Satellite-borne SAR scene matching curve frequency domain imaging preprocessing method oriented to time-varying repetition frequency sampling
By calculating the deskewing factor, zero-filling number, and amplitude factor for upsampling, the problem of low efficiency in traditional spaceborne SAR time-varying repetition rate data preprocessing is solved, achieving efficient data preprocessing and imaging quality improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING INST OF TECH
- Filing Date
- 2025-12-29
- Publication Date
- 2026-04-14
AI Technical Summary
Traditional spaceborne SAR scene matching curve imaging time-varying repetition rate data preprocessing methods are inefficient, resulting in severe echo obstruction loss and affecting imaging quality.
A spaceborne SAR scene matching curve frequency domain imaging preprocessing method oriented towards time-varying repetition sampling is adopted. By calculating the deskewing factor, the number of zero-points, and the amplitude factor, upsampling is performed to avoid interpolation operations, adjust the time-frequency relationship, and compensate for spectral amplitude differences.
It achieves efficient data preprocessing, avoids echo occlusion loss, improves imaging quality and efficiency, and reduces computational load.
Smart Images
Figure CN121856967A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a frequency domain imaging preprocessing method for spaceborne SAR scene matching curves oriented to time-varying repetition rate sampling, belonging to the field of Synthetic Aperture Radar (SAR) technology. Background Technology
[0002] Spaceborne synthetic aperture radar (SAR) is a widely used active microwave remote sensing device. It features all-weather, all-day operation and high two-dimensional resolution. It is an important technical component of remote sensing and has significant implications for applications such as disaster monitoring, land mapping, and construction planning.
[0003] Spaceborne SAR transmits and receives linear frequency modulated (LFM) pulse signals at regular time intervals. The number of pulses transmitted per second is typically called the Pulse Repetition Frequency (PRF). Traditional spaceborne SAR, with its limited slant range and viewing angle, often uses a fixed PRF for data transmission and reception. However, as the viewing angle of spaceborne SAR increases, the range of echo delay variation also increases. Spaceborne SAR scene-matching curve imaging is a more flexible imaging method that generates a curved imaging band by continuously adjusting the radar beam direction, resulting in more drastic changes in echo delay. Using a traditional fixed PRF for data acquisition can easily lead to echo obstruction loss, severely affecting image quality. In such cases, time-varying PRF technology, also known as segmented PRF technology, is often used to adjust the pulse PRF to adapt to different viewing angles and avoid echo obstruction loss.
[0004] Time-varying repetition rate (RFR) data typically requires preprocessing before frequency domain imaging. Traditional methods usually involve interpolation to obtain uniformly sampled data, followed by deramp upsampling to obtain azimuth-free spectral aliasing data. However, due to the enormous computational cost of data interpolation, traditional preprocessing methods are extremely inefficient.
[0005] Therefore, a new upsampling preprocessing method for time-varying repetition rate data in spaceborne SAR curve imaging is needed. Summary of the Invention
[0006] The purpose of this invention is to overcome the shortcomings of existing spaceborne SAR scene matching curve frequency domain imaging time-varying repetition frequency data preprocessing and solve the problem of low interpolation efficiency of time-varying repetition frequency data. A spaceborne SAR scene matching curve frequency domain imaging preprocessing method oriented towards time-varying repetition frequency sampling is designed.
[0007] The method of this invention is achieved through the following technical solution: In a first aspect, the present invention provides a frequency domain imaging preprocessing method for spaceborne SAR scene matching curves oriented towards time-varying repetition rate sampling, the specific process of which is as follows: Step 1: When the spaceborne SAR performs scene matching curve imaging, it transmits a linear frequency modulated signal with segmented pulse repetition frequency and performs range preprocessing on the received echo signal. Step 2: Calculate the azimuth upsampling parameters for each segment of repetition frequency data, including: deskewing factor, number of zero-padded points, and amplitude factor; Step 3: Select the echo data corresponding to the current distance frequency and the corresponding azimuth upsampling parameters; Step 4: Based on the azimuth upsampling parameters, upsample each segment of repetition frequency data in the echo data corresponding to the current distance frequency and then accumulate them; Step 5: Update distance frequency. Repeat steps 3 and 4 until all data preprocessing is complete.
[0008] Optionally, the descrambling factor is fdrc_k_i, the number of zero-padding points is N_k_i, and the amplitude factor is A_k_i; the specific process of step four is as follows: (1) Based on the deskewing factor fdrc_k_i, construct a deskewing filter, perform deskewing filtering on the data S0i corresponding to the i-th repetition frequency in the echo data corresponding to the k-th distance frequency point fr(k), obtain data S1i, and then symmetrically pad zeros on both sides of it. After padding, the number of data S2i points is N_k_i. (2) The data S2i is processed by the azimuth time delay compensation function, and then the azimuth time-frequency transformation is performed to obtain the data S3i; (3) Use the deramp function to transform the data S3i to a new azimuth time domain, and then symmetrically pad the transformed data with zeros on both sides so that the number of points in the zero-padding data S4i is the number of points Np after upsampling; (4) Construct the azimuth frequency recovery function based on the deskewing factor fdrc_k_i, multiply it with the data S4i to obtain the data S5i; perform upsampling processing on the data S5i by compensating the amplitude factor to obtain the data S6i and accumulate it to obtain the upsampled data corresponding to the current distance frequency.
[0009] Optionally, the specific process of step two in this invention is as follows: First, based on the variation range of the Doppler center frequency sequence fdc_line(ta) of the wave signal and the instantaneous Doppler bandwidth of the echo, the expected upsampling pulse repetition frequency PRFp is set; according to the expected upsampling pulse repetition frequency PRFp and the Deramp upsampling time-frequency relationship, the number of upsampled zero-padding points N_k_i corresponding to the i-th repetition frequency data at the k-th distance frequency point fr(k) is calculated. , M is the number of repetitive data segments, and Nr is the number of distance points; Secondly, the deskewing factor is updated based on the rounded N_k_i to obtain the new deskewing factor corresponding to the k-th distance frequency point fr(k) for the i-th repetition frequency data: Finally, the amplitude compensation factor A_k_i is calculated based on the number of upsampled points Np and the number of upsampled points N_k_i with zero padding.
[0010] Optionally, the number of points N_k_i after upsampling and zero-padding in this invention is:
[0011] Where, round means rounding to the nearest integer, PRFi means the pulse repetition frequency corresponding to the i-th repetition frequency data, and fdrc_k means the descrambling factor.
[0012] Optionally, the descrambling factor corresponding to the new i-th repetition frequency data at the k-th distance frequency point fr(k) in this invention is:
[0013] Where sign() means to take the sign; Optionally, the amplitude compensation factor A_k_i described in this invention is:
[0014] PRFmax represents the maximum repetition rate in the time-varying repetition rate data.
[0015] Optionally, the descrambling filter constructed according to the present invention is as follows:
[0016] Where ta_cut is the original azimuth time axis corresponding to the i-th repetition frequency data. Optionally, the azimuth delay compensation function of the present invention has the following expression:
[0017] Where Tnewi is the delay on the time axis of data S2i, Torgi is the start time of each segment of variable repetition frequency data, and fa_i is the azimuth frequency axis after zero padding.
[0018] Optionally, the deramp function expression of this invention is:
[0019] Where tap_i is the time axis of data S3i.
[0020] Optionally, the expression for the recovery function described in this invention is:
[0021] Where fap is the new frequency domain azimuth frequency axis.
[0022] Beneficial effects: The present invention adopts the upsampling method with variable deskewing factor described in step two. It sets corresponding deskewing factors according to different repetition frequencies, and then adjusts the time-frequency relationship in the upsampling process through the deskewing factor, so that data with different pulse repetition frequencies can be upsampled to the same interval, thereby avoiding the interpolation operation required in traditional upsampling processing methods and achieving efficient preprocessing.
[0023] This invention employs the amplitude factor compensation method for time-varying repetition frequency described in step two. Based on Fresnel's formula, it calculates the spectral amplitude difference of different repetition frequency data during the upsampling process and compensates for this value during the spectral superposition process. This solves the splicing error problem caused by the inconsistency of azimuth spectral amplitude in the upsampling process of time-varying repetition frequency data and achieves accurate preprocessing. Attached Figure Description
[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0025] Figure 1 This is a flowchart of a rapid upsampling preprocessing method for spaceborne SAR time-varying repetition rate data; Figure 2 It is a spaceborne SAR simulation observation configuration and point target distribution; Figure 3 It is a spaceborne SAR simulation of variable repetition frequency sequence and echo amplitude; Figure 4 This is the spectrum of unsampled echoes at different repetition rates in spaceborne SAR simulation; Figure 5 These are the echo spectrum and array target imaging results after upsampling in spaceborne SAR simulation; Figure 6 This is the imaging evaluation result of the center point of the spaceborne SAR simulation scene; Detailed Implementation The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0026] It should be noted that, unless otherwise specified, the following embodiments and features can be combined with each other; and, based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.
[0027] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.
[0028] This invention is a frequency domain imaging preprocessing method for spaceborne SAR scene matching curves oriented towards time-varying repetition rate sampling. The specific process is as follows: Step 1: When the spaceborne SAR performs scene matching curve imaging, it transmits a linear frequency modulated signal with segmented pulse repetition frequency and performs range preprocessing on the received echo signal. Step 2: Calculate the azimuth upsampling parameters for each segment of repetition frequency data, including: deskewing factor, number of zero-padded points, and amplitude factor; Step 3: Select the echo data corresponding to the current distance frequency and the corresponding azimuth upsampling parameters; Step 4: Based on the azimuth upsampling parameters, upsample each segment of repetition frequency data in the echo data corresponding to the current distance frequency and then accumulate them; Step 5: Update distance frequency. Repeat steps 3 and 4 until all data preprocessing is complete.
[0029] The following provides a detailed explanation of the specific implementation process for each of the above steps, as shown in the flowchart below. Figure 1 As shown, it is specifically as follows: Step 1: Echo distance gate alignment, motion removal, and distance fast Fourier transform processing.
[0030] While the radar satellite moves along its orbit, the onboard SAR radar payload uses a linear frequency modulated signal as the transmitted signal to observe the ground target P. The transmitted pulse repetition frequency varies in segments, and the expression for the demodulated radar echo signal is as follows: (1) Where c is the speed of light, tr represents the range-fast time, and ta represents the azimuth-slow time, corresponding to a non-uniform azimuth-time axis. R(ta) is the instantaneous slant range between the radar antenna phase center and the target P, and Rmin(ta) is the initial slant range of the echo reception window at different azimuth times. tp is the center time when the beam illuminates the target P, fc is the signal carrier frequency, Kr is the transmission signal modulation frequency, Tp is the signal pulse width, and Tsyn is the target illumination time. Equation (1) omits the amplitude term of the echo signal.
[0031] Range preprocessing is performed on the echo signal. First, the range time domain range after echo range gate alignment and de-travel is calculated. Zero-padding is then performed in the range time domain based on the current and expected range time domain ranges to ensure no aliasing in the range time domain range after preprocessing. Assume that the number of echo azimuth points Na and the number of range points Nr after range preprocessing are both integer multiples of 2.
[0032] Then, a range-to-fast Fourier transform is performed on the echo to obtain the range-frequency domain expression of the echo signal: (2) Where fr represents the distance frequency, and its expression is: (3) Where Fs is the range-direction signal sampling rate, -Fs / 2≤fr≤Fs / 2.
[0033] To perform distance gate alignment and motion removal, the distance gate alignment filter Fgate is constructed as follows: (4) Where Rmin0 is the initial slant range of the aligned echo reception window. The de-migrating filter Frw is constructed as follows: (5) Where Vs is the satellite velocity and θ is the radar center angle.
[0034] Multiplying equations (4) and (5) by equation (2), we obtain the expression for the filtered signal as follows: (6) At this point, the echo signal has completed echo distance gate alignment, de-travel, and distance fast Fourier transform processing, and the Doppler center frequency has also been compensated.
[0035] Step 2: Calculate the azimuth upsampling parameters for each segment of variable repetition frequency data, including the deskewing factor, the number of zero-filling points, and the amplitude factor.
[0036] Assuming the input echo signal has M repetition frequency data segments, let PRFi (1≤i≤M) be the pulse repetition frequency corresponding to the i-th repetition frequency data segment, and let pst_i be the starting index and ped_i be the ending index of the i-th repetition frequency data segment in terms of azimuth. Then, the starting time Torgi of each variable repetition frequency data segment is: (7) Assuming the Doppler center frequency sequence of the input echo signal is fdc_line(ta), the expected upsampling pulse repetition frequency PRFp is set based on the variation range of the Doppler center frequency fdc_line(ta) and the instantaneous Doppler bandwidth of the echo. Then, the overall echo deskewing factor fdrc is calculated based on the variation range of the Doppler center frequency fdc_line(t) and the total echo azimuth time Ta. Taking the sliding convergence mode as an example, the deskewing factor calculation method is as follows: (8) If the data is in TOPS mode, then the descrambling factor is negative.
[0037] Since the echo data is located in the range frequency domain, the overall deskewing factor of the echo corresponding to the range frequency domain is calculated. The expression for the deskewing factor corresponding to the k-th range frequency point fr(k) is as follows: (9) Based on the Deramp upsampling time-frequency relationship, the number of upsampling points corresponding to the k-th distance frequency point fr(k) in the case of the i-th repetition frequency is obtained as follows: (10) Where, round means rounding to the nearest integer. N_k_i is the number of upsampled and zero-padded points corresponding to the i-th repetition frequency data at the k-th distance frequency point fr(k). It is a two-dimensional variable parameter, and at the same time, N_k_i is guaranteed to be an integer multiple of 2 for easy processing.
[0038] Since upsampling requires maintaining consistency in the time-frequency relationship, the deskewing factor needs to be updated based on the rounded N_k_i, thus obtaining a new deskewing factor: (11) Here, sign() represents taking the positive or negative sign, and fdrc_k_i is the descrambling factor corresponding to the i-th repetition frequency data at the k-th distance frequency point fr(k), which is a two-dimensional variable parameter.
[0039] Because the echo data needs to maintain consistent data length after preprocessing for easier subsequent processing, zero-padding in the azimuth direction is necessary. At this point, the total number of azimuth points can be padded based on the subsequent processing requirements. If there is no need to convert to the azimuth time domain, the number of upsampled points can be calculated based on N_k_i. (12) Where max represents the maximum value. If conversion to the azimuth-time domain is required, the number of upsampled points can be calculated based on the azimuth-time and PRFp: (13) Where ceil represents rounding up, Ta is the total azimuth duration of the input data, and PRFp is the pulse repetition frequency after upsampling.
[0040] The deskewing factor and number of points differ for different segments of the repetition frequency data. According to the Fresnel formula, the expression for the azimuth spectrum amplitude gain after upsampling is as follows: (14) Where Ka represents the azimuth modulation frequency. When the echo of the same target exists in different variable repetition frequency (VRF) data segments, the azimuth modulation frequency Ka is the same. Different deskewing factors and zero-padding numbers will lead to differences in the final spectral amplitude. Therefore, the amplitude compensation factor A_k_i is calculated as follows: (15) PRFmax represents the maximum repetition rate in the time-varying repetition rate data.
[0041] Step 3: Select the echo data and azimuth upsampling parameters corresponding to the current distance frequency.
[0042] Select the k-th distance frequency point fr(k) to be processed, 1≤k≤Nr, and extract the corresponding echo data with a size of 1×Na. Select the azimuth upsampling parameters for each repetition frequency corresponding to the k-th distance frequency point, namely N_k_i, fdrc_k_i, A_k_i, where 1≤i≤M.
[0043] Step 4: Upsample each segment of variable repetition frequency data and then sum them up.
[0044] (1) For the echo data corresponding to the k-th distance frequency point fr(k) in the input, for processing the data corresponding to the i-th repetition frequency (1≤i≤M), first extract this segment of data and denote it as S0i, and denote the number of points of S0i as Ni. Construct a descrambling filter: (16) Where ta_cut is the original azimuth time axis corresponding to the i-th repetition frequency data, and its expression is: (17) Where pst_i is the starting index of the i-th repetition frequency data in the echo data azimuth, and ped_i is the ending index.
[0045] Multiply the data S0i corresponding to the i-th repetition frequency (1≤i≤M) by the descrambling filter Hdechirp to obtain the data S1.
[0046] Then, symmetrical zero-padding is performed on both sides of data S1i (i.e., zero-padding is performed with data S1i as the center), so that the number of points in data S2i after zero-padding is N_k_i, and this number of points is denoted as Nt. The starting index kst and ending index ked of data S1i on data S2i after zero-padding are: (18) Next, the delay Tnewi of the data S1i on the zero-padded time axis is calculated, and the zero-padded azimuth time axis and the zero-padded azimuth frequency axis are defined as follows: (19) Get the delay Tnewi: (20) (2) Perform a fast Fourier transform on the data S2i, and then multiply it by the azimuth time delay compensation function. The function expression is: (twenty one) The data is then subjected to an inverse fast Fourier transform (IFFT) to obtain the data after azimuth time delay compensation. Following this, an azimuth time-frequency transform is performed. If the data is in sliding clustering mode, an azimuth IFFT is performed. If the data is in TOPS mode, an inverse azimuth IFFT is performed. The data after the azimuth time-frequency transform is denoted as S3i. The azimuth data is transformed to a new azimuth time domain, and the new time-domain azimuth time axis is defined as follows: (twenty two) (3) Perform deramp processing on the data S3i, multiplying by the deramp function expression: (twenty three) Next, zero-padding is performed on the deramped data S3i, centering on it, so that the number of points in the zero-padding data S4i is Np. The starting index Lst and ending index Led of the zero-padding data S3i on the data S4i are: (twenty four) (4) Afterwards, the data S4i undergoes a fast Fourier transform for azimuth. At this time, the azimuth data is transformed into a new azimuth frequency domain, and the new frequency domain azimuth frequency axis is defined as: (25) Multiplying this by the azimuth frequency restoration function, the expression is: (26) At this point, we obtain data S5i. Then, we perform upsampling on data S5i by compensating for the amplitude factor, resulting in data S6i. (27) At this point, the upsampling process for a single repetition frequency is completed. Then, upsampling is performed on the repetition frequency data for other segments. Finally, the data S6i is accumulated with other upsampled variable repetition frequency data to obtain the upsampled data corresponding to the current distance frequency. (28) Step 5: Update distance frequency. Repeat steps 3 and 4 until all data preprocessing is complete.
[0047] The next range frequency point is updated, and the corresponding echo data and azimuth upsampling parameters are selected. Following steps three and four, azimuth upsampling processing is performed at the corresponding range frequency. Finally, all Nr range frequency points and their corresponding echo data are traversed to complete the upsampling preprocessing. Thus, the upsampled, aliasing-free echo data two-dimensional spectrum D(fr, fap) is obtained. This spectrum can be directly processed according to subsequent imaging requirements, or Fourier transformed to restore it to the two-dimensional time domain before processing.
[0048] Example: Table 1 shows some of the simulation parameters for the required spaceborne SAR scene matching curve imaging.
[0049] Table 1. List of key parameters for SAR satellites
[0050] To verify the feasibility of the proposed rapid upsampling preprocessing method for time-varying repetition rate data of spaceborne SAR, computer simulations were performed using the parameters in Table 1. The simulation observation configuration was a scene-matching curve imaging sliding spotting mode, as shown in the figure. Figure 2 As shown in (a), the distribution of the dot matrix targets is as follows: Figure 2 As shown in (b), the azimuth spacing between adjacent targets is 1 km, and the range spacing is 1 km. The time-varying repetition rate distribution of the echo data is as follows: Figure 3 (a) As shown in Table 2. Based on the above parameters, point target simulation imaging is performed, and the amplitude diagram of the generated echo is shown below. Figure 3 As shown in (b), the two-dimensional spectra of echo data segments with different repetition frequencies before upsampling are as follows: Figure 4 As shown, some of the spectrum exhibits folding.
[0051] Table 2. Time-varying repetition frequency values and corresponding point numbers
[0052] After preprocessing, the number of echo points Np is 9956, and the azimuth frequency PRFp obtained by upsampling is 7081.7Hz. The echo spectrum after upsampling is as follows. Figure 5 As shown in (a), the different frequency spectra are well spliced together, and the overall spectrum is not folded.
[0053] The ANCS algorithm is used as the frequency domain processing method to perform imaging processing on the preprocessed echo, and the resulting dot matrix imaging result is as follows: Figure 5 As shown in (b). The imaging results of target 5 at the scene center point are evaluated, and the imaging results at the scene center point are as follows. Figure 6 As shown in (a), the azimuth profile is as follows Figure 6 As shown in (b), the distance from the profile is as follows Figure 6 As shown in (c), the evaluation results are shown in Table 3, and the imaging results show no defocus.
[0054] Table 3. Imaging evaluation results of the center point of the simulation scene
[0055] Simulation results in the embodiments demonstrate that the proposed spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition sampling is feasible. It can effectively support frequency domain processing of curve imaging, and the processing does not involve interpolation operations, thus avoiding a large amount of computational consumption.
[0056] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A preprocessing method for frequency domain imaging of spaceborne SAR scene matching curves oriented towards time-varying repetition rate sampling, characterized in that, The specific process is as follows: Step 1: When the spaceborne SAR performs scene matching curve imaging, it transmits a linear frequency modulated signal with segmented pulse repetition frequency and performs range preprocessing on the received echo signal. Step 2: Calculate the azimuth upsampling parameters for each segment of repetition frequency data, including: deskewing factor, number of zero-padded points, and amplitude factor; Step 3: Select the echo data corresponding to the current distance frequency and the corresponding azimuth upsampling parameters; Step 4: Based on the azimuth upsampling parameters, upsample each segment of repetition frequency data in the echo data corresponding to the current distance frequency and then accumulate them; Step 5: Update distance frequency. Repeat steps 3 and 4 until all data preprocessing is complete.
2. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 1, characterized in that, The deskewing factor is fdrc_k_i, the number of zero-padded points is N_k_i, and the amplitude factor is A_k_i; the specific process of step four is as follows: (1) Based on the deskewing factor fdrc_k_i, construct a deskewing filter, perform deskewing filtering on the data S0i corresponding to the i-th repetition frequency in the echo data corresponding to the k-th distance frequency point fr(k), obtain data S1i, and then symmetrically pad zeros on both sides of it. After padding, the number of data S2i points is N_k_i. (2) The data S2i is processed by the azimuth time delay compensation function, and then the azimuth time-frequency transformation is performed to obtain the data S3i; (3) Use the deramp function to transform the data S3i to a new azimuth time domain, and then symmetrically pad the transformed data with zeros on both sides so that the number of points in the zero-padding data S4i is the number of points Np after upsampling; (4) Construct the azimuth frequency recovery function based on the deskewing factor fdrc_k_i, multiply it with the data S4i to obtain the data S5i; perform upsampling processing on the data S5i by compensating the amplitude factor to obtain the data S6i and accumulate it to obtain the upsampled data corresponding to the current distance frequency.
3. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 2, characterized in that, The specific process of step two is as follows: First, based on the variation range of the Doppler center frequency sequence fdc_line(ta) of the wave signal and the instantaneous Doppler bandwidth of the echo, the expected upsampling pulse repetition frequency PRFp is set; according to the expected upsampling pulse repetition frequency PRFp and the Deramp upsampling time-frequency relationship, the number of upsampled zero-padding points N_k_i corresponding to the i-th repetition frequency data at the k-th distance frequency point fr(k) is calculated. , M is the number of repetitive data segments, and Nr is the number of distance points; Secondly, the deskewing factor is updated based on the rounded N_k_i to obtain the new deskewing factor corresponding to the k-th distance frequency point fr(k) for the i-th repetition frequency data: Finally, the amplitude compensation factor A_k_i is calculated based on the number of upsampled points Np and the number of upsampled points N_k_i with zero padding.
4. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The number of points N_k_i after upsampling and zero-padding is: Where, round means rounding to the nearest integer, PRFi means the pulse repetition frequency corresponding to the i-th repetition frequency data, and fdrc_k means the descrambling factor.
5. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The deskewing factor corresponding to the new i-th repetition frequency data at the k-th distance frequency point fr(k) is: The sign() function is used to determine the sign of a symbol.
6. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The amplitude compensation factor A_k_i is: PRFmax represents the maximum repetition rate in the time-varying repetition rate data.
7. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The descrambling filter is constructed as follows: Where ta_cut is the original azimuth time axis corresponding to the i-th repetition frequency data.
8. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The azimuth delay compensation function has the following expression: Where Tnewi is the delay on the time axis of data S2i, Torgi is the start time of each segment of variable repetition frequency data, and fa_i is the azimuth frequency axis after zero padding.
9. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The expression for the deramp function is: Where tap_i is the time axis of data S3i.
10. The spaceborne SAR scene matching curve frequency domain imaging preprocessing method for time-varying repetition rate sampling according to claim 3, characterized in that, The expression for the recovery function is: Where fap is the new frequency domain azimuth frequency axis.