CN103323828A - Ultrahigh-resolution spaceborne SAR imaging processing method and device - Google Patents

Abstract

The invention discloses an ultrahigh-resolution spaceborne SAR imaging processing method. The ultrahigh-resolution spaceborne SAR imaging processing method comprises the following steps of establishing a two-dimension frequency spectrum model of a spaceborne SAR echo signal, eliminating a range-migration phase item and a high-order coupling phase item which are arranged at the position of the reference slope distance of the two-dimension frequency spectrum model through multiplying the two-dimension frequency spectrum model by a reference function which corresponds to the two-dimension frequency spectrum model, conducting distance-direction reverse Fourier transform on the processed signal with the range migration phase item and the high-order coupling phase item are eliminated, correcting the residual distance migration quantity of the two-dimension frequency spectrum model in an interpolation mode, compensating the rest of high-order coupling phase items of the two-dimension frequency spectrum model in a distance blocking compensation mode, eliminating an orientation modulation phase item of the two-dimension frequency spectrum model in an orientation compression mode, and finally obtaining a spaceborne SAR image after focusing. In addition, the invention further discloses an ultrahigh-resolution spaceborne SAR imaging processing device. Through the ultrahigh-resolution spaceborne SAR imaging processing method and device, the high-quality ultrahigh-resolution spaceborne SAR image can be obtained. The high-quality ultrahigh-resolution spaceborne SAR image has a good locating capability and meets the application demand under the condition of high resolution.

Description

Ultrahigh resolution satellite-borne SAR image processing method and device

Technical field

The present invention relates to ultrahigh resolution satellite-borne synthetic aperture radar (SAR, Synthetic Aperture Radar) imaging processing technology, be specifically related to a kind of ultrahigh resolution satellite-borne SAR image processing method and device.

Background technology

Technology such as satellite-borne SAR comprehensive utilization synthetic aperture technique, pulse compression and data processing adopt short antenna just can access higher image resolution ratio, are a kind of high-resolution microwave side-looking imaging radars round-the-clock, the round-the-clock characteristic that has.

Based on fast development and the widespread use of satellite-borne SAR technology, the low rail satellite-borne SAR of high resolving power has proposed more harsh requirement to image processing method.Traditional satellite-borne SAR equivalence oblique distance model has utilized straight path to come analog orbit, and the curved in tracks characteristic of spaceborne platform only is described by second order satellite-borne SAR Doppler parameter, so just cause the precision of traditional satellite-borne SAR oblique distance model limited, can not satisfy satellite-borne SAR imaging processing under the ultrahigh resolution to the requirement of oblique distance model, restrict the development of High Resolution Spaceborne SAR technology.

High-precision image processing method is the basis that obtains the high-quality radar image, present image processing method, become mark algorithm and Wavenumber Domain Algorithms majority based on the equivalent oblique distance model of tradition such as distance range and Doppler, frequency modulation, mostly there not be to consider or only partly considered the imaging processing to the satellite-borne SAR warp rail, so, just make data can not well focussed, make picture quality sharply descend, and then do not satisfy the requirement of high-resolution applications.

Summary of the invention

In view of this, fundamental purpose of the present invention is to provide a kind of ultrahigh resolution satellite-borne SAR image processing method and device, can satisfy the demand of the high-resolution imaging of satellite-borne SAR.

For achieving the above object, technical scheme of the present invention is achieved in that

The invention provides a kind of ultrahigh resolution satellite-borne synthetic aperture radar SAR image processing method, described method comprises:

Set up the 2-d spectrum model of satellite-borne SAR echoed signal;

Multiply each other by described 2-d spectrum model and its corresponding reference function and to eliminate range migration phase term and the high-order coupling phase term at reference oblique distance place of described 2-d spectrum model, again to making distance to inverse Fourier transform through the signal of eliminating after handling, and proofread and correct residual range migration amount in the described 2-d spectrum model by interpolation method, by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compress mode again;

According to the SAR image after eliminating described 2-d spectrum model behind the orientation phase modulation item and obtaining focusing on.

In the such scheme, before the described 2-d spectrum model of setting up the satellite-borne SAR echoed signal, described method also comprises:

Calculate the state vector of satellite and terrain object;

Calculate satellite-borne SAR high-order Doppler parameter according to described state vector;

According to described high-order Doppler parameter, set up satellite-borne SAR oblique distance model;

According to the slip beam bunching mode of described satellite-borne SAR state vector and setting, generate the SAR echoed signal, and the orientation frequency spectrum that utilizes the orientation preconditioning technique to remove described echoed signal blurs.

In the such scheme, described interpolation method comprises: Singh Sinc interpolation method.

In the such scheme, the described 2-d spectrum model of setting up the satellite-borne SAR echoed signal comprises:

According to satellite-borne SAR oblique distance model and the progression inversion method set up, set up the 2-d spectrum model of satellite-borne SAR echoed signal: wherein, described 2-d spectrum model Φ _PTRS(r ₀, f _r, f _η) expression formula be:

${\mathrm{\Φ}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\η}})\≈-\frac{4\mathrm{\π}(f_c+f_r)}{c}(r_0-\frac{M^2}{2}{\mathrm{\α}}_0-\frac{M^3}{3}{\mathrm{\β}}_0-\frac{M^4}{4}{\mathrm{\γ}}_0)-\frac{\mathrm{\π}{f}_r^2}{K_r};$

The following formula series expansion is obtained:

${\mathrm{\Φ}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\η}})={\mathrm{\Φ}}_{\mathrm{azi}}(r_0,f_{\mathrm{\η}})+{\mathrm{\Φ}}_{\mathrm{RCM}}(r_0,f_{\mathrm{\η}})\·f_r+{\mathrm{\Φ}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\η}})\·f_r^2+{\mathrm{\Φ}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\η}})\·f_r^3;$

Wherein, ≈ represents to be similar to, $M=-\frac{c{f}_{\mathrm{\η}}}{2(f_c+f_r)}+\frac{\mathrm{\λ}{f}_{\mathrm{dc}}}{2},$ ${\mathrm{\α}}_0=\frac{1}{2k_2},$ ${\mathrm{\β}}_0=\frac{3k_3}{8k_2^3},$ ${\mathrm{\γ}}_0=\frac{9k_3^2-4k_2{k}_4}{16k_2^5},$ C is the light velocity, f _cBe carrier frequency, f _rFor the distance to frequency axis, f _ηFor the orientation to frequency axis, K _rFor the distance to frequency modulation rate, r ₀Be the nearest oblique distance of radar to terrain object, f _DcBe satellite Doppler center, f _1rBe doppler frequency rate, f _2rBe second order doppler frequency rate, f _3rBe three rank doppler frequency rates, Φ _Azi(r ₀, f _η) be orientation phase modulation item, Φ _RCM(r ₀, f _η) f _rBe the range migration phase term,

Be the second order coupled item,

Be three rank coupling terms, described high-order coupling phase term comprises: second order coupled item and three rank coupling terms.

In the such scheme, described 2-d spectrum model corresponding reference function H _RFM(r _Ref, f _r, f _η) expression formula be:

$H_{\mathrm{RFM}}(r_{\mathrm{ref}},f_r,f_{\mathrm{\η}})=\exp \{j\frac{4\mathrm{\π}(f_c+f_r)}{c}(r_{\mathrm{ref}}-\frac{M^2}{2}{\mathrm{\α}}_{\mathrm{ref}}-\frac{M^3}{3}{\mathrm{\β}}_{\mathrm{ref}}-\frac{M^4}{4}{\mathrm{\γ}}_{\mathrm{ref}})+j\frac{\mathrm{\π}{f}_r^2}{K_r}\};$

Wherein, it is the exponential function of the truth of a matter that exp{} represents with e, and j represents imaginary unit; r _RefBe reference oblique distance, α _Ref, β _Ref, γ _RefOblique distance model parameter for reference oblique distance place.

In the such scheme, describedly proofread and correct residual range migration amount in the described 2-d spectrum model by interpolation method, comprising:

Described 2-d spectrum model is made distance behind inverse fourier transform, utilize the Sinc interpolation method to proofread and correct residual range migration amount in the described 2-d spectrum model; Wherein, the expression formula of described residual range migration amount is:

${\mathrm{RCM}}_{\mathrm{diff}}(r_0,f_{\mathrm{\η}})=-\frac{{\mathrm{\λ}}^2[{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2+2f_{\mathrm{dc}}(f_{\mathrm{\η}}-f_{\mathrm{dc}})]}{8}({\mathrm{\α}}_0-{\mathrm{\α}}_{\mathrm{ref}})$

$+\frac{{\mathrm{\λ}}^3[2{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3+3f_{\mathrm{dc}}{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2}{24}({\mathrm{\β}}_0-{\mathrm{\β}}_{\mathrm{ref}})$

$-\frac{{\mathrm{\λ}}^4[3{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^4+4f_{\mathrm{dc}}{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3]}{64}({\mathrm{\γ}}_0-{\mathrm{\γ}}_{\mathrm{ref}}).$

In the such scheme, described by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, comprising:

The signal that residual range migration amount in the described 2-d spectrum model is done to proofread and correct after handling is made distance to Fourier transform, multiply each other with the phase compensation expression formula again; Described phase compensation expression formula is:

Φ _comp(r _Mn，f _r，f _η)＝[Φ _PTRS(r _Mn，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r _Mn，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r]

-[Φ _azi(r _Mn，f _η)-Φ _azi(r _ref，f _η)]

Wherein, r _MnBe the individual center oblique distance apart from ion block of n, n is positive integer, Φ _RCM(r _Mn, f _η) f _rFor at r _MnThe range migration phase term of point, Φ _Azi(r _Ref, f _η) be at r _RefThe orientation phase modulation item of point.

In the such scheme, describedly eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compress mode, comprising:

Signal and the orientation matched filter of adjusting the distance behind the branch block compensation multiply each other, and the result after multiplying each other is made the orientation to inverse Fourier transform; Wherein, the expression formula of described orientation matched filter is:

$H_{\mathrm{azi}}(r_0,f_{\mathrm{\η}})=\exp \{-j\frac{4\mathrm{\π}}{\mathrm{\λ}}[\frac{{\mathrm{\λ}}^2{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2}{8}({\mathrm{\α}}_0-{\mathrm{\α}}_{\mathrm{ref}})-\frac{{\mathrm{\λ}}^3{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3}{24}({\mathrm{\β}}_0-{\mathrm{\β}}_{\mathrm{ref}})+\frac{{\mathrm{\λ}}^4{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^4}{64}({\mathrm{\γ}}_0-{\mathrm{\γ}}_{\mathrm{ref}})\left]\right\};$

Wherein, λ is the SAR radar wavelength.

The present invention also provides a kind of ultrahigh resolution satellite-borne SAR imaging processing device, and described device comprises: set up unit, first and eliminate unit, converter unit, correcting unit, compensating unit, second elimination unit and the image-generating unit; Wherein,

The described unit of setting up is for the 2-d spectrum model of setting up the satellite-borne SAR echoed signal;

Described first eliminates the unit, is used for by described 2-d spectrum model and its corresponding reference function are multiplied each other, and eliminates range migration phase term and the high-order coupling phase term at reference oblique distance place of described 2-d spectrum model;

Described converter unit is used for the described 2-d spectrum model after the described first elimination cell processing is made distance to inverse Fourier transform;

Described correcting unit is used for utilizing interpolation method to proofread and correct the residual range migration amount of the described 2-d spectrum model after described converter unit is handled;

Described compensating unit is used for utilizing the residue high-order coupling phase term that compensates the described 2-d spectrum model after described correcting unit is handled apart from the piecemeal compensation way;

Described second eliminates the unit, is used for eliminating by the orientation compress mode orientation phase modulation item of the described 2-d spectrum model after described compensating unit is handled;

Described image-generating unit is for the SAR image after obtaining focusing on according to the described 2-d spectrum model after the described second elimination cell processing.

In the such scheme, described correcting unit is by utilizing the residual range migration amount in the described 2-d spectrum model after Singh Sinc interpolation method is proofreaied and correct described converter unit processing.

In the such scheme,

The expression formula of described 2-d spectrum model is:

${\mathrm{\Φ}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\η}})\≈-\frac{4\mathrm{\π}(f_c+f_r)}{c}(r_0-\frac{M^2}{2}{\mathrm{\α}}_0-\frac{M^3}{3}{\mathrm{\β}}_0-\frac{M^4}{4}{\mathrm{\γ}}_0)-\frac{\mathrm{\π}{f}_r^2}{K_r};$

Above-mentioned series expansion is obtained:

${\mathrm{\Φ}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\η}})={\mathrm{\Φ}}_{\mathrm{azi}}(r_0,f_{\mathrm{\η}})+{\mathrm{\Φ}}_{\mathrm{RCM}}(r_0,f_{\mathrm{\η}})\·f_r+{\mathrm{\Φ}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\η}})\·f_r^2+{\mathrm{\Φ}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\η}})\·f_r^3;$

Described 2-d spectrum model corresponding reference function H _RFM(r _Ref, f _r, f _η) expression formula be:

$H_{\mathrm{RFM}}(r_{\mathrm{ref}},f_r,f_{\mathrm{\η}})=\exp \{j\frac{4\mathrm{\π}(f_c+f_r)}{c}(r_{\mathrm{ref}}-\frac{M^2}{2}{\mathrm{\α}}_{\mathrm{ref}}-\frac{M^3}{3}{\mathrm{\β}}_{\mathrm{ref}}-\frac{M^4}{4}{\mathrm{\γ}}_{\mathrm{ref}})+j\frac{\mathrm{\π}{f}_r^2}{K_r}\};$

Wherein, ≈ represents to be similar to, $M=-\frac{c{f}_{\mathrm{\η}}}{2(f_c+f_r)}+\frac{\mathrm{\λ}{f}_{\mathrm{dc}}}{2},$ ${\mathrm{\α}}_0=\frac{1}{2k_2},$ ${\mathrm{\β}}_0=-\frac{3k_3}{8k_2^3},$ ${\mathrm{\γ}}_0=\frac{9k_3^2-4k_2{k}_4}{16k_2^5},$ C is the light velocity, f _cBe carrier frequency, f _rFor the distance to frequency axis, f _ηFor the orientation to frequency axis, K _rFor the distance to frequency modulation rate, r ₀Be the nearest oblique distance of radar to terrain object, f _DcBe satellite Doppler center, f _1rBe doppler frequency rate, f _2rBe second order doppler frequency rate, f _3rBe three rank doppler frequency rates, Φ _Azi(r ₀, f _η) be orientation phase modulation item, Φ _RCM(r ₀, f _η) f _rBe the range migration phase term,

Be the second order coupled item, Be three rank coupling terms, it is the exponential function of the truth of a matter that exp{} represents with e, and j represents imaginary unit, r _RefBe reference oblique distance, α _Ref, β _Ref, γ _RefOblique distance model parameter for reference oblique distance place.

In the such scheme, the described unit of setting up comprises: first computation subunit, second computation subunit, second set up subelement, generate subelement, remove subelement and the 3rd sets up subelement; Wherein,

Described first computation subunit is for the state vector of calculating satellite and terrain object;

Described second computation subunit is used for according to described state vector, calculates satellite-borne SAR high-order Doppler parameter;

Described second sets up subelement, is used for according to the high-order Doppler parameter, sets up satellite-borne SAR oblique distance model;

Described generation subelement is used for the slip beam bunching mode according to satellite-borne SAR state vector and setting, generates the SAR echoed signal;

Described removal subelement, the orientation frequency spectrum that is used for utilizing the orientation preconditioning technique to remove described echoed signal blurs;

The described the 3rd sets up subelement, according to satellite-borne SAR oblique distance model and the progression inversion method set up, sets up the 2-d spectrum model of satellite-borne SAR echoed signal.

In the such scheme, the described 2-d spectrum model after described correcting unit is handled described converter unit carries out the correction of residual range migration amount by the Sinc interpolation method in distance Doppler territory; Wherein, described residual range migration scale is shown:

${\mathrm{RCM}}_{\mathrm{diff}}(r_0,f_{\mathrm{\η}})=-\frac{{\mathrm{\λ}}^2[{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2+2f_{\mathrm{dc}}(f_{\mathrm{\η}}-f_{\mathrm{dc}})]}{8}({\mathrm{\α}}_0-{\mathrm{\α}}_{\mathrm{ref}})$

$+\frac{{\mathrm{\λ}}^3[2{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3+3f_{\mathrm{dc}}{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2}{24}({\mathrm{\β}}_0-{\mathrm{\β}}_{\mathrm{ref}}).$

$-\frac{{\mathrm{\λ}}^4[3{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^4+4f_{\mathrm{dc}}{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3]}{64}({\mathrm{\γ}}_0-{\mathrm{\γ}}_{\mathrm{ref}})$

In the such scheme, described compensating unit comprises: the first varitron unit and the subelement that multiplies each other; Wherein,

The described first varitron unit is used for the described 2-d spectrum model after the described correcting unit processing is made distance to Fourier transform;

The described subelement that multiplies each other multiplies each other for described 2-d spectrum model and phase compensation expression formula after the described first varitron cell processing;

Wherein, described phase compensation expression formula is:

Φ _comp(r _Mn，f _r，f _η)＝[Φ _PTRS(r _Mn，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r _Mn，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r]

-[Φ _azi(r _Mn，f _η)-Φ _azi(r _ref，f _η)]；

Wherein, r _MnBe the individual center oblique distance apart from ion block of n, n is positive integer, Φ _RCM(r _Mn, f _η) be at r _MnThe residual range migration amount of point, Φ _Azi(r _Ref, f _η) be at r _RefThe orientation phase modulation item of point.

In the such scheme, described second eliminates the unit comprises: coupling subelement and the second varitron unit; Wherein,

Described coupling subelement is used for described 2-d spectrum model and orientation matched filter after described compensating unit is handled are multiplied each other;

The described second varitron unit is used for the described 2-d spectrum model after described coupling subelement is handled is made the orientation to inverse Fourier transform; Wherein, the expression formula of described orientation matched filter is:

$H_{\mathrm{azi}}(r_0,f_{\mathrm{\η}})=\exp \{-j\frac{4\mathrm{\π}}{\mathrm{\λ}}[\frac{{\mathrm{\λ}}^2{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^2}{8}({\mathrm{\α}}_0-{\mathrm{\α}}_{\mathrm{ref}})-\frac{{\mathrm{\λ}}^3{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^3}{24}({\mathrm{\β}}_0-{\mathrm{\β}}_{\mathrm{ref}})+\frac{{\mathrm{\λ}}^4{(f_{\mathrm{\η}}-f_{\mathrm{dc}})}^4}{64}({\mathrm{\γ}}_0-{\mathrm{\γ}}_{\mathrm{ref}})\left]\right\};$

Wherein, λ is the SAR radar wavelength.

Ultrahigh resolution satellite-borne SAR image processing method provided by the invention and device are at first set up the 2-d spectrum model of the satellite-borne SAR echoed signal under the ultrahigh resolution; In the imaging processing described 2-d spectrum model and its corresponding reference function are multiplied each other, eliminate described 2-d spectrum model at range migration phase term and the high-order coupling phase term at reference distance place, remake distance to inverse Fourier transform, proofread and correct residual range migration amount in the described 2-d spectrum model in distance Doppler territory by the Sinc interpolation method, by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compression again; According to the SAR image after eliminating described 2-d spectrum model behind the orientation phase modulation item and obtaining focusing on.Utilize technical scheme of the present invention, can obtain focusing on good High Resolution Spaceborne SAR image, and have good station-keeping ability, can satisfy the application demand under the high-resolution.

Description of drawings

Fig. 1 is the schematic flow sheet of ultrahigh resolution satellite-borne SAR image processing method of the present invention;

Fig. 2 is the synoptic diagram of a specific embodiment of ultrahigh resolution satellite-borne SAR image processing method of the present invention;

Fig. 3 is the synoptic diagram of another specific embodiment of ultrahigh resolution satellite-borne SAR image processing method of the present invention;

Fig. 4 (a) and (b) are the composition structural representation of ultrahigh resolution satellite-borne SAR imaging processing device of the present invention;

Fig. 5 (a) and (b) are focus characteristics synoptic diagram of the present invention.

Embodiment

The present invention has put down in writing a kind of ultrahigh resolution satellite-borne SAR image processing method, and as shown in Figure 1, described method comprises:

Step 11: the 2-d spectrum model of setting up the satellite-borne SAR echoed signal;

Here, described step 11 is finished by computing machine, and Fig. 2 is the specific embodiment synoptic diagram of step 11, and as shown in Figure 2, described step 11 comprises:

Step 111: set up spaceborne platform;

Concrete, set up track six roots of sensation number and the ground surface imaging scene of spaceborne platform;

Step 112: the state vector of under described spaceborne platform and earth inertial coordinates system, calculating satellite and terrain object; Described state vector comprises: the distance vector r of satellite _s, velocity v _s, acceleration A _s, single order acceleration A _s' and second order acceleration A _s"; The distance vector r of terrain object _t, velocity v _t, acceleration A _t, single order acceleration A _t' and second order acceleration A _t";

In this step, considered the relative motion between satellite and the terrain object, and used the driftage guidance technology and the beam position of radar and attitude are changed compensate, made the beam position always perpendicular to the flight path direction.

Step 113: according to described state vector, utilize formula (1) to (4) to calculate satellite-borne SAR high-order Doppler parameter;

Described satellite-borne SAR high-order Doppler parameter comprises: satellite Doppler center f _Dc, doppler frequency rate f _1r, second order doppler frequency rate f _2r, three rank doppler frequency rate f _3r

$f_{\mathrm{dc}}=-\frac{2R^{\text{'}}}{\mathrm{\λ}}=-\frac{2}{\mathrm{\λ}}\frac{(r_s-r_t)(v_s-v_t)}{R}---\left(1\right)$

$f_{1r}=-\frac{2R^{\text{'}}}{\mathrm{\λ}}=-\frac{2}{\mathrm{\λ}}[\frac{{(v_s-v_t)}^2}{R}+\frac{(A_s-A_t)(r_s-r_t)}{R}-\frac{{\mathrm{\λ}}^2{f}_{\mathrm{dc}}^2}{4R}]---\left(2\right)$

$f_{2r}=-\frac{2R^{\left(3\right)}}{\mathrm{\λ}}=-\frac{2}{\mathrm{\λ}}[\frac{3(v_s-v_t)(A_s-A_t)}{R}+\frac{(r_s-r_t)(A_s^{\text{'}}-A_t^{\text{'}})}{R}-\frac{3{\mathrm{\λ}}^2{f}_{\mathrm{dc}}{f}_{1r}}{4R}]---\left(3\right)$

$f_{3r}=-\frac{2R^{\left(4\right)}}{\mathrm{\λ}}=-\frac{2}{\mathrm{\λ}}[\frac{3{(A_s-A_t)}^2}{R}+\frac{4(v_s-v_t)(A_s^{\text{'}}-A_t^{\text{'}})}{R}+\frac{(r_s-r_t)(A_s^{\text{'}}-A_t^{\text{'}})}{R}-\frac{3{\mathrm{\λ}}^2{f}_{1r}^2}{4R}-\frac{{\mathrm{\λ}}^2{f}_{\mathrm{dc}}{f}_{2r}}{R}]---\left(4\right)$

Wherein, vector R, scalar R are known quantity, represent respectively that radar arrives the distance vector of terrain object and apart from scalar; λ is radar wavelength.

Step 114: according to described high-order Doppler parameter, set up the satellite-borne SAR oblique distance model under the ultrahigh resolution;

Here, utilize straight path to come analog orbit with traditional satellite-borne SAR equivalence oblique distance model, and the curved in tracks characteristic is only carried out part compensation difference by second order satellite-borne SAR Doppler parameter, the present invention considers that accurately curved in tracks is to the influence of satellite-borne SAR, according to the computing formula (1) of high-order Doppler parameter to (4), set up the satellite-borne SAR oblique distance model relevant with described high-order Doppler parameter, as shown in Equation (5):

R(η)＝r ₀+k ₁η+k ₂η ²+k ₃η ³+k ₄η ⁴ (5)

Wherein, r ₀Be the nearest oblique distance of radar to terrain object, η is that the orientation is to the time; Parameter k ₁～k ₄With the relation of above-mentioned high-order Doppler parameter as shown in Equation (6):

$k_1=-\frac{\mathrm{\&lambda;}{f}_{\mathrm{dc}}}{2},$ $k_2=-\frac{\mathrm{\&lambda;}{f}_{1r}}{4},$ $k_3=-\frac{\mathrm{\&lambda;}{f}_{2\mathrm{<figure-callout\; id="12"\; label="r"\; filenames="HDA00003267927700011.png"\; state="\{\{state\}\}">r</figure-callout>}}}{12},$ $k_4=-\frac{\mathrm{\&lambda;}{f}_{3r}}{48}---\left(6\right)$

Here, the radar that is calculated by satellite-borne SAR oblique distance model of the present invention is to the oblique distance of terrain object, compare based on radar to the oblique distance of terrain object that numerical method calculates with employing, error is very little, so can utilize satellite-borne SAR oblique distance model of the present invention to replace traditional satellite-borne SAR oblique distance model to carry out imaging processing.

Step 115: the state vector that obtains according to step 112, adopt described numerical method to calculate radar to the oblique distance of terrain object, and its mode of operation is set is the slip beam bunching mode, generates time domain SAR echoed signal, and the orientation frequency spectrum that utilizes the orientation preconditioning technique to remove under the slip beam bunching mode is fuzzy;

Here, therefore High Resolution Spaceborne SAR or wide the having relatively high expectations of ultrahigh resolution satellite-borne SAR the other side bit strip, adopt the slip beam bunching mode to obtain high orientation bandwidth among the present invention; Simultaneously, though the slip beam bunching mode has been avoided the requirement of high pulse repetition frequency, also brought the fuzzy problem of orientation frequency spectrum; Fuzzy for eliminating the orientation frequency spectrum, utilize this classical orientation preconditioning technique of Deramp method to handle.

Step 116: according to described satellite-borne SAR oblique distance model and progression inversion method, set up the 2-d spectrum model of satellite-borne SAR echoed signal;

Here, usually utilize principle in the phase bit that the orientation frequency-region signal is found the solution in the prior art, but can occur three rank polynomial expressions in the solution procedure and cause finding the solution difficulty, the present invention obtains site in the phasing by the method that adopts the progression inverting, thereby obtained described 2-d spectrum model, concrete:

At first, to carrying out distance based on the satellite-borne SAR echoed signal of oblique distance model shown in the formula (5) to Fourier transform (FFT, Fourier Transform), obtain following expression:

$S(f_r,\mathrm{\&eta;})=\exp \{-j\frac{4\mathrm{\&pi;}(f_c+f_r)R\left(\mathrm{\&eta;}\right)}{c}\}\exp \{-j\frac{\mathrm{\&pi;}{f}_r^2}{K_r}\}---\left(7\right)$

Wherein, c is the light velocity, f _cBe carrier frequency, f _rFor the distance to frequency axis, K _rFor the distance to the frequency modulation rate.

Make the orientation to FFT about the orientation to the formula of time η (7), obtain formula (8):

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})={\&Integral;}_{-\frac{T}{2}}^{\frac{T}{2}}\exp \{-\mathrm{j\&phi;}(r_0,f_r,\mathrm{\&eta;})\}\mathrm{d\&eta;}---\left(8\right)$

Wherein, $\mathrm{\&phi;}(r_0,f_r,\mathrm{\&eta;})=-\frac{\mathrm{\&pi;}{f}_r^2}{K_r}+\frac{4\mathrm{\&pi;}(f_c+f_r)(r_0+k_1\mathrm{\&eta;}+k_2{\mathrm{\&eta;}}^2+k_3{\mathrm{\&eta;}}^3+k_4{\mathrm{\&eta;}}^4)}{c}-2\mathrm{\&pi;}{f}_{\mathrm{\&eta;}}\mathrm{\&eta;}---\left(9\right)$

Wherein, f _ηFor the orientation to frequency axis; T is that the orientation is to the synthetic aperture time.

According to principle in the phase bit, should get The time η expression, but there are three rank polynomial expressions in the η expression formula after the differentiate, makes computation process very complicated.Therefore the present invention has adopted the progression inversion method to calculate the expression of η, and concrete computation process can repeat no more with reference to the explanation of progression inverting herein.Bring the result of calculation of progression inversion method into formula (9), and utilize the series expansion formula to obtain carrying out quadravalence expansion about M, thereby obtain described 2-d spectrum model representation formula, as shown in Equation (10):

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})\&ap;-\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_0-\frac{M^2}{2}{\mathrm{\&alpha;}}_0-\frac{M^3}{3}{\mathrm{\&beta;}}_0-\frac{M^4}{4}{\mathrm{\&gamma;}}_0)-\frac{\mathrm{\&pi;}{f}_r^2}{K_r}---\left(10\right)$

Wherein, ≈ represents to be similar to, $M=-\frac{c{f}_{\mathrm{\&eta;}}}{2(f_c+f_r)}+\frac{\mathrm{\&lambda;}{f}_{\mathrm{dc}}}{2},$ ${\mathrm{\&alpha;}}_0=\frac{1}{2k_2},$ ${\mathrm{\&beta;}}_0=-\frac{3k_3}{8k_2^3},$ ${\mathrm{\&gamma;}}_0=\frac{9k_3^2-4k_2{k}_4}{16k_2^5}.$

About the distance to frequency f _rFormula (10) is carried out the expansion of progression, the 2-d spectrum model of the satellite-borne SAR that obtains being similar to, as shown in Equation (11):

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})={\mathrm{\&Phi;}}_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})+{\mathrm{\&Phi;}}_{\mathrm{RCM}}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^2+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^3---\left(11\right)$

Wherein, Φ _Azi(r ₀, f _η) be orientation phase modulation item, Φ _RCM(r ₀, f _η) f _rBe the range migration phase term, ${\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^2$ For the second order coupled item, ${\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^3$ Be three rank coupling terms.

Step 12: eliminate described 2-d spectrum model by described 2-d spectrum model and its corresponding reference function are multiplied each other at range migration phase term and the high-order coupling phase term at reference oblique distance place, again to making distance to inverse Fourier transform through the signal of eliminating after handling, by interpolation method, be specially the residual range migration amount in the described 2-d spectrum model of Sinc interpolation correction, by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compression again; Obtain focusing on back SAR image according to the described 2-d spectrum model behind the elimination orientation phase modulation item.

Here, described step 12 is finished by computing machine; Fig. 3 is the specific embodiment synoptic diagram of described step 12, and as shown in Figure 3, described step 12 comprises:

Step 121: the reference function shown in the 2-d spectrum model shown in the formula (11) and the formula (12) is multiplied each other, eliminate range migration phase term and high-order coupling phase term at reference oblique distance place in the described 2-d spectrum model, but also had residue high-order coupling phase term and the residual range migration phase term at other oblique distance place;

The signal of eliminating after handling is made distance to inverse Fourier transform;

$H_{\mathrm{RFM}}(r_{\mathrm{ref}},f_r,f_{\mathrm{\&eta;}})=\exp \{j\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_{\mathrm{ref}}-\frac{M^2}{2}{\mathrm{\&alpha;}}_{\mathrm{ref}}-\frac{M^3}{3}{\mathrm{\&beta;}}_{\mathrm{ref}}-\frac{M^4}{4}{\mathrm{\&gamma;}}_{\mathrm{ref}})+j\frac{\mathrm{\&pi;}{f}_r^2}{K_r}\}---\left(12\right)$

Wherein, it is the exponential function of the truth of a matter that exp{} represents with e, and j represents imaginary unit, r _RefBe the reference oblique distance; α _Ref, β _Ref, γ _RefOblique distance model parameter for reference oblique distance place; Here, being called the described residual range migration phase term of distance after inverse Fourier transform is residual range migration amount.

Step 122: the signal of doing after distance is handled to inverse Fourier transform is made Singh Sinc interpolation processing, by the residual range migration amount in the described 2-d spectrum model of described Sinc interpolation correction;

Concrete, proofread and correct residual range migration amount with doing to carry out the Sinc interpolation apart from the signal behind inverse Fourier transform in distance Doppler territory, described residual range migration scale is shown:

${\mathrm{RCM}}_{\mathrm{diff}}(r_0,f_{\mathrm{\&eta;}})=-\frac{{\mathrm{\&lambda;}}^2[{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2+2f_{\mathrm{dc}}(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})]}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})$

$+\frac{{\mathrm{\&lambda;}}^3[2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3+3f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}})---\left(13\right)$

$-\frac{{\mathrm{\&lambda;}}^4[3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4+4f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3]}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}})$

Here, the reference function of formula (12) only compensates fully to the high-order coupling phase term of the terrain object at reference oblique distance place, but still there is high-order coupling phase term at other oblique distance places, the high-order at described other oblique distance places coupling phase term is referred to as to remain high-order coupling phase term, the expression formula of described residue high-order coupling phase term as shown in Equation (14):

Φ _res(r ₀，f _r，f _η)＝[Φ _PTRS(r ₀，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r ₀，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r] (14)

-[Φ _azi(r ₀，f _η)-Φ _azi(r _ref，f _η)]

Step 123: by the 2-d spectrum model after proofreading and correct is done apart from piecemeal, compensate the residue high-order coupling phase term in the described 2-d spectrum model;

For the described residue high-order coupling of formula (14) phase term is compensated, the present invention adopts apart from the piecemeal compensation way;

Concrete, the output of step 122 is made distance earlier to Fourier transform to two-dimensional frequency, multiply each other with the phase compensation expression formula shown in the formula (15) again.

Φ _comp(r _Mn，f _r，f _η)＝[Φ _PTRS(r _Mn，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r _Mn，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r] (15)

-[Φ _azi(r _Mn，f _η)-Φ _azi(r _ref，f _η)]

Wherein, r _MnBe the individual center oblique distance apart from ion block of n, (n is positive integer); Φ _RCM(r _Ref, f _η) f _rFor at r _MnThe range migration phase term of point, Φ _Azi(r _Ref, f _η) be at r _RefThe orientation phase modulation item of point.

Among the present invention, with each apart from the phase compensation expression formula of the residue high-order at the oblique distance place, center of piecemeal coupling phase term as each sub-piece.

Step 124: the described 2-d spectrum model after the piecemeal compensation deals of adjusting the distance is done orientation compression and is handled, thus the SAR image after obtaining focusing on according to the signal model of eliminating orientation phase modulation item;

The output of described step 123 and the orientation matched filter shown in the formula (16) are multiplied each other, and the result after multiplying each other is made the orientation to inverse Fourier transform, the satellite-borne SAR image after can obtaining focusing on.

$H_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})=\exp \{-j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}[\frac{{\mathrm{\&lambda;}}^2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})-\frac{{\mathrm{\&lambda;}}^3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}})+\frac{{\mathrm{\&lambda;}}^4{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}})\left]\right\}---\left(16\right)$

Based on above-mentioned ultrahigh resolution satellite-borne SAR image processing method, the present invention has also put down in writing a kind of ultrahigh resolution satellite-borne SAR imaging processing device, shown in Fig. 4 (a) and (b), described device comprises: set up unit 41, first and eliminate unit 42, converter unit 43, correcting unit 44, compensating unit 45, second elimination unit 46 and the image-generating unit 47; Wherein,

The described unit 41 of setting up is for the 2-d spectrum model of setting up the satellite-borne SAR echoed signal;

Described first eliminates unit 42, is used for by described 2-d spectrum model and its corresponding reference function are multiplied each other, and eliminates described 2-d spectrum model at range migration phase term and the high-order coupling phase term at reference oblique distance place;

Described converter unit 43 is used for making distance to inverse Fourier transform through the described first described 2-d spectrum model of eliminating after unit 42 is handled;

Described correcting unit 44 is used for utilizing interpolation method, is specially the residual range migration amount that the Sinc interpolation method is proofreaied and correct the described 2-d spectrum model after described converter unit 43 is handled;

Described compensating unit 45 is used for utilizing the residue high-order coupling phase term that compensates the described 2-d spectrum model after described correcting unit 44 is handled apart from the piecemeal compensation way;

Described second eliminates unit 46, is used for eliminating by the orientation compress mode orientation phase modulation item of the described 2-d spectrum model after described compensating unit 45 is handled;

Described image-generating unit 47 is for the SAR image after obtaining focusing on according to the described 2-d spectrum model after the described second elimination unit, 46 processing.

Further, the described unit 41 of setting up comprises: first sets up subelement 411, first computation subunit 412, second computation subunit 413, second sets up subelement 414, generates subelement 415, removes subelement 416 and the 3rd and set up subelement 417;

Wherein, described first sets up subelement 411, is used for setting up spaceborne platform, specifically is track six roots of sensation number and the ground surface imaging scene of setting up spaceborne platform;

Described first computation subunit 412 is used for setting up the state vector of calculating satellite and terrain object under spaceborne platform that subelement 411 sets up and the earth inertial coordinates system described first;

Described second computation subunit 413 is used for the described state vector that described first computation subunit 412 of foundation calculates, and utilizes formula (1) to (4) to calculate satellite-borne SAR high-order Doppler parameter;

Described second sets up subelement 414, is used for the high-order Doppler parameter that described second computation subunit 413 of foundation calculates, and sets up satellite-borne SAR oblique distance model as shown in Equation (5);

Described generation subelement 415 for the satellite-borne SAR state vector that calculates according to described first computation subunit 412 and the slip beam bunching mode of setting, generates the time domain SAR echoed signal under the slip beam bunching mode;

Described removal subelement 416, the orientation frequency spectrum that is used for utilizing the orientation preconditioning technique to remove described echoed signal blurs;

The described the 3rd sets up subelement 417, is used for according to formula (5) described satellite-borne SAR oblique distance model and progression inverting, sets up the 2-d spectrum model of satellite-borne SAR echoed signal;

Wherein, the described the 3rd sets up subelement 417, to carrying out distance to Fourier transform based on the satellite-borne SAR echoed signal of oblique distance model shown in the formula (5), obtain the expression formula shown in the formula (7), again formula (7) is done about the fourier transform of azimuth of orientation to time η, obtain the expression formula of formula (8), the result of calculation that recycling progression inversion method obtains is brought formula (9) into, obtain the satellite-borne SAR 2-d spectrum model shown in the formula (10), again the 2-d spectrum model shown in the formula (10) is done the expansion of progression, obtain the expansion 2-d spectrum model shown in the formula (11), the expression formula of described 2-d spectrum model comprises: orientation phase modulation item Φ _Azi(r ₀, f _η), range migration phase term Φ _RCM(r ₀, f _η) f _rAnd high-order coupling phase term

With ${\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^3.$

Described first eliminates unit 42 multiplies each other the 2-d spectrum model shown in formula (10) or the formula (11) and the reference function shown in the formula (12); Here, described first eliminates unit 42 through multiplying each other with described reference function, just eliminated described 2-d spectrum model at range migration phase term and the high-order coupling phase term at reference distance place, but still have residual range migration amount and the coupling of the residue high-order as shown in Equation (14) phase term of other distance, the existence of residue high-order coupling phase term has seriously limited the depth of focus of SAR.

43 pairs of described 2-d spectrum models after the described first elimination unit 42 is handled of described converter unit are made distance and are obtained the expression formula shown in the formula (12) to inverse Fourier transform.

Described correcting unit 44 carries out the Sinc interpolation processing with the described 2-d spectrum model after described converter unit 43 is handled, and to proofread and correct the residual range migration amount in the 2-d spectrum model, the expression formula of described residual range migration amount as shown in Equation (13).

Described compensating unit 45 further comprises: the first varitron unit 451 and the subelement 452 that multiplies each other; Wherein, the 451 pairs of described 2-d spectrum models after described correcting unit 44 is handled in the described first varitron unit are made distance to Fourier transform, and export transformation results to the described subelement 452 that multiplies each other; The result that the described subelement 452 that multiplies each other will receive and the phase compensation expression formula shown in the formula (15) multiply each other.

Described second eliminates unit 46 further comprises: coupling subelement 461 and the second varitron unit 462; Wherein, described coupling subelement 461, be used for and multiply each other through the described 2-d spectrum model after the described subelement 452 that multiplies each other is handled after described compensating unit 45 processing, concrete and the orientation matched filter shown in the formula (16), and the result after multiplying each other is exported to the described second varitron unit 462; The result that the described second varitron unit 462 will receive makes the orientation to inverse Fourier transform, triggers the image after described image-generating unit 47 obtains satellite-borne SAR focusing.

When device of the present invention was described, concept, formula (1) to expression formulas such as (16) such as related satellite and the state vector of terrain object, satellite-borne SAR high-order Doppler parameter saw also the explanation of making in the method for the present invention, repeat no more here.

The realization function that it will be appreciated by those skilled in the art that the each processing unit in the ultrahigh resolution satellite-borne SAR imaging processing device shown in Fig. 4 can be with reference to the associated description of aforementioned formation method and is understood.The function that it will be appreciated by those skilled in the art that each processing unit in the ultrahigh resolution satellite-borne SAR imaging processing device shown in Figure 4 can realize by the program that runs on the processor, also can realize by concrete logical circuit.

Fig. 5 (a) is the focus characteristics synoptic diagram based on traditional satellite-borne SAR equivalence oblique distance model imaging processing; Fig. 5 (b) is focus characteristics synoptic diagram of the present invention.In conjunction with Fig. 5 (a) and (b), the focus characteristics of formation method of the present invention is further described.

In the present embodiment, adopt parameter as shown in table 1; Distance to 10 kms (km, kilometer), the orientation is to 4km; Evenly arrange 9 terrain object (PT1-PT9), distance is to being spaced apart 5km, and the orientation is to being spaced apart 2km, and the orientation is 0.3 meter to resolution, and slant range resolution is 0.13 meter.

Parameter	Value	Parameter	Value
				Perigee altitude	668km	Carrier frequency	9.6GHz
Eccentricity	0.0011	Bandwidth	1000MHz
				Orbit inclination	98°	Distance is to sampling rate	1133MHz
Right ascension of ascending node	0°	The emission pulsewidth	2μs
				Argument of perigee	90°	Slippage factor	0.1
The satellite angle of latitude	45°	Antenna length	6m
				Nearest oblique distance	827.3km	Pulse repetition rate	4560Hz
The rotation oblique distance	919.2km	Satellite velocities	7608.34m/s

Downwards angle of visibility

35°

Irradiation time

10.89s

Table 1

The focus characteristics figure of PT3, the PT5 shown in Fig. 5 (a) and three terrain object of PT7 does not form the center accumulation, defocuses serious; And PT3, PT5 shown in Fig. 5 (b) and three terrain object of PT7 have all formed the center accumulation, and focus characteristics is good.

Table 2

Table 2 has reflected PT1 to the PT9 station-keeping ability of totally 9 terrain object, and is as shown in table 2, and the position after the focusing and the alternate position spike between the original position are that the two-dimensional position skew is less, and the locating features of formation method of the present invention is good, precision is higher.

Ultrahigh resolution satellite-borne SAR image processing method provided by the invention and device are set up the 2-d spectrum model of satellite-borne SAR echoed signal earlier; Multiply each other by described 2-d spectrum model and its corresponding reference function and to eliminate described 2-d spectrum model at range migration phase term and the high-order coupling phase term at reference oblique distance place, again to making distance to inverse Fourier transform through the signal of eliminating after handling, proofread and correct residual range migration amount in the described 2-d spectrum model by the Sinc interpolation method, by compensate the residue high-order coupling phase term in the described high-order coupling phase term apart from the piecemeal compensation way, eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compress mode again; According to the SAR image after eliminating described 2-d spectrum model behind the orientation phase modulation item and obtaining focusing on.The present invention considers the imaging processing to warp rail, and then accurate 2-d spectrum model proposed, and every the carrying out in the described 2-d spectrum model eliminated successively or compensate, thereby obtain the satellite-borne SAR image under the high-quality high-resolution, have good station-keeping ability simultaneously, can satisfy the high-resolution demands of applications.

The above is preferred embodiment of the present invention only, is not for limiting protection scope of the present invention.

Claims (15)

1. ultrahigh resolution satellite-borne synthetic aperture radar SAR image processing method is characterized in that described method comprises:

Set up the 2-d spectrum model of satellite-borne SAR echoed signal;

Multiply each other by described 2-d spectrum model and its corresponding reference function and to eliminate range migration phase term and the high-order coupling phase term at reference oblique distance place of described 2-d spectrum model, again to making distance to inverse Fourier transform through the signal of eliminating after handling, and proofread and correct residual range migration amount in the described 2-d spectrum model by interpolation method, by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, eliminate the orientation phase modulation item of described 2-d spectrum model by the orientation compress mode again;

According to the SAR image after eliminating described 2-d spectrum model behind the orientation phase modulation item and obtaining focusing on.

2. ultrahigh resolution satellite-borne SAR image processing method according to claim 1 is characterized in that, before the described 2-d spectrum model of setting up the satellite-borne SAR echoed signal, described method also comprises:

Calculate the state vector of satellite and terrain object;

Calculate satellite-borne SAR high-order Doppler parameter according to described state vector;

According to described high-order Doppler parameter, set up satellite-borne SAR oblique distance model;

According to the slip beam bunching mode of described satellite-borne SAR state vector and setting, generate the SAR echoed signal, and the orientation frequency spectrum that utilizes the orientation preconditioning technique to remove described echoed signal blurs.

3. ultrahigh resolution satellite-borne SAR image processing method according to claim 2 is characterized in that described interpolation method comprises: Singh Sinc interpolation method.

4. ultrahigh resolution satellite-borne SAR image processing method according to claim 3 is characterized in that, the described 2-d spectrum model of setting up the satellite-borne SAR echoed signal comprises:

According to satellite-borne SAR oblique distance model and the progression inversion method set up, set up the 2-d spectrum model of satellite-borne SAR echoed signal: wherein, described 2-d spectrum model Φ _PTRS(r ₀, f _r, f _η) expression formula be:

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})\&ap;-\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_0-\frac{M^2}{2}{\mathrm{\&alpha;}}_0-\frac{M^3}{3}{\mathrm{\&beta;}}_0-\frac{M^4}{4}{\mathrm{\&gamma;}}_0)-\frac{\mathrm{\&pi;}{f}_r^2}{K_2};$

The following formula series expansion is obtained:

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})={\mathrm{\&Phi;}}_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})+{\mathrm{\&Phi;}}_{\mathrm{RCM}}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^2+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^3;$

Wherein, ≈ represents to be similar to, $M=-\frac{c{f}_{\mathrm{\&eta;}}}{2(f_c+f_r)}+\frac{\mathrm{\&lambda;}{f}_{\mathrm{dc}}}{2},$ ${\mathrm{\&alpha;}}_0=\frac{1}{2k_2},$ ${\mathrm{\&beta;}}_0=-\frac{3k_3}{8k_2^3},$ ${\mathrm{\&gamma;}}_0=\frac{9k_3^2-4k_2{k}_4}{16k_2^5},$ C is the light velocity, f _cBe carrier frequency, f _rFor the distance to frequency axis, f _ηFor the orientation to frequency axis, K _rFor the distance to frequency modulation rate, r ₀Be the nearest oblique distance of radar to terrain object, f _DcBe satellite Doppler center, f _1rBe doppler frequency rate, f _2rBe second order doppler frequency rate, f _3rBe three rank doppler frequency rates, Φ _Azi(r ₀, f _η) be orientation phase modulation item, Φ _RCM(r ₀, f _η) f _rBe the range migration phase term,

Be the second order coupled item,

Be three rank coupling terms, described high-order coupling phase term comprises: second order coupled item and three rank coupling terms.

5. according to each described ultrahigh resolution satellite-borne SAR image processing method of claim 1 to 4, it is characterized in that described 2-d spectrum model corresponding reference function H _RFM(r _Ref, f _r, f _η) expression formula be:

$H_{\mathrm{RFM}}(r_{\mathrm{ref}},f_r,f_{\mathrm{\&eta;}})=\exp \{j\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_{\mathrm{ref}}-\frac{M^2}{2}{\mathrm{\&alpha;}}_{\mathrm{ref}}-\frac{M^3}{3}{\mathrm{\&beta;}}_{\mathrm{ref}}-\frac{M^4}{4}{\mathrm{\&gamma;}}_{\mathrm{ref}})+j\frac{\mathrm{\&pi;}{f}_r^2}{K_r}\};$

Wherein, it is the exponential function of the truth of a matter that exp{} represents with e, and j represents imaginary unit; r _RefBe reference oblique distance, α _Ref, β _Ref, γ _RefOblique distance model parameter for reference oblique distance place.

6. ultrahigh resolution satellite-borne SAR image processing method according to claim 5 is characterized in that, describedly proofreaies and correct residual range migration amount in the described 2-d spectrum model by interpolation method, comprising:

Described 2-d spectrum model is made distance behind inverse fourier transform, utilize the Sinc interpolation method to proofread and correct residual range migration amount in the described 2-d spectrum model; Wherein, the expression formula of described residual range migration amount is:

${\mathrm{RCM}}_{\mathrm{diff}}(r_0,f_{\mathrm{\&eta;}})=-\frac{{\mathrm{\&lambda;}}^2[{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2+2f_{\mathrm{dc}}(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})]}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})$

$+\frac{{\mathrm{\&lambda;}}^3[2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3+3f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}})$

$-\frac{{\mathrm{\&lambda;}}^4[3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4+4f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3]}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}}).$

7. ultrahigh resolution satellite-borne SAR image processing method according to claim 6 is characterized in that, and is described by compensate the residue high-order coupling phase term in the described 2-d spectrum model apart from the piecemeal compensation way, comprising:

The signal that residual range migration amount in the described 2-d spectrum model is done to proofread and correct after handling is made distance to Fourier transform, multiply each other with the phase compensation expression formula again; Described phase compensation expression formula is:

Φ _comp(r _Mn，f _r，f _η)＝[Φ _PTRS(r _Mn，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r _Mn，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r]

-[Φ _azi(r _Mn，f _η)-Φ _azi(r _ref，f _η)]

Wherein, r _MnBe the individual center oblique distance apart from ion block of n, n is positive integer, Φ _RCM(r _Mn, f _η) f _rFor at r _MnThe range migration phase term of point, Φ _Azi(r _Ref, f _η) be at r _RefThe orientation phase modulation item of point.

8. ultrahigh resolution satellite-borne SAR image processing method according to claim 7 is characterized in that, describedly eliminates the orientation phase modulation item of described 2-d spectrum model by the orientation compress mode, comprising:

Signal and the orientation matched filter of adjusting the distance behind the branch block compensation multiply each other, and the result after multiplying each other is made the orientation to inverse Fourier transform; Wherein, the expression formula of described orientation matched filter is:

$H_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})=\exp \{-j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}[\frac{{\mathrm{\&lambda;}}^2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})-\frac{{\mathrm{\&lambda;}}^3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}})+\frac{{\mathrm{\&lambda;}}^4{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}})\left]\right\};$

Wherein, λ is the SAR radar wavelength.

9. a ultrahigh resolution satellite-borne SAR imaging processing device is characterized in that described device comprises: set up unit, first and eliminate unit, converter unit, correcting unit, compensating unit, second elimination unit and the image-generating unit; Wherein,

The described unit of setting up is for the 2-d spectrum model of setting up the satellite-borne SAR echoed signal;

Described first eliminates the unit, is used for by described 2-d spectrum model and its corresponding reference function are multiplied each other, and eliminates range migration phase term and the high-order coupling phase term at reference oblique distance place of described 2-d spectrum model;

Described converter unit is used for the described 2-d spectrum model after the described first elimination cell processing is made distance to inverse Fourier transform;

Described correcting unit is used for utilizing interpolation method to proofread and correct the residual range migration amount of the described 2-d spectrum model after described converter unit is handled;

Described compensating unit is used for utilizing the residue high-order coupling phase term that compensates the described 2-d spectrum model after described correcting unit is handled apart from the piecemeal compensation way;

Described second eliminates the unit, is used for eliminating by the orientation compress mode orientation phase modulation item of the described 2-d spectrum model after described compensating unit is handled;

Described image-generating unit is for the SAR image after obtaining focusing on according to the described 2-d spectrum model after the described second elimination cell processing.

10. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 9 is characterized in that, described correcting unit is by utilizing the residual range migration amount in the described 2-d spectrum model after Singh Sinc interpolation method is proofreaied and correct described converter unit processing.

11. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 10 is characterized in that,

The expression formula of described 2-d spectrum model is:

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})\&ap;-\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_0-\frac{M^2}{2}{\mathrm{\&alpha;}}_0-\frac{M^3}{3}{\mathrm{\&beta;}}_0-\frac{M^4}{4}{\mathrm{\&gamma;}}_0)-\frac{\mathrm{\&pi;}{f}_r^2}{K_r};$

Above-mentioned series expansion is obtained:

${\mathrm{\&Phi;}}_{\mathrm{PTRS}}(r_0,f_r,f_{\mathrm{\&eta;}})={\mathrm{\&Phi;}}_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})+{\mathrm{\&Phi;}}_{\mathrm{RCM}}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}2}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^2+{\mathrm{\&Phi;}}_{\mathrm{az}\_\mathrm{rg}3}(r_0,f_{\mathrm{\&eta;}})\&CenterDot;f_r^3;$

Described 2-d spectrum model corresponding reference function H _RFM(r _Ref, f _r, f _η) expression formula be:

$H_{\mathrm{RFM}}(r_{\mathrm{ref}},f_r,f_{\mathrm{\&eta;}})=\exp \{j\frac{4\mathrm{\&pi;}(f_c+f_r)}{c}(r_{\mathrm{ref}}-\frac{M^2}{2}{\mathrm{\&alpha;}}_{\mathrm{ref}}-\frac{M^3}{3}{\mathrm{\&beta;}}_{\mathrm{ref}}-\frac{M^4}{4}{\mathrm{\&gamma;}}_{\mathrm{ref}})+j\frac{\mathrm{\&pi;}{f}_r^2}{K_r}\};$

Wherein, ≈ represents to be similar to, $M=-\frac{c{f}_{\mathrm{\&eta;}}}{2(f_c+f_r)}+\frac{\mathrm{\&lambda;}{f}_{\mathrm{dc}}}{2},$ ${\mathrm{\&alpha;}}_0=\frac{1}{2k_2},$ ${\mathrm{\&beta;}}_0=-\frac{3k_3}{8k_2^3},$ ${\mathrm{\&gamma;}}_0=\frac{9k_3^2-4k_2{k}_4}{16k_2^5},$ C is the light velocity, f _cBe carrier frequency, f _rFor the distance to frequency axis, f _ηFor the orientation to frequency axis, K _rFor the distance to frequency modulation rate, r ₀Be the nearest oblique distance of radar to terrain object, r _DcBe satellite Doppler center, f _1rBe doppler frequency rate, f _2rBe second order doppler frequency rate, f _3rBe three rank doppler frequency rates, Φ _Azi(r ₀, f _η) be orientation phase modulation item, Φ _RCM(r ₀, f _η) f _rBe the range migration phase term,

Be the second order coupled item,

Be three rank coupling terms, it is the exponential function of the truth of a matter that exp{} represents with e, and j represents imaginary unit, r _RefBe reference oblique distance, α _Ref, β _Ref, γ _RefOblique distance model parameter for reference oblique distance place.

12. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 11, it is characterized in that the described unit of setting up comprises: first computation subunit, second computation subunit, second set up subelement, generate subelement, remove subelement and the 3rd sets up subelement; Wherein,

Described first computation subunit is for the state vector of calculating satellite and terrain object;

Described second computation subunit is used for according to described state vector, calculates satellite-borne SAR high-order Doppler parameter;

Described second sets up subelement, is used for according to the high-order Doppler parameter, sets up satellite-borne SAR oblique distance model;

Described generation subelement is used for the slip beam bunching mode according to satellite-borne SAR state vector and setting, generates the SAR echoed signal;

Described removal subelement, the orientation frequency spectrum that is used for utilizing the orientation preconditioning technique to remove described echoed signal blurs;

The described the 3rd sets up subelement, according to satellite-borne SAR oblique distance model and the progression inversion method set up, sets up the 2-d spectrum model of satellite-borne SAR echoed signal.

13. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 12, it is characterized in that the described 2-d spectrum model after described correcting unit is handled described converter unit carries out the correction of residual range migration amount by the Sinc interpolation method in distance Doppler territory; Wherein, described residual range migration scale is shown:

${\mathrm{RCM}}_{\mathrm{diff}}(r_0,f_{\mathrm{\&eta;}})=-\frac{{\mathrm{\&lambda;}}^2[{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2+2f_{\mathrm{dc}}(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})]}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})$

$+\frac{{\mathrm{\&lambda;}}^3[2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3+3f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}}).$

$-\frac{{\mathrm{\&lambda;}}^4[3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4+4f_{\mathrm{dc}}{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3]}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}})$

14. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 13 is characterized in that described compensating unit comprises: the first varitron unit and the subelement that multiplies each other; Wherein,

The described first varitron unit is used for the described 2-d spectrum model after the described correcting unit processing is made distance to Fourier transform;

The described subelement that multiplies each other multiplies each other for described 2-d spectrum model and phase compensation expression formula after the described first varitron cell processing;

Wherein, described phase compensation expression formula is:

Φ _comp(r _Mnf _r，f _η)＝[Φ _PTRS(r _Mn，f _r，f _η)-Φ _PTRS(r _ref，f _r，f _η)]

-[Φ _RCM(r _Mn，f _η)·f _r-Φ _RCM(r _ref，f _η)·f _r]

-[Φ _azi(r _Mn，f _η)-Φ _azi(r _ref，f _η)]；

Wherein, r _MnBe the individual center oblique distance apart from ion block of n, n is positive integer, Φ _RCM(r _Mn, f _η) be at r _MnThe residual range migration amount of point, Φ _Azi(r _Ref, f _η) be at r _RefThe orientation phase modulation item of point.

15. ultrahigh resolution satellite-borne SAR imaging processing device according to claim 14 is characterized in that, described second eliminates the unit comprises: coupling subelement and the second varitron unit; Wherein,

Described coupling subelement is used for described 2-d spectrum model and orientation matched filter after described compensating unit is handled are multiplied each other;

The described second varitron unit is used for the described 2-d spectrum model after described coupling subelement is handled is made the orientation to inverse Fourier transform; Wherein, the expression formula of described orientation matched filter is:

$H_{\mathrm{azi}}(r_0,f_{\mathrm{\&eta;}})=\exp \{-j\frac{4\mathrm{\&pi;}}{\mathrm{\&lambda;}}[\frac{{\mathrm{\&lambda;}}^2{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^2}{8}({\mathrm{\&alpha;}}_0-{\mathrm{\&alpha;}}_{\mathrm{ref}})-\frac{{\mathrm{\&lambda;}}^3{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^3}{24}({\mathrm{\&beta;}}_0-{\mathrm{\&beta;}}_{\mathrm{ref}})+\frac{{\mathrm{\&lambda;}}^4{(f_{\mathrm{\&eta;}}-f_{\mathrm{dc}})}^4}{64}({\mathrm{\&gamma;}}_0-{\mathrm{\&gamma;}}_{\mathrm{ref}})\left]\right\};$

Wherein, λ is the SAR radar wavelength.

Images