A doppler center precise compensation method based on PFA algorithm real-time imaging

Abstract

The present application relates to the field of airborne synthetic aperture radar imaging, and discloses a Doppler center fine compensation method based on PFA algorithm real-time imaging, which realizes real-time estimation and compensation of each frame through steps of distance motion compensation, Doppler center and residual squint angle estimation, repositioning imaging center and echo residual compensation. The method can solve the problem of imaging center deviation caused by beam control error, antenna installation error and non-ideal motion error of the carrier, and can avoid the time lag problem in real-time processing by combining with the current real-time processing architecture based on the PFA algorithm, so as to realize the real-time processing capability of estimating each frame and compensating each frame, avoid the problems of scene center deviation and azimuth defocus of the SAR image, realize the consistency of the imaging center and the beam center, and improve the imaging quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of airborne synthetic aperture radar imaging, in particular to a Doppler center fine compensation method based on PFA algorithm real-time imaging. BACKGROUND

[0002] Synthetic aperture radar (SAR) is an active microwave remote sensing device, which has the ability of high-resolution imaging and motion detection and positioning in all-weather and all-day. The significance of real-time imaging of high-resolution airborne SAR is self-evident. In the military field, it can provide timely information for the command system through real-time reconnaissance of the battlefield. In other fields, such as geological exploration, disaster assessment, agriculture and forestry, real-time processing can achieve on-site monitoring, which also has great significance. At the same time, real-time compression of radar images before transmission can greatly reduce the requirements for data transmission channels.

[0003] Airborne spotlight SAR can increase the synthetic aperture length by controlling the antenna beam to always point to the same area, so as to realize continuous observation and high-resolution imaging of the scene. PFA algorithm is a classic spotlight SAR imaging algorithm. After selecting the imaging center, the algorithm stores data in polar coordinate format and eliminates coupling through two-dimensional interpolation, which can effectively solve the problem of more than one resolution cell walking of the scattering point far from the center of the imaging area, and simply and efficiently compensate for the errors introduced by the non-ideal flight of the radar platform. Due to the existence of radar antenna installation errors, beam control errors and inertial navigation attitude errors, the actual beam energy coverage area will deviate from the predetermined scene. When the selected imaging center deviates from the center of the actual pointing of the beam, not only will the image produce distance and azimuth deviation, which is not conducive to continuous multi-frame continuous observation and positioning, but also will cause image defocusing or even unable to image in severe cases.

[0004] In order to have the ability of real-time processing, airborne SAR real-time imaging system not only depends on high-performance digital signal processor hardware, but also puts forward higher requirements for imaging signal processing flow and algorithm. The existing real-time processing flow usually adopts a stream processing architecture, which completes the distance processing of the echo while accumulating the azimuth aperture, and then waits for the completion of a frame of echo accumulation before performing subsequent processing. Compared with processing all echoes at the same time, this method improves the processing efficiency and shortens the imaging time. However, this method of processing while accumulating needs to assume the motion state of the aircraft and the pointing of the radar beam in advance as the input of the distance processing. When the actual flight path deviates from the ideal setting, it cannot be perceived during distance processing, which has a time lag, and it is difficult to compensate for the echo with errors in real time after the completion of azimuth accumulation.

[0005] The existing technology has the following disadvantages:

[0006] 1) Only the azimuth imaging angle is estimated, and the imaging center position is not estimated, which can only solve the problems of image azimuth deviation and defocus, and cannot solve the problem of range deviation.

[0007] 2) The azimuth imaging angle is estimated directly from the original data using the Doppler center frequency, and the estimated value contains the aircraft motion error, which cannot truly reflect the beam pointing error.

[0008] 3) The estimated azimuth imaging angle is difficult to compensate in real time in the frame, and can only be compensated in the next frame, which has a lag, and needs multiple frame iterations to converge. SUMMARY

[0009] Therefore, the present application provides a Doppler center fine compensation method based on PFA algorithm real-time imaging, which solves the problems of real-time image deviation and defocus of the spotlight SAR by re-estimating the imaging center and compensating the echo, realizes real-time compensation in the frame, and ensures the consistency of the imaging center and the beam energy center, thereby improving the imaging quality and continuous multi-frame imaging capability.

[0010] A Doppler center fine compensation method based on PFA algorithm real-time imaging, comprising the following steps:

[0011] Step one, determine the virtual imaging center by setting the ideal flight path, and perform distance processing on the original echo to obtain the distance-processed echo;

[0012] Step two, estimate the Doppler center for the distance-processed echo, and calculate the ground plane squint angle error;

[0013] Step three, reposition the imaging center according to the actual flight path;

[0014] Step four, real-time compensate the echo data to the new imaging center according to the ground plane squint angle error, and update the azimuth interpolation parameters, so as to complete the subsequent PFA imaging processing flow.

[0015] Further, the distance processing method for the original echo in step one is:

[0016] s(m, n) = s0(m, n) x H moco (m, n); wherein H moco (m, n) is a motion compensation reference function, H moco (m, n) = exp{j·[K c + K r (n)]·[R p (m) - R ref ]}, is the center wave number, f c is the center frequency, n is the distance wave number, and Nr The instantaneous slant range R of the aircraft is the number of sampling points in the distance direction. p (m), m=1,2,L,N a N a f is the number of sampling points in the azimuth direction; s R is the range sampling frequency, j is the imaginary part of the complex number; s0(m,n) is the original echo, s(m,n) is the range-processed echo; ref This is the reference slope distance.

[0017] Furthermore, the method for calculating the oblique angle error in step two includes the following steps:

[0018] 1) When processing distance-oriented data, allocate N... r temp, a complex array of 2 rows and 2 columns Nr×2 Used to temporarily store echoes;

[0019] 2) Starting with the first echo, place the range-processed echoes into the first column of temp until all echoes are collected;

[0020] 3) Starting from the second echo, multiply the current echo after distance processing by the conjugate with the previous echo, add it to the second column of temp, and put it into the second column of temp, until all echoes are collected;

[0021] temp(n,2)=temp(n,2)+s(m,n)s * (m-1,n)

[0022] m = 2, 3, L, N a n = 1, 2, ..., N r

[0023] 4) Calculate the phase angle of the second column of temp, take the average value, and calculate the Doppler center frequency offset Δf. dc

[0024]

[0025] Wherein, PRF is the pulse repetition frequency;

[0026] 5) Calculate the error of the oblique angle on the ground plane. λ is the average speed of the carrier aircraft within the entire aperture, and λ is the radar wavelength.

[0027] Furthermore, step three, which calculates the actual beam center position using imaging geometry and uses it as the new imaging center, specifically includes the following steps:

[0028] Calculate the distance from the actual synthetic aperture center projected onto the virtual imaging center on the ground;

[0029] Taking the actual synthetic aperture center as the center of a circle, taking the distance from the actual synthetic aperture center to the virtual imaging center as the radius, and rotating the virtual imaging center along the track direction by an angle Δθ g ;

[0030] The rotated end point is taken as a new imaging center.

[0031] Further, the instantaneous slant range of the carrier to the new imaging center is corrected by using the azimuth interpolation parameter in step four, and the specific steps include the following steps:

[0032] According to the instantaneous pitch angle, the instantaneous slant angle, the ground plane distance wave number and the instantaneous slant range in the distance direction processing process, the instantaneous slant range of the carrier to the new imaging center is calculated.

[0033] The echo after the distance direction processing is multiplied by the residual motion compensation phase function, and is taken as the echo input of the subsequent azimuth processing.

[0034] The ground plane instantaneous slant angle under the new imaging center is calculated, is taken as the input parameter of the subsequent azimuth interpolation, and the subsequent imaging processing work is completed.

[0035] Further, the residual motion compensation phase function is:

[0036]

[0037] Wherein, is the instantaneous pitch angle in the original echo distance direction processing process, θ g (m) is the original echo ground plane instantaneous slant angle, K y (n) is the original echo ground plane distance wave number, R p (m) is the instantaneous slant range of the original echo, is the instantaneous slant range of the new imaging center , θ gc is the ground plane slant angle of the aperture center at the moment.

[0038] Compared with the prior art, the beneficial effects of the present application are:

[0039] 1. The present application combines the real-time processing flow of the spotlight SAR, re-estimates the imaging center and compensates the echo, solves the problems of the real-time image deviation and defocusing of the spotlight SAR, realizes the real-time compensation of each frame, thereby ensuring the consistency of the imaging center and the beam energy center, improving the imaging quality and the continuous multi-frame imaging ability.

[0040] 2、The application can solve the imaging center deviation problem caused by beam control error, antenna installation error and non-ideal motion error of a carrier, and can be combined with the current real-time processing architecture based on the PFA algorithm to avoid the time lag problem in real-time processing, realize the real-time processing capability of frame estimation and frame compensation, and can be used in the fields of real-time imaging of airborne spotlight SAR and strip SAR. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0042] Figure 1 is a flow chart of a Doppler center fine compensation method based on the PFA algorithm real-time imaging in embodiment 2;

[0043] Figure 2 is an imaging geometry schematic diagram in embodiment 2;

[0044] Figure 3 is an imaging center repositioning compensation result in embodiment 2. DETAILED DESCRIPTION

[0045] The embodiments of the present application will be described in detail below with reference to the drawings.

[0046] The embodiments of the present application will be described in detail below with reference to the drawings.

[0047] Embodiment 1

[0048] Referring to Figure 1 A Doppler center fine compensation method based on the PFA algorithm real-time imaging includes the following steps:

[0049] Step one, determine a virtual imaging center by setting an ideal flight path, perform distance processing on the original echo, and obtain the echo after distance processing;

[0050] Step two, Doppler center estimation is performed on the processed echo in the range direction, and the ground plane squint angle error is calculated;

[0051] Step three, the imaging center is relocated according to the actual flight path;

[0052] Step four, real-time compensation of echo data to the new imaging center according to the ground plane squint angle error, and updating of the azimuth interpolation parameter, so as to complete the subsequent PFA imaging processing flow.

[0053] In the embodiment, during the imaging of the radar frame, the virtual imaging center is determined by the ideal flight path of the frame, the original echo is processed in the range direction, the Doppler center estimation is performed on the motion compensated data after all echoes of the frame are collected, and the actual imaging squint angle of the frame is obtained. Then, the synthetic aperture center slant range and the instantaneous slant range are recalculated according to the actual flight path of the frame, and finally, the scene center azimuth and range direction offset caused by the squint angle error and the instantaneous slant range residual are compensated before the azimuth interpolation, so as to realize the real-time estimation and compensation of the frame. The method can solve the problem of imaging center deviation caused by beam control error, antenna installation error and non-ideal motion error of the carrier, and can avoid the time lag problem in real-time processing by combining with the current real-time processing architecture based on the PFA algorithm. The real-time processing capability of the frame estimation and compensation is realized, the problems of scene center offset and azimuth defocus of the SAR image can be avoided, the imaging center and the beam center are consistent, and the imaging quality is improved.

[0054] It should be noted that after the imaging processing of the current frame is completed, the imaging process from the starting point to the end point of the next frame can be repeated by steps one to four.

[0055] Embodiment 2

[0056] Step one, the virtual imaging center is determined by setting the ideal flight path, which is used for range direction processing:

[0057] Figure 2 The imaging geometry of the spotlight SAR is shown in the figure, assuming that the carrier moves at a constant speed in a straight line, A and B are the starting point and the end point of the ideal flight path of the current imaging frame respectively, and C is the synthetic aperture center. When the carrier moves to point A, it is the starting point of the current frame, that is, the first echo data sampling point of the current frame, at this time, the virtual imaging center P of the ideal flight path is established. track The flight path angle of the ideal flight path (the angle between the projection line of AB to the ground and the north direction, which can be determined according to the average flight path angle of the last frame or the flight path angle of the current frame A point).

[0058] P is the virtual imaging center when imaging by the PFA algorithm, which can be determined by the distance from A to P (i.e. R P (m) value), the ground plane azimuth angle θ ga , and the pitch angle to determine the three-dimensional coordinates of P relative to A. The range R ref , as the reference slant range of the current frame echo distance sampling.

[0059] The curve AE is the actual flight path of the carrier, which can be measured by an inertial navigation device. Assuming that the mth azimuth sampling point is the instantaneous slant range of the carrier to P R p (m), m = 1, 2, L, N a , N a is the number of azimuth sampling points, then the motion compensation in the PFA distance processing needs to multiply the reference function H moco (m, n)

[0060] s(m, n) = s0(m, n) x H moco (m, n)

[0061] where H moco (m, n) = exp{j·[K c + K r (n)]·[R p (m) - R ref ]}, is the central wave number, is the distance wave number, N r is the distance sampling point number, f s is the distance sampling frequency, and j is the imaginary part of the complex number; s0(m, n) is the original echo, and s(m, n) is the echo after distance processing.

[0062] Step two, Doppler center estimation is performed on the echo after distance processing, and the ground plane squint angle error is calculated:

[0063] If the actual beam is correctly directed, the Doppler center of the echo after distance processing (matched filtering, motion compensation, and distance interpolation) is zero, that is, the motion compensation has removed the echo Doppler shift caused by the non-ideal motion of the carrier. However, due to the existence of antenna installation error and beam control error, the actual beam center is Imaging with the virtual center P will cause image shift and even defocus.

[0064] Therefore, it is necessary to estimate the actual squint angle from the echo and calculate the coordinates of the actual beam center . The squint angle error can be calculated by Doppler center estimation on the echo after distance processing. The specific implementation method is as follows:

[0065] 1) In distance processing, an N r row, 2-column complex number array temp Nr×2 is opened to temporarily store the echo.

[0066] 2) Starting from the first echo, put the distance-processed echoes into the first column of temp until all echoes are collected.

[0067] 3) Starting from the second echo, multiply the current echo after distance processing by the conjugate with the previous echo, add it to the second column of temp, and put it into the second column of temp, until all echoes are collected.

[0068] temp(n,2)=temp(n,2)+s(m,n)s * (m-1,n)

[0069] m = 2, 3, L, N a n = 1, 2, ..., N r

[0070] 4) Calculate the phase angle of the second column of temp, take the average value, and calculate the Doppler center frequency offset Δf. dc

[0071]

[0072] Where PRF is the pulse repetition frequency.

[0073] 5) Calculate the error of the oblique angle of the ground plane Δθ g

[0074]

[0075] λ is the average speed of the carrier aircraft within the entire aperture, and λ is the radar wavelength.

[0076] Step 3: Recalculate the imaging center based on the actual trajectory of the current frame. The coordinates.

[0077] The Doppler center estimation in step two is completed after range processing and is the average value of instantaneous Doppler, representing the deviation of the actual beam pointing of the synthetic aperture center on the ground plane from the ideal beam pointing. Therefore, the actual beam center position can be calculated through imaging geometry and used as the new imaging center.

[0078] 1) Calculate the projection D of the actual synthetic aperture center on the ground. g Distance R to the virtual imaging center P dg

[0079] 2) Fixed D g Rotate line segment D along the flight path direction g Angle P Δθ g

[0080] 3) Calculate the endpoints after rotation The coordinates will as a new imaging center.

[0081] Step four: compensate echo data to new imaging center in real time

[0082] According to step three, new imaging center must be estimated after azimuth accumulation is completed, at this time the distance processing of echo data has been completed, including matched filtering, motion compensation and distance interpolation. The influence of changing imaging center is mainly reflected in both distance processing and azimuth processing, in which distance processing includes the change of instantaneous slant range R p (m) in motion compensation, the change of input parameters in distance interpolation, and azimuth processing includes the change of input parameters in azimuth interpolation.

[0083] The input parameters of distance interpolation mainly include instantaneous pitch angle θ g (m), instantaneous slant angle θ g (m) and instantaneous slant range R y (m). The coordinates of the grid to be interpolated need to be calculated by the above parameters in the interpolation process, and the change of the coordinates of the grid to be interpolated caused by the change of imaging center is usually less than one distance frequency unit, which can be ignored, that is, the new interpolation grid coordinates can be replaced by the nearest neighbor approximation of the original grid, avoiding re-interpolation.

[0084] Since distance interpolation is a process of extracting and rearranging the slant plane distance frequency domain echo after motion compensation according to the grid to be interpolated, the values on these grids are not changed. Therefore, if the imaging center needs to be changed, the key is to compensate the echo after distance motion compensation, that is, to correct it by using the new instantaneous slant range R . The specific process is as follows:

[0085] Save the instantaneous pitch angle θ g (m), the ground distance wave number K y (n), the instantaneous slant range R p (m) and other parameters in the distance processing process;

[0086] Calculate the instantaneous slant range R between the aircraft and the new imaging center

[0087] Multiply the echo s(m,n) after distance processing by the residual motion compensation phase function H moco_res (m,n) as the input of the subsequent azimuth processing.

[0088] s(m,n) = s(m,n) · H moco_res (m,n)

[0089]

[0090] Calculating the instantaneous squint angle of the ground plane under the new imaging center As an input parameter of subsequent azimuth interpolation, and complete the subsequent imaging processing work.

[0091] Figure 3 The same airborne spotlight SAR measured data is processed by different methods, and the imaging results are shown. Fig. (a) is the result of imaging by using a virtual imaging center, at this time the squint angle is 112°, under the influence of antenna installation error and inertial navigation error, it can be seen that the image center has exceeded the beam coverage range, the amplitude is weak, and there is obvious defocusing. Because the error of the virtual imaging center P is large, it leads to complete imaging failure in (a). Fig. (b) is the result of real-time imaging and compensation by using the imaging center estimated by the application, the estimated squint angle error is 3°, at this time the image is well focused and located in the beam center.

[0092] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A Doppler center fine compensation method for real-time imaging based on the PFA algorithm, characterized in that, Includes the following steps: Step 1: Determine the virtual imaging center by setting an ideal trajectory, and perform range processing on the original echo to obtain the range-processed echo; the method for range processing of the original echo is as follows: obtain the original echo by motion compensation. ;in, For motion compensation reference function, , The central wave number, f c For the center frequency, , n For the range wave number, The number of sampling points in the range direction represents the instantaneous slant range of the carrier aircraft. , This represents the number of sampling points in the azimuth direction. For the range sampling frequency, j It is the imaginary part of a complex number; The original echo, It is the echo after range processing; For reference slope distance; Step 2: Perform Doppler center estimation on the range-processed echo and calculate the ground plane oblique angle error; Step 3: Relocate the imaging center based on the actual flight path; Step 4: Based on the real-time compensation of the ground plane oblique angle error, the echo data is repositioned to the imaging center in Step 3, and the azimuth interpolation parameters are updated to complete the subsequent PFA imaging processing flow. The method for calculating the oblique angle error in step two includes the following steps: 1) When processing distance, open up Complex array of 2 rows and 2 columns Used to temporarily store echoes; 2) Starting with the first echo, input the range-processed echo into... In the first column, continue until all echoes are collected; 3) Starting from the second echo, multiply the current echo (processed with range) by the conjugate of the previous echo, and sum... Add the results from the second column and insert them into the input. In the second column, continue until all echoes are collected; 4) To The second column calculates the phase angle, takes the average, and then calculates the Doppler center frequency shift. Wherein, PRF is the pulse repetition frequency; 5) Calculate the error of the oblique angle on the ground plane. , λ is the average speed of the carrier aircraft within the entire aperture, and λ is the radar wavelength.

2. The Doppler center fine compensation method for real-time imaging based on the PFA algorithm according to claim 1, characterized in that, Step three involves calculating the actual beam center position using imaging geometry and using it as the new imaging center. This process includes the following steps: Calculate the distance from the actual synthetic aperture center projected onto the virtual imaging center on the ground; Using the actual synthetic aperture center projected onto the ground as the center and the distance from the actual synthetic aperture center projected onto the virtual imaging center as the radius, rotate the virtual imaging center along the flight path by an angle. ; The rotated endpoint serves as the new imaging center.

3. The Doppler center fine compensation method for real-time imaging based on the PFA algorithm according to claim 2, characterized in that, Step four involves correcting the azimuth interpolation parameters using the instantaneous slant range from the aircraft to the new imaging center. This includes the following steps: The instantaneous slant distance from the carrier aircraft to the new imaging center is calculated based on the instantaneous pitch angle, instantaneous slant angle on the ground plane, ground plane range wavenumber, and instantaneous slant distance during the range processing. Multiply the range-processed echo by the residual motion compensation phase function and use it as the echo input for subsequent azimuth processing; Calculate the instantaneous oblique angle of the ground plane below the new imaging center, use it as the input parameter for subsequent azimuth interpolation, and complete the subsequent imaging processing.

4. The Doppler center fine compensation method for real-time imaging based on the PFA algorithm according to claim 3, characterized in that, The residual motion compensation phase function is: in, The pitch angle is the instantaneous angle during the processing of the original echo range. The instantaneous oblique angle of the original echo ground plane. For the original echo ground plane distance wavenumber, The instantaneous slant range of the original echo. For the new imaging center The instantaneous slant distance, The oblique angle of the ground plane at the moment of aperture center.

Images