CN104730503A - Method for determining influence on scaling by high-resolution SAR reference target RCS and compensation method - Google Patents

Abstract

The invention discloses a method for determining the influence on scaling by a high-resolution SAR reference target RCS and a compensation method. The method for determining the influence on scaling by the high-resolution SAR reference target RCS comprises the steps that based on an FEKO three-dimensional electromagnetic simulation result, data of the reference target RCS changing along with frequency bands or azimuth angles are acquired based on a multilevel fast multipole method; afterwards, the reference target RCS is introduced into an SAR echo signal, so the influence on an imaging result by the reference target RCS changing along with the frequency bands or the azimuth angles can be analyzed and estimated comprehensively and quantitatively; accordingly, the method is an analyzing method closer to actual conditions. The compensation method comprises the steps that firstly, existing high-resolution SAR reference target RCS data having different center frequencies and/or frequency band ranges are simulated; secondly, based on the result of the influence on scaling, the maximum value of integral energy difference values in bandwidth which actually uses high-resolution SAR is judged, and whether at least one of the maximum values of integral energy difference values in azimuth beamwidth is greater than or equal to 0.2 dB is judged, so whether echo compensation is performed or not is determined. Accordingly, the computation burden can be reduced while SAR detection accuracy is met.

Description

Determine that high resolution SAR reference target RCS is on the method for calibration impact and compensation method

Technical field

The invention belongs to signal processing technology field, be specifically related to a kind of method that high resolution SAR reference target RCS affects calibration of determining.

Background technology

For arrowband and narrow azimuth beam SAR (Synthetic Aperture Radar) system, think RCS (Radar Cross-Section, the RCS) approximately constant in bandwidth of operation and azimuth beam of scaled reference point target.But, along with in remote sensing application to the raising of image resolution requirement, create a lot of high resolution SAR (frequency band range is greater than the SAR that 2GHz and/or azimuth beamwidth be greater than 20 ° and is generally considered to be high resolution SAR) system, relative bandwidth even can reach 100% more than 10%, and orientation ranges of incidence angles can be increased to tens degree even 360 ° (circle mark SAR) from the several years.In the large distance of modem high-resolution SAR system to bandwidth sum wide-azimuth under the feature of wave beam, with the backscattering characteristic of the RCS approximate representation reference target at centre frequency and position angle place, can not reach the accuracy requirement of radiation calibration, frequency band or the position angle correlativity of reference target RCS must take in.When computer storage capacity and arithmetic speed develop rapidly, be also that simulation of Radar System provides guarantee subtly.

RCS traditional at present calculates many based on high-frequency approximation theoretical formula model, and the many hypothesis based on constant RCS of SAR echo simulation, there is deficiency below in this technical method:

(1) RCS computation model is inaccurate: theoretical formula model, obtain through multistep approximate processing, but by object and background environment separation during actual rcs measurement, and provide sufficient far field condition to be all a challenge, can not represent by desirable analytic formula.

(2) frequency band and the position angle correlativity of RCS is ignored in SAR echo simulation: high resolution SAR system additionally can produce multiple error due to its large distance to bandwidth sum wide-azimuth to the feature of wave beam, wherein just comprises reference point target radiation characteristic error.

(3) do not carry out comprehensive quantitative test assessment: current when carrying out high resolution SAR system emulation, do not consider the frequency band of RCS and the impact of position angle correlativity simultaneously, and comprehensive not to the quality evaluation index of net result.

Summary of the invention

In view of this, the invention provides a kind of method that high resolution SAR reference target RCS affects calibration of determining, the data of reference target RCS on calibration impact can be obtained; Meanwhile, present invention also offers a kind of based on the method for this calibration impact to echo cancellation, according to the integral energy difference obtained in the method for calibration impact, can judge whether to compensate echo, the detection accuracy that high precision SAR is certain can be ensured thus.

Determine the method for high resolution SAR reference target RCS on calibration impact, comprise the steps:

Step 1, for the existing High Resolution SAR with different center frequency and/or frequency band range, apply the electromagnetic scattering emulation geometric model of FEKO software building reference target respectively, then in FEKO software, select multilevel fast multipole method to obtain reference target RCS corresponding to each High Resolution SAR respectively with frequency band and the data with azimuthal variation, data corresponding to each High Resolution SAR all have two dimensions, first dimension is that this High Resolution SAR is in the data of beam angle internal reference target RCS with azimuthal variation, second dimension is that this High Resolution SAR is in the data of distance bandwidth internal reference target RCS with frequency change,

Step 2, for each High Resolution SAR, set up original echoed signals realistic model respectively, and to original echoed signals in distance to doing Fourier transform, obtain distance to frequency-region signal; For each High Resolution SAR, in beam angle, orientation in echo signal data is obtained after reference target RCS aligns by position angle with the data of azimuthal variation to each sampled point and the step 1 of time-domain signal and is multiplied, obtain and introduce orientation that reference target RCS affects to time-domain signal data; In distance bandwidth, the reference target RCS obtained to each sampled point of frequency-region signal and step 1 by echo signal data middle distance is multiplied after aliging by Frequency point with the data of frequency change, obtain introduce reference target RCS impact distance to frequency-region signal data;

Step 3, for each High Resolution SAR, to time domain 2-D data, Range compress and Azimuth Compression are carried out to frequency domain and orientation to the distance introducing reference target RCS impact that step 2 obtains, obtains the echoed signal SAR complex pattern introducing reference target RCS and affect; Range compress and Azimuth Compression are carried out to described original echo 2-D data, obtains the SAR complex pattern of original signal; For the SAR complex pattern of the echoed signal SAR complex pattern and original signal that introduce reference target RCS impact, respectively difference is asked to the point target peak energy of this two width SAR complex pattern, integral energy, peak sidelobe ratio and integration secondary lobe ratio, obtain reference target RCS and frequency band and/or azimuthal correlativity to the impact of image quality, determine thus each for the reference target RCS of each High Resolution SAR and frequency band and/or azimuthal change on the impact of calibrating.

Preferably, in described step 1, the step-length arranging frequency bandwidth in FEKO software is 20MHz, and the angle step arranging azimuth beam is 0.2 °.

Preferably, for the 2-D data that step 1 obtains, cubic spline interpolation is adopted to carry out interpolation, make reference target RCS consistent to the data point of time-domain signal with orientation in the echoed signal in step 2 with the data of azimuthal variation, make reference target RCS consistent to frequency-region signal data point with echo signal data middle distance in step 2 with the data point of frequency change.

Of the present invention a kind of based on the compensation method of above-mentioned calibration impact to echo, the integral energy difference data corresponding with the high resolution SAR that reality uses is found in step 3, and the maximal value of integration energy differences in the maximal value of integral energy difference obtained in bandwidth of this actual use high resolution SAR and azimuth beamwidth, judging that whether these two maximal values are satisfied has at least one to be more than or equal to 0.2dB, if met: compensate the actual ghosts signal of the High Resolution SAR of described practical application; If do not met, do not need to compensate actual ghosts signal;

Described compensation method is: to described actual ghosts signal in distance to doing Fourier transform, obtain distance to frequency-region signal; Reference target RCS corresponding to each high precision SAR obtained in described step 1, with frequency band and with in the 2-D data of azimuthal variation, finds the 2-D data corresponding with the high precision SAR of practical application, each for this 2-D data data point is asked respectively reciprocal; Orientation in actual ghosts signal data is multiplied after each sampled point of time-domain signal aligns by position angle with the data asking the reference target RCS after inverse with azimuthal variation, obtains the orientation after compensating to time-domain signal data; Actual ghosts signal data middle distance is multiplied after each sampled point of frequency-region signal aligns by Frequency point with the data asking the reference target RCS after inverse with frequency change, obtains the distance after compensating to frequency-region signal data.

The present invention has following beneficial effect:

(1) a kind of method that high resolution SAR reference target RCS affects calibration of determining of the present invention, compared with prior art, advantage is: based on FEKO 3 D electromagnetic simulation result, multilevel fast multipole method (MLFMM) is adopted to obtain reference target RCS with frequency band or azimuthal variation data, then this reference target RCS is incorporated in SAR echo signal emulation, thus can quantitative test assessment reference target RCS comprehensively with frequency band or azimuthal variation influence on RT, the method is a kind of more close to the analytical approach of actual conditions compared with the conventional method.

(2) by rationally arranging the step-length of the frequency bandwidth in FEKO software and the angle step of azimuth beam, while reducing the operand of FEKO, the data and curves variation tendency that FEKO can also be made to export satisfies condition.

(3) a kind of method echo compensated based on calibration impact of the present invention, by first emulating the existing reference target RCS data with the High Resolution SAR of different center frequency and/or frequency band range, then based on the result that above-mentioned calibration affects, judge whether to meet in the maximal value of integration energy differences in the maximal value of the integral energy difference in the bandwidth of actual use high resolution SAR and azimuth beamwidth to have at least one to be more than or equal to 0.2dB to determine whether to carrying out echo cancellation, thus can while meeting the detection accuracy of SAR, save operand.

Accompanying drawing explanation

Fig. 1 is the electromagnetic scattering emulation geometric model figure by FEKO software building reference target in the present invention.

Fig. 2 (a) is reference target normalization RCS surface chart in the present invention; The section curve at the centre frequency that the surface chart that Fig. 2 (b) is Fig. 2 (a) is corresponding and central party parallactic angle place.

Fig. 3 is RCS with the point target analogous diagram of frequency band and azimuthal variation.

Wherein 1-excitation, 2-reverberator.

Embodiment

To develop simultaneously embodiment below in conjunction with accompanying drawing, describe the present invention.

The quantitative relationship setting up target scattering characteristics and echoed signal by radar equation is the prerequisite of radiation calibration theoretical analysis, and the average received signal power of High Resolution SAR Images can be expressed as:

$P_r=\frac{P_t{G}^2(f,\mathrm{\α})G_s\left(f\right){\mathrm{\λ}}^3\left(f\right){\mathrm{\τ}}_p{f}_s{f}_{\mathrm{PRF}}}{2{\left(4\mathrm{\π}\right)}^3{R}^3\left(\mathrm{\α}\right)V{\mathrm{\ρ}}_a}\mathrm{\σ}(f,\mathrm{\α})+P_n$

Wherein, P _tfor transmit signal power; P _nfor noise power; G ²(f, α) is round trip antenna radiation pattern; for the visual angle of radar illumination target, θ is the angle of pitch, for position angle, in carried SAR radiation calibration process, for given reference point target, need the angle of pitch that corner reflector is set exactly according to flight track or flight path, make it reach close to desirable duty, therefore, only need consider that RCS is with azimuthal variation characteristic; G _sf () is system-gain; λ (f) is the wavelength that transmits; The oblique distance that R (α) is radar and target; τ _pfor fire pulse width; f _sfor distance is to sample frequency; f _pRFfor orientation is to sample frequency; V is Texas tower flying speed; ρ _afor azimuth resolution; The radar cross section that σ (f, α) is point target.

Consider that high resolution SAR system has large distance to bandwidth sum wide-azimuth to wave beam, reference point target ideal radiation characteristic is no longer satisfied, for keeping the consistance with existing radiation calibration algorithm, quantitative test must be carried out to the frequency band of reference target RCS and position angle correlativity, make reference point target still have smooth response in system bandwidth and azimuth beam, thus guarantee that radiometric difference is because the radiation characteristic change of imageable target instead of reference target causes.The bandwidth sum targeting device RCS change bandwidth in corresponding according to system operating mode, by the bearing calibration in echo territory, point target bandwidth modulation effect is corrected, and re-imaging, to eliminate the error of the antenna pattern measurement that bandwidth is introduced, azimuth beam (response) width according to azimuth scan scope and targeting device is also needed for beam bunching mode, considers the impact on radiation characteristic of azimuth beamwidth.

In view of this, the present invention is in order to solve under Current high resolution SAR system condition the how Obtaining Accurate reference target RCS relevant to frequency band or position angle, and SAR system emulation time think reference target RCS in bandwidth of operation and azimuth beam approximately constant hypothesis deficiency, propose a kind of quantitative test high resolution SAR reference target radiation characteristic on calibration impact method.The method is on the basis of the high resolution SAR radiation calibration model set up, by FEKO 3 D electromagnetic simulation software, obtain reference target RCS with frequency band and azimuthal variation relation, in conjunction with high resolution SAR system emulation, analyze the frequency band of reference target RCS or position angle correlativity quantitatively to the impact of SAR image quality.

A kind of method that high resolution SAR reference target RCS affects calibration of determining of the present invention, comprises the steps:

Step 1, for the existing High Resolution SAR with different center frequency and/or frequency band range, apply the electromagnetic scattering emulation geometric model of FEKO software building reference target respectively, then in FEKO software, select multilevel fast multipole method (MLFMM) to obtain reference target RCS corresponding to each High Resolution SAR respectively with frequency band and the data with azimuthal variation, data corresponding to each High Resolution SAR all have two dimensions, first dimension is that this High Resolution SAR is in the data of beam angle internal reference target RCS with azimuthal variation, second dimension is that this High Resolution SAR is in the data of distance bandwidth internal reference target RCS with frequency change, MLFMM is the rapidly and efficiently numerical algorithm of research Electrically large size object problem, and it makes on small-size computer, solve large-scale electromagnetic scattering problems becomes possibility.In the present embodiment, under the illuminate condition of Ku wave band far field, operating center frequency is f ₀=16GHz, bandwidth deltaf f are 4GHz; The angle of pitch is fixed value θ=54.74 °, position angle with centered by, azimuth coverage it is 30 °; Trihedral corner reflector 2 is of a size of l=20cm, and meet l ≈ 10 λ, namely target scattering characteristics belongs to high frequency region.It should be noted that, consider the configuration of PC and the calculated amount of FEKO engineering, because the calculated amount of FEKO can be multiplied with the discrete point quantity in its optimum configurations, getting frequency step in distance bandwidth is herein 20MHz, in azimuth beamwidth, angle step is 0.2 °, although arrange in FEKO discrete count much smaller than the distance in SAR simulation parameter to orientation to sampling number, but when discrete count can reflect plots changes, can adopt cubic spline interpolation process that discrete the counting of being derived by FEKO is finally met with SAR distance to consistent to sampling number with orientation, thus avoid FEKO project scale excessive, reduce operand, draw the centre section curve of normalization RCS curved surface and correspondence as shown in Fig. 2 (a) He (b).

Step 2, for each High Resolution SAR, set up original echoed signals realistic model respectively, and to original echoed signals in distance to doing Fourier transform, obtain distance to frequency-region signal; For each High Resolution SAR, orientation in echo signal data is obtained after reference target RCS aligns by position angle with the data of azimuthal variation to each sampled point and the step 1 of time-domain signal and is multiplied, obtain and introduce orientation that reference target RCS affects to time-domain signal data; The reference target RCS obtained to each sampled point of frequency-region signal and step 1 by echo signal data middle distance is multiplied after aliging by Frequency point with the data of frequency change, obtains and introduces distance that reference target RCS affects to frequency-region signal data;

Step 3, for each High Resolution SAR, to time domain 2-D data, Range compress and Azimuth Compression are carried out to frequency domain and orientation to the distance introducing reference target RCS impact that step 2 obtains, obtains the echoed signal SAR complex pattern introducing reference target RCS and affect; Range compress and Azimuth Compression are carried out to described original echo 2-D data, obtains the SAR complex pattern of original signal; For the SAR complex pattern of the echoed signal SAR complex pattern and original signal that introduce reference target RCS impact, respectively difference is asked to the point target peak energy of this two width SAR complex pattern, integral energy, peak sidelobe ratio and integration secondary lobe ratio, obtain reference target RCS and frequency band and/or azimuthal correlativity to the impact of image quality, determine thus each for the reference target RCS of each High Resolution SAR and frequency band and/or azimuthal change on the impact of calibrating.

Result as shown in Figure 3, in four groups of bandwidth sum azimuth beam setting parameter situations, the 1st group of parameter: Δ f=0.5GHz, 2nd group of parameter: Δ f=1GHz, 3rd group of parameter: Δ f=2GHz, 4th group of parameter: Δ f=4GHz, the relative change size of point target peak energy, integral energy, peak sidelobe ratio and integration secondary lobe ratio.Wherein, Δ f represents bandwidth, represent azimuth beamwidth.

Usually, in whole SAR system radiation error is distributed, reference target RCS changes the error caused, and active targeting device can not more than 0.2dB, and passive targeting device can not more than 0.3dB.Found by Fig. 2 and 3 simulation results, above-mentioned impact has reached radiation calibration and can not ignore degree, must be compensated the radiation characteristic impact of large distance bandwidth sum wide-azimuth wave beam on reference target.Carry out choosing of signal transacting region at complex image domain, and then be mapped to echo territory, obtain and need SAR echo signal row, column to be processed interval.

Therefore, present invention also offers a kind of based on the compensation method to actual ghosts signal of reference target RCS on calibration impact, before imaging algorithm, echoed signal compensated, be specially:

The integral energy difference data corresponding with the high resolution SAR that reality uses is found in step 3, and the maximal value of integration energy differences in the maximal value of integral energy difference obtained in bandwidth of this actual use high resolution SAR and azimuth beamwidth, judging that whether these two maximal values are satisfied has at least one to be more than or equal to 0.2dB, if met: compensate the actual ghosts signal of the High Resolution SAR of described practical application; If do not met, do not need to compensate actual ghosts signal;

Described compensation method is: to described actual ghosts signal in distance to doing Fourier transform, obtain distance to frequency-region signal; Reference target RCS corresponding to each high precision SAR obtained in described step 1, with frequency band and with in the 2-D data of azimuthal variation, finds the 2-D data corresponding with the high precision SAR of practical application, each for this 2-D data data point is asked respectively reciprocal; Orientation in actual ghosts signal data is multiplied after each sampled point of time-domain signal aligns by position angle with the data asking the reference target RCS after inverse with azimuthal variation, obtains the orientation after compensating to time-domain signal data; Actual ghosts signal data middle distance is multiplied after each sampled point of frequency-region signal aligns by Frequency point with the data asking the reference target RCS after inverse with frequency change, obtains the distance after compensating to frequency-region signal data.

In sum, these are only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.Within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (4)

1. determine the method for high resolution SAR reference target RCS on calibration impact, it is characterized in that, comprise the steps:

Step 1, for the existing High Resolution SAR with different center frequency and/or frequency band range, apply the electromagnetic scattering emulation geometric model of FEKO software building reference target respectively, then in FEKO software, select multilevel fast multipole method to obtain reference target RCS corresponding to each High Resolution SAR respectively with frequency band and the data with azimuthal variation, data corresponding to each High Resolution SAR all have two dimensions, first dimension is that this High Resolution SAR is in the data of beam angle internal reference target RCS with azimuthal variation, second dimension is that this High Resolution SAR is in the data of distance bandwidth internal reference target RCS with frequency change,

Step 2, for each High Resolution SAR, set up original echoed signals realistic model respectively, and to original echoed signals in distance to doing Fourier transform, obtain distance to frequency-region signal; For each High Resolution SAR, in beam angle, orientation in echo signal data is obtained after reference target RCS aligns by position angle with the data of azimuthal variation to each sampled point and the step 1 of time-domain signal and is multiplied, obtain and introduce orientation that reference target RCS affects to time-domain signal data; In distance bandwidth, the reference target RCS obtained to each sampled point of frequency-region signal and step 1 by echo signal data middle distance is multiplied after aliging by Frequency point with the data of frequency change, obtain introduce reference target RCS impact distance to frequency-region signal data;

Step 3, for each High Resolution SAR, to time domain 2-D data, Range compress and Azimuth Compression are carried out to frequency domain and orientation to the distance introducing reference target RCS impact that step 2 obtains, obtains the echoed signal SAR complex pattern introducing reference target RCS and affect; Range compress and Azimuth Compression are carried out to described original echo 2-D data, obtains the SAR complex pattern of original signal; For the SAR complex pattern of the echoed signal SAR complex pattern and original signal that introduce reference target RCS impact, respectively difference is asked to the point target peak energy of this two width SAR complex pattern, integral energy, peak sidelobe ratio and integration secondary lobe ratio, obtain reference target RCS and frequency band and/or azimuthal correlativity to the impact of image quality, determine thus each for the reference target RCS of each High Resolution SAR and frequency band and/or azimuthal change on the impact of calibrating.

2. as claimed in claim 1 a kind of determine high resolution SAR reference target RCS on calibration impact method, it is characterized in that, in described step 1, the step-length arranging frequency bandwidth in FEKO software is 20MHz, and the angle step arranging azimuth beam is 0.2 °.

3. as claimed in claim 2 a kind of determine high resolution SAR reference target RCS on calibration impact method, it is characterized in that, for the 2-D data that step 1 obtains, cubic spline interpolation is adopted to carry out interpolation, make reference target RCS consistent to the data point of time-domain signal with orientation in the echoed signal in step 2 with the data of azimuthal variation, make reference target RCS consistent to frequency-region signal data point with echo signal data middle distance in step 2 with the data point of frequency change.

4. calibrate the compensation method of impact to echo based on according to claim 1 for one kind, it is characterized in that: find the integral energy difference data corresponding with the high resolution SAR that reality uses in step 3, and the maximal value of integration energy differences in the maximal value of integral energy difference obtained in bandwidth of this actual use high resolution SAR and azimuth beamwidth, judging that whether these two maximal values are satisfied has at least one to be more than or equal to 0.2dB, if met: compensate the actual ghosts signal of the High Resolution SAR of described practical application; If do not met, do not need to compensate actual ghosts signal;

Described compensation method is: to described actual ghosts signal in distance to doing Fourier transform, obtain distance to frequency-region signal; Reference target RCS corresponding to each high precision SAR obtained in described step 1, with frequency band and with in the 2-D data of azimuthal variation, finds the 2-D data corresponding with the high precision SAR of practical application, each for this 2-D data data point is asked respectively reciprocal; Orientation in actual ghosts signal data is multiplied after each sampled point of time-domain signal aligns by position angle with the data asking the reference target RCS after inverse with azimuthal variation, obtains the orientation after compensating to time-domain signal data; Actual ghosts signal data middle distance is multiplied after each sampled point of frequency-region signal aligns by Frequency point with the data asking the reference target RCS after inverse with frequency change, obtains the distance after compensating to frequency-region signal data.