$

The doppler frequency rate of full aperture is

$f_{\mathrm{dr}0}=-\frac{2v_0^2\cos {\mathrm{\θ}}_0}{{\mathrm{\λR}}_S}---\left(39\right)$

So far completed full aperture doppler frequency rate f _dr0, full aperture doppler centroid f _dc0, there is the directions of rays kinematic error R apart from space-variant _{err_var}with the directions of rays kinematic error R without apart from space-variant _{err_fix}estimation.

The 4th step, aperture, envelope cancellation module the other side seat data are carried out envelope cancellation

While carrying out envelope cancellation respectively for M sub-aperture data, in order to solve the impact apart from mutability, first sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, then carry out envelope cancellation to each apart from piece, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance.

First the distance blocks of data of sub-orientation aperture data is carried out to distance dimension Fourier transform, be then multiplied by envelope correction factor H apart from frequency domain ₃, then carry out distance dimension inverse Fourier transform.

Envelope correction factor is

$H_3=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_{\mathrm{Bi}}}\·R_{\mathrm{err}\_\mathrm{var}})\right)---\left(40\right)$

Wherein, R _{err_fix}without the directions of rays kinematic error apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant, R _sfor line-of-sight distance corresponding to full aperture data center, R _bifor line-of-sight distance corresponding to current distance piece.

The 5th step, distance dimension piecemeal processing module is carried out after distance dimension piecemeal full aperture data, then goes respectively Doppler center, phase compensation, orientation pulse pressure, looks processing more and quantize output.

According to the processing power of hardware platform, the data of full aperture are divided into N apart from blocks of data in advance by distance dimension, go Doppler center, phase compensation, orientation pulse pressure, look processing more apart from piece each, finally carry out entirety and quantize, just can obtain a width SAR image.

For each each apart from bar data apart from piece, be multiplied by respectively Doppler center factor H in time domain ₄with phase compensating factor H ₅, complete Doppler center and phase compensation,

Go the Doppler center factor to be

H ₄＝exp(-j2πf _dc0·t _a) (41)

Wherein, f _dc0for the doppler centroid of full aperture, t _afor the slow time, represent the azimuth dimension time;

Phase compensating factor is

$H_5=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_i}\·R_{\mathrm{err}\_\mathrm{var}})\right)---\left(42\right)$

Wherein, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece, R _{err_fix}for the directions of rays kinematic error without apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant.

Orientation pulse pressure completes at orientation frequency domain, each apart from the advanced row azimuth dimension of bar data Fourier transform, is multiplied by orientation pulse compression match filter H ₆, then carry out azimuth dimension inverse Fourier transform.

Pulse compression match filter in orientation is

$H_6=\exp \left(\mathrm{j\π}\frac{{f_a}^2}{(R_s/R_i)\·f_{\mathrm{dr}0}}\right)---\left(43\right)$

Wherein, f _afor orientation frequency coordinate, f _dr0for the doppler frequency rate of full aperture, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece.

Depending on the process of processing be how: establish and imaging data is carried out to L look processing, exactly every the data of azimuth dimension L point is averaging, carry out L to look after processing, azimuth dimension is counted and become M/L point from M point.

The formula of image quantization is

$x=\frac{x_{\mathrm{ori}}}{\mathrm{mean}\left(x_{\mathrm{ori}}\right)}\·p---\left(44\right)$

Wherein, x _orirepresent the data before quantizing, x represents the image after quantification, and p is quantization parameter.

So far, completed the SAR real time imagery processing based on piecemeal processing.

This method can complete the processing of SAR real time imagery, compared with the general formation method based on full aperture processing, this method greatly reduces the requirement of Installed System Memory, can effectively solve the problem that causes picture quality variation apart from space-variant, be particularly useful for the SAR Real Time Image System higher to requirement of real-time.

The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any amendment of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (10)

1. the SAR real time imagery method based on piecemeal processing, is characterized in that, comprises the steps:

Step 1:SAR launches linear FM signal, receives each apart from go forward side by side horizontal pulse compression of echo data;

Step 2: when the distance echo data after pulse compression reaches the data volume in sub-aperture, an orientation, the distance echo data after paired pulses compression carries out range migration correction, Doppler parameter and kinematic parameter and estimates;

Step 3: in the time that the echo data receiving reaches the data volume of a full aperture, Doppler parameter and kinematic parameter to aperture, each side seat carry out Conjoint Analysis estimation, obtain full aperture Doppler parameter and kinematic error;

Step 4: sub-aperture, each orientation data are carried out respectively to envelope cancellation according to full aperture Doppler parameter and kinematic error;

Step 5: the full aperture data after envelope cancellation are carried out to distance dimension piecemeal, carry out phase compensation, orientation pulse pressure, look to process and quantize more and export apart from piece each.

2. a kind of SAR real time imagery method based on piecemeal processing according to claim 1, is characterized in that, in step 1, the signal of SAR transmitting is linear FM signal, and its expression formula is

$s\left(t_r\right)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \left\{\mathrm{j\π}{K}_r{t_r}^2\right\}$

Wherein, t _rfor the fast time, representative is apart from dimension time, T _pthe indicating impulse duration, K _rthe frequency modulation rate that represents chirp pulse signal, the impulse response of its corresponding matched filter is

$h\left(t_r\right)=s^{*}(-t_r)=\mathrm{rect}\left[\frac{t_r}{T_p}\right]\exp \{-\mathrm{j\π}{K}_r{t_r}^2\}$

Apart from echo pulse pressure module, each receiving passed through to its matched filter apart from echo data, obtain apart from the data after pulse pressure, realize apart from high-resolution.

3. a kind of SAR real time imagery method based on piecemeal processing according to claim 1, is characterized in that, the concrete steps of step 2 are:

Step 2.1: in advance full aperture is divided into sub-aperture, M orientation by azimuth dimension according to the processing power of hardware platform;

Step 2.2: in the time that the echo data receiving reaches the data volume in sub-aperture, an orientation, according to the processing power of hardware platform, the data in aperture, the party seat are carried out to the processing of distance dimension piecemeal, obtain several apart from piece;

Step 2.3: carry out range migration correction to each apart from piece, estimate the doppler centroid f in sub-aperture, orientation _dcwith doppler frequency rate f _dr;

Step 2.4: estimate that according to each Doppler parameter apart from piece in aperture, the party seat the carrier aircraft in sub-aperture, orientation is along the speed v of course-and-bearing _subacceleration a with the normal plane inner rays direction perpendicular to course line _sub.

4. a kind of SAR real time imagery method based on piecemeal processing according to claim 3, it is characterized in that, the process of carrying out range migration correction in described step 2.3 is: sub-orientation aperture data are carried out to azimuth dimension Fourier transform and distance dimension Fourier transform, be multiplied by respectively range migration correction factor H in two-dimensional frequency ₁with second pulse compressibility factor H ₂, then carry out distance dimension inverse Fourier transform and azimuth dimension inverse Fourier transform, range migration correction factor H ₁with second pulse compressibility factor H ₂expression formula as shown in the formula

$\begin{array}{cc}{H}_1=\exp \left(j2\mathrm{\π}\frac{R_i}{C}{\left(\frac{f_d}{f_{\mathrm{dM}}}\right)}^2{f}_r\right)& H_2=\exp (-\mathrm{j\π}\frac{{2\mathrm{\λR}}_i{(f_d/f_{\mathrm{dM}})}^2{f_r}^2}{C^2{\left(\sqrt{1-{(f_d/f_{\mathrm{dM}})}^2}\right)}^3})\end{array}$

Wherein, R _irepresent i the line-of-sight distance corresponding apart from piece, C represents the light velocity, equals 3 × 10 ⁸meter per second, for echo Doppler, v is carrier aircraft actual measurement speed, and λ is carrier wavelength, and θ is actual measurement radar angle of squint, for being positioned at the echo Doppler of carrier aircraft dead ahead point target, f _rfor distance dimension frequency coordinate;

Wherein, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance.

5. a kind of SAR real time imagery method based on piecemeal processing according to claim 3, is characterized in that the specific implementation of step 3:

Step 3.1: by the doppler centroid f in each sub-aperture, orientation _dccalculate the doppler centroid f of full aperture _dc0, f _dc0=mean (f _dc);

Step 3.2: the speed v according to carrier aircraft along course-and-bearing _{sub_all}=[v _sub1, v _sub2... v _subi... v _subM] (wherein, v _subibe the speed along course-and-bearing corresponding to sub-aperture, i orientation) and perpendicular to the acceleration a of the normal plane inner rays direction in course line _{sub_all}=[a _sub1, a _sub2... a _subi... a _subM], calculate the vector acceleration a having apart from the directions of rays of space-variant _{r_var}vector acceleration a with the directions of rays without apart from space-variant _{r_fix}, computing formula as shown in the formula

$a_{R\_\mathrm{var}}=-\frac{v_{\mathrm{sub}\_\mathrm{all}}^2}{R_B}$

Wherein, R _bfor the vertical range in target and course line,

By cubic spline interpolation algorithm by the vector acceleration a of the normal plane inner rays direction perpendicular to course line _{sub_all}be interpolated into the length of counting of full aperture, obtain the vector acceleration a without the directions of rays apart from space-variant _{r_fix};

Step 3.3: calculate the normal plane inner rays direction resultant acceleration vector a of carrier aircraft perpendicular to course line _{r_sum}=a _{r_var}+ a _{r_fix};

Step 3.4: the carrier aircraft practical flight speed and the v that calculate full aperture ₀with radar angle of squint θ ₀,

$v_0=\sqrt{v_{\mathrm{LOS}}^2+v_{\mathrm{azi}}^2}=\sqrt{{\left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{2}\right)}^2+v_{\mathrm{azi}}^2}$

$v_{\mathrm{azi}}=\mathrm{mean}\left(v_{\mathrm{azi}\_\mathrm{sum}}\right)=\mathrm{mean}\left(\sqrt{{-a}_{R\_\mathrm{sum}}{R}_s}\right)$

${\mathrm{\θ}}_0=\arcsin \left(\frac{{\mathrm{\λf}}_{\mathrm{dc}0}}{{2v}_0}\right)$

Wherein, λ is carrier wavelength, f _dc0for the doppler centroid of full aperture, v _azicarrier aircraft is closed average velocity scalar, a along course-and-bearing _{r_sum}for carrier aircraft is perpendicular to the normal plane inner rays direction resultant acceleration vector in course line, R _sfor line-of-sight distance corresponding to full aperture data center;

Step 3.5: according to airborne practical flight speed v ₀with radar angle of squint θ ₀calculate the doppler frequency rate f of full aperture _dr0,

$d_{\mathrm{dr}0}=\frac{{2v}_0^2\cos {\mathrm{\θ}}_0}{{\mathrm{\λR}}_s}$

Wherein λ is carrier wavelength, R _sfor the line-of-sight distance of full aperture data center;

Step 3.6: calculate the directions of rays kinematic error R having apart from space-variant _{err_var}with the directions of rays kinematic error R without apart from space-variant _{err_fix};

R _{err_var}＝∫∫a _{R_var}

a _{err_fix}＝a _{R_fix}-mean(a _{R_fix})

R _{err_fix}＝∫∫a _{err_fix}

Wherein, a _{r_var}for thering is the vector acceleration apart from the directions of rays of space-variant, a _{r_fix}for the vector acceleration of the directions of rays without apart from space-variant, its a _{err_fix}for a _{r_fix}and the bias vector between acceleration average.

6. a kind of SAR real time imagery method based on piecemeal processing according to claim 4, is characterized in that, the specific implementation of step 4 is:

Step 4.1: sub-aperture, each orientation data are carried out to the processing of distance dimension piecemeal, obtain several apart from piece;

Step 4.2: carry out envelope cancellation to each apart from piece, each apart from line-of-sight distance corresponding to piece according to adjusting apart from distance,

Wherein, each while carrying out envelope cancellation apart from piece, advanced row distance dimension Fourier transform, then be multiplied by envelope correction factor H apart from frequency domain ₃, then carry out distance dimension inverse Fourier transform,

Wherein, $H_3=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_{\mathrm{Bi}}}\·R_{\mathrm{err}\_\mathrm{var}})\right)$

Wherein, R _{err_fix}without the directions of rays kinematic error apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant, R _sfor line-of-sight distance corresponding to full aperture data center, R _bifor line-of-sight distance corresponding to current distance piece.

7. a kind of SAR real time imagery method based on piecemeal processing according to claim 1, is characterized in that, the specific implementation of step 5 is:

Step 5.1: in advance the data of full aperture are divided into N apart from piece by distance dimension according to the processing power of hardware platform;

Step 5.2: go Doppler center, phase compensation, orientation pulse pressure, look processing more apart from piece each, obtain full aperture data x after treatment _ori;

Step 5.3: by the full aperture data x handling _oricarry out entirety and quantize, obtain a width SAR image.

8. a kind of SAR real time imagery method based on piecemeal processing according to claim 7, is characterized in that, in step 5.2 to each apart from piece go Doppler center, phase compensation, orientation pulse pressure, look more process and process that entirety quantizes as follows:

A) eachly remove being embodied as of Doppler center apart from piece, each distance bar data apart from piece, are multiplied by respectively Doppler center factor H in time domain ₄, H ₄=exp (j2 π f _dc0t _a),

Wherein, f _dc0for the doppler centroid of full aperture, t _afor the slow time, represent the azimuth dimension time;

B) eachly carry out the each distance bar data apart from piece of being embodied as of phase compensation apart from piece, be multiplied by respectively phase compensating factor H in time domain ₅,

$H_5=\exp \left(j\frac{4\mathrm{\π}}{\mathrm{\λ}}(R_{\mathrm{err}\_\mathrm{fix}}+\frac{R_s}{R_i}\·R_{\mathrm{err}\_\mathrm{var}})\right),$

Wherein, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece, R _{err_fix}for the directions of rays kinematic error without apart from space-variant, R _{err_var}for thering is the directions of rays kinematic error apart from space-variant;

C) eachly carry out being embodied as of orientation pulse pressure apart from piece, each distance bar data apart from piece, carry out azimuth dimension Fourier transform in territory, orientation, are multiplied by orientation pulse compression match filter H ₆, then carry out azimuth dimension inverse Fourier transform,

$H_0=\exp \left(\mathrm{j\π}\frac{{f_a}^2}{(R_s/R_i)\·f_{\mathrm{dr}0}}\right),$

Wherein, f _afor orientation frequency coordinate, f _dr0for the doppler frequency rate of full aperture, R _sfor line-of-sight distance corresponding to full aperture data center, R _irepresent i the line-of-sight distance corresponding apart from piece;

D) eachly carry out looking being embodied as of processing apart from piece, each distance bar data apart from piece, are averaging every the data of azimuth dimension L point, carry out L to look after processing more, and azimuth dimension is counted and become M/L point from M point.

9. a kind of SAR real time imagery method based on piecemeal processing according to claim 7, is characterized in that, in step 5.3 by the full aperture data x handling _orithe formula that carries out entirety quantification is as follows,

$x=\frac{x_{\mathrm{ori}}}{\mathrm{mean}\left(x_{\mathrm{ori}}\right)}\·p$

Wherein, x _orirepresent the data before quantizing, x represents the image after quantification, and p is quantization parameter.

10. realize in claim 1-9 a kind of system of the SAR real time imagery method based on piecemeal processing described in any one for one kind, it is characterized in that, comprise apart from echo pulse pressure module, azimuth dimension piecemeal processing module, full aperture parameter estimation module, envelope cancellation module and distance dimension piecemeal processing module;

Described apart from echo pulse pressure module, it is for carrying out pulse compression by each receiving apart from echo data;

Described azimuth dimension piecemeal processing module, it carries out range migration correction, Doppler parameter and kinematic parameter for aperture, the other side seat data and estimates;

Described full aperture parameter estimation module, it,, for the parameter in aperture, each side seat is carried out to Conjoint Analysis estimation, obtains full aperture Doppler parameter and kinematic error;

Described envelope cancellation module, it carries out envelope cancellation for aperture, the other side seat data;

Described distance dimension piecemeal processing module, it carries out phase compensation, orientation pulse pressure, looks to process and quantize more and export for the data to after envelope cancellation.