Multi-main-image tomography SAR (Synthetic Aperture Radar) baseline design and three-dimensional imaging method based on minimum redundancy

Abstract

The invention relates to a multi-main-image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy, and belongs to the technical field of synthetic aperture radars. The method comprises the following steps: step 1, carrying out baseline design of a multi-main image model based on a'minimum redundancy 'criterion; step 2, carrying out data acquisition according to the designed base line to obtain M SAR images; 3, performing interference processing on the data based on a multi-main image model to obtain a tomographic SAR data stack and a mapping relation; and step 4, performing height dimension estimation based on the tomographic SAR data stack and the mapping relation to obtain a tomographic SAR three-dimensional imaging result.

Description

Technical Field

[0001] The present invention relates to a multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy, belonging to the technical field of synthetic aperture radar. Background Technique

[0002] Synthetic Aperture Radar (SAR) is an active microwave remote sensing technology that can obtain the two-dimensional microwave scattering characteristics of the target area all day and all weather without being restricted by illumination conditions. Subsequently, this technology was extended to tomographic SAR. Tomographic SAR forms a synthetic aperture in the elevation direction through multiple observations at different spatial positions, providing resolution in the elevation direction. It can reconstruct the three-dimensional (azimuth, range, and elevation) scattering coefficients of the target through spectral estimation. Therefore, tomographic SAR solves the inherent layover problem in SAR imaging and is a strict three-dimensional imaging method. However, with the continuous improvement of corresponding scientific research task requirements, these applications require the imaging results of tomographic SAR three-dimensional imaging to be more comprehensive and accurate, thus posing more stringent requirements for tomographic SAR three-dimensional imaging technology.

[0003] Currently, limited by the number of repeat-pass flights and observation geometry, tomographic SAR three-dimensional imaging technology still faces many difficulties. With a limited number of observations, the sampling of tomographic SAR in the elevation direction does not satisfy the Nyquist law, and undersampling will lead to elevation ambiguity problems. At the same time, sparse sampling will also cause the sidelobes to rise, affecting the correct elevation extraction of the target. In addition, in tomographic SAR three-dimensional imaging, the viewing angle differences of different flight tracks are large, which will lead to serious spatial decorrelation problems, introduce phase noise, and thus reduce the elevation estimation accuracy. Currently, the research on tomographic SAR focuses on elevation estimation algorithms, using methods such as compressive sensing and spectral estimation to achieve sidelobe suppression, but there is a lack of research on signal models. The existing exploration of signal models is dual-station multi-master tomographic SAR, whose purpose is to correct the atmospheric phase and does not improve the sidelobe suppression ability. Summary of the Invention

[0004] In view of the above problems, the present invention proposes a multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy, which can achieve unambiguous, low-sidelobe, and high-precision elevation estimation under the condition of a limited number of observations, and ensure the three-dimensional imaging quality.

[0005] The present invention is realized through the following technical solutions:

[0006] In the first aspect, a multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy, the steps of the method include:

[0007] Step 1: Design the baseline of the multi-master image model based on the "minimum redundancy" criterion;

[0008] Step 2: Collect data according to the designed baseline to obtain M SAR images;

[0009] Step 3: Perform interferometric processing on the data based on the multi-master image model to obtain a tomographic SAR data stack and a mapping relationship;

[0010] Step 4: Perform height dimension estimation based on the tomographic SAR data stack and the mapping relationship to obtain a tomographic SAR three-dimensional imaging result.

[0011] Optionally, the baseline design of the multi-master image model based on the "minimum redundancy" criterion in the present invention is specifically as follows:

[0012] Determine the elevation sampling range according to the flight range of the SAR payload platform and baseline decorrelation;

[0013] Search for the possible distribution forms of elevation sampling within the elevation sampling range, and record the distribution form with the lowest redundancy, which is the baseline of the multi-master image model.

[0014] Optionally, the present invention sets the normalized sampling distribution form in the elevation direction as b = {b1, b2, …, b M}, where b i are all integers, and the elevation sampling distribution form is a corresponding multiple b×b0 of the unit baseline length b0. Search for the possible distribution forms of high-level sampling within the elevation sampling range, and record the distribution form with the lowest redundancy, which is the baseline of the multi-master image model.

[0015] Optionally, the minimum redundancy of b in the present invention is: The quantitative representation of the redundancy is:

[0016] b = min{R(b)}

[0017]

[0018] where M(M - 1) is the maximum number of pairwise combinations of M array elements, and H represents the degree of freedom of the array.

[0019] Optionally, the specific process of Step 3 in the present invention is as follows:

[0020] First, register the M SAR images collected, and take each of the registered M SAR images as the master image to perform interferometric processing with the remaining (M - 1) images to obtain M*(M - 1) interferograms;

[0021] Secondly, divide the interferograms belonging to the same master image into a group, and then perform deskewing processing on each group of interferograms. All the deskewed interferograms constitute the tomographic SAR data stack under the multi-master image model;

[0022] Finally, the mapping relationship between the tomographic SAR data stack and the spatial power spectrum under the multi-primary image model is obtained. Optionally, the mapping relationship between the tomographic SAR data stack and the spatial power spectrum in the present invention is:

[0023] η = Ψp + ε

[0024]

[0025] where η represents the tomographic SAR data stack, and η i represents the tomographic SAR data stack with the i-th image as the primary image, Ψ represents the mapping matrix, p represents the spatial power spectrum vector, and ε is the noise vector.

[0026] Optionally, the present invention uses the K-th image as the primary image, and based on the interferometric images of the K-th image and the remaining (M - 1) images, the autocorrelation function η(ξ' m,K ) is calculated;

[0027] η(ξ' m,K ) = ∫ Δs p(h)exp(-j2πξ' m,K h)dh, m ≠ K

[0028]

[0029] where h = s·sinθ represents the elevation perpendicular to the reference ground, r K represents the slant range of the K-th image, θ K represents the depression angle of the K-th image, p(h) represents the spatial power spectrum of the scattering coefficient, and b m,K represents the baseline between the m-th and K-th images;

[0030] Then, with the K-th image as the primary image, the mapping relationship based on the interferometric images of the K-th image and the remaining (M - 1) images is:

[0031] η k = Ψ k p + ε k

[0032] where, is an (M - 1)-dimensional observation vector;

[0033] [ψ k (h1), ψ k (h2),..., ψ k (h N )] is an (M - 1)×N-dimensional mapping matrix;

[0034] is the mapping vector corresponding to p(h i )

[0035] p = [p(h1), p(h2),..., p(h N )] T is the power spectrum vector in the N-dimensional space;

[0036] h i is the elevation perpendicular to the reference ground, and ε k is the noise vector.

[0037] In a second aspect, the present invention discloses a multi-master image tomography SAR baseline design and three-dimensional imaging device based on minimum redundancy, including a baseline design module, an image acquisition module, an image processing module, and a three-dimensional imaging module;

[0038] The baseline design module designs the baseline of the multi-master image model based on the "minimum redundancy" criterion;

[0039] The image acquisition module is used to collect data according to the designed baseline to obtain M SAR images;

[0040] The image processing module performs interference processing on the data based on the multi-master image model to obtain a tomography SAR data stack and a mapping relationship;

[0041] The three-dimensional imaging module estimates the height dimension based on the tomography SAR data stack and the mapping relationship to obtain the tomography SAR three-dimensional imaging result.

[0042] Beneficial effects

[0043] The present invention designs the baseline based on the "minimum redundancy" criterion and applies the multi-master image model for tomography SAR three-dimensional imaging, making the tomography SAR data stack have higher coherence and lower phase noise level, effectively suppressing the elevation ambiguity problem in tomography SAR three-dimensional imaging, reducing the sidelobe level of the elevation estimation result, and having higher elevation estimation accuracy. The computer simulation results and the actual test results show that this method is feasible and has excellent performance. Description of the drawings

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0045] Figure 1 It is an expansion schematic diagram of the traditional tomography SAR model (single-master image model) to the multi-master image model;

[0046] Figure 2 (a)~Figure 2 (b) is a schematic diagram of two track distributions designed in the embodiment of the present invention, where Figure 2 (a) is the distribution of uniform tracks in the elevation direction, Figure 2 (b) is the distribution of non-uniform tracks in the elevation direction;

[0047] Figure 3 (a) to Figure 3 (d) are the equivalent samplings in the elevation direction under four models in the embodiment of the present invention, where Figure 3 (a) is the single-master image model under uniform tracks, Figure 3 (b) is the single-master image model under minimum redundant tracks, Figure 3 (c) is the multi-master image model under uniform tracks, Figure 3 (d) is the multi-master image model under minimum redundant tracks;

[0048] Figure 4 (a) to Figure 4 (d) are the sets of elevation profiles obtained by the ISTA algorithm for elevation estimation under four models in the embodiment of the present invention. Where Figure 4 (a) is the set of elevation profiles of the single-master image model under uniform tracks, Figure 4 (b) is the set of elevation profiles of the single-master image model under minimum redundant tracks, Figure 4 (c) is the set of elevation profiles of the multi-master image model under uniform tracks, Figure 4 (d) is the set of elevation profiles of the multi-master image model under minimum redundant tracks. In each image, a column represents the elevation profile of a group of point targets. The abscissa is the true height of the moving scatterer among two scatterers. The ordinate is the estimated elevation value. Each elevation profile has been normalized according to its respective maximum peak. Detailed implementation manners

[0049] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0050] It should be noted that, without conflict, the following embodiments and the features in the embodiments may be combined with each other; and, based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present disclosure.

[0051] Note that the following description relates to various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is for illustrative purposes only. Based on this disclosure, those skilled in the art should 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 the aspects set forth herein can be used to implement a device and / or practice a method. Additionally, this device can be implemented and this method can be practiced using other structures and / or functionality in addition to one or more of the aspects set forth herein.

[0052] An embodiment of this application provides a multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy. The specific process is as follows:

[0053] Step 1: Design the baseline of the multi-master image model based on the "minimum redundancy" criterion;

[0054] Assume that the normalized sampling distribution form in the elevation direction is \(b = \{b_1, b_2, \ldots, b_{ M}\} \in Z\), where \(b_{ i}\) are all integers. The actual elevation sampling is the corresponding multiple \(b \times b_0\) of the unit baseline length \(b_0\). Therefore, for the baseline design under the multi-master image model, only a set of integers \(\{b_{ i}\} i＝1,2,...M needs to be determined such that the redundancy of \(b\) is minimized, that is:

[0055] \(b = \min\{R(b)\} \ (1)\)

[0056] where \(R(b)\) represents the redundancy of \(b\), which can quantitatively describe the redundancy characteristics of the array, and its definition is:

[0057]

[0058] where the numerator \(M(M - 1)\) is the maximum number of pairs of combinations of \(M\) array elements, and the denominator \(H\) represents the degrees of freedom (DOF) of the array, which is related to the \(b\) value distribution.

[0059] Formula (1) can be obtained by the computer exhaustive enumeration method. That is, first set a sampling range, and then search for possible permutations within the fixed sampling range, record the permutation form with the lowest redundancy. If the redundancy of this distribution form does not meet the requirements, modify the value of the sampling range and search again.

[0060] After determining the distribution form, only the sampling elevation range of the tomography SAR in the elevation direction needs to be determined to obtain the unit baseline length \(b_0\).

[0061] The sampling range of tomographic SAR in the elevation direction is determined by two factors. One is the range that the load platform can fly, and the other is baseline decorrelation. Baseline decorrelation means that when the viewing angle difference between the primary and secondary images in two observations is too large, the backscattering coefficient of the target is completely uncorrelated, and elevation estimation cannot be performed at this time.

[0062] The spatial baseline decorrelation function is

[0063]

[0064] where θ represents the down-viewing angle of the primary image observation, λ represents the wavelength, ρ s represents the range resolution in the range direction, and δθ represents the difference in the down-viewing angles of the two observations.

[0065] When ρ = 0, the data of the two observations are completely decorrelated, and interference and elevation estimation cannot be performed. The corresponding baseline is the critical baseline. Therefore, it is usually required that the baseline range is less than the critical baseline.

[0066] After comprehensively considering the flight range and baseline decorrelation to determine the elevation sampling range, within the sampling range, according to the normalized sampling distribution form based on minimum redundancy design, that is, the sampling positions of tomographic SAR in the elevation direction are determined, which is the baseline of the multi-primary image model.

[0067] Step 2: Collect data according to the designed baseline above to obtain M SAR images;

[0068] Assume that the radar has observed the target scene M times from different positions in space, and after performing imaging processing on the results of each observation, M SAR single-look complex images can be obtained.

[0069] The feasibility of using interference data to represent the tomographic SAR signal model is deduced and explained below:

[0070] Select the Kth image as the primary image, and perform registration and de-skew processing on other images. At this time, the observed signal of tomographic SAR, that is, the complex value of a single pixel point in the SAR image, can be expressed as:

[0071] g(ξ m ) = ∫ Δs γ(s)exp(-j2πξ m s)ds, m = 1, 2,..., M (4)

[0072] where Δs is the height range in the NSR (normal-slant-range) direction, γ(s) is the scattering coefficient, and ξ m is the spatial frequency corresponding to the height s in the NSR direction, and its expression is:

[0073]

[0074] Among them, r represents the slant range and λ is the wavelength. It should be noted that here b m does not represent the baseline, but the position of the sensor. Only in the special case where the position of the sensor corresponding to the master image b K = 0, b m is the same as the baseline. However, in order to maintain the same definition as the baseline of the interferometric SAR, here b m is still considered to be the position of the sensor, and the difference in the position of the sensor between it and the master image b m,K = b m - b K is the baseline.

[0075] Select the Kth image as the master image and perform interferometric processing on other images:

[0076] g(ξ m )g * (ξ K ) = ∫ Δs ∫ Δs' γ(s)γ(s') exp[-j2π(ξ m s - ξ K s')] ds ds', m ≠ K (7)

[0077] Among them, the superscript * represents the complex conjugate, γ(s') and s' are variable substitutions for the double integral, and have the same meaning as γ(s) and s above.

[0078] Assume that the scattering coefficient γ(s) follows a Gaussian distribution and is white, then its power spectral density is stationary, and there is

[0079] E{γ(s)γ * (s')} = p(s)δ(s - s') (8)

[0080] Among them, p(s) represents the spatial power spectrum of the scattering coefficient, and δ(s - s') represents the Dirac function.

[0081] At this time, the expectation of the observed data can be expressed as

[0082] E{g(ξ m )g * (ξ K )} = ∫ Δs p(s) exp[-j2π(ξ m - ξ K )s] ds, m ≠ K (9)

[0083] Define the baseline between the mth and Kth images as b m,K = b m - b K, and its corresponding spatial frequency is

[0084]

[0085] The expectation of the interference data, i.e., the autocorrelation function, can be expressed as:

[0086] η(ξ m,K ) = ∫ Δs p(s)exp(-j2πξ m,K s)ds (11)

[0087] where Δs is the height range in the NSR (normal - slant - range) direction, and ξ m,K represents the baseline b of the m - th and K - th images m,K corresponding to the spatial frequency. η(ξ m,K ) can be regarded as the sampling of the autocorrelation function of the original signal at different spatial positions. Due to the spatial stationarity of the signal, η(ξ m,K ) is independent of the absolute spatial position, but related to the spatial position difference (baseline), and its relationship with the spatial power spectrum p(s) of the scattering coefficient is still a Fourier transform relationship. Therefore, tomographic SAR three - dimensional imaging can be performed using interference data.

[0088] Step 3: Perform interference processing on the data based on the multi - master - image model to obtain a tomographic SAR data stack and a mapping relationship;

[0089] The extension of the tomographic SAR model (single - master - image model) to the multi - master - image model is as Figure 1 shown. Under the multi - master - image model, first, register M SAR images. Among the M registered SAR images, any one SAR image can be used as the master image and interfered with the other (M - 1) images, and a total of M*(M - 1) interferograms can be obtained.

[0090] Since formula (11) is the autocorrelation function in the elevation direction, it is necessary to correct formula (11), that is, convert the variable s in formula (11) to the elevation h. Taking the K - th image as the master image, we can get:

[0091] η(ξ' m,K ) = ∫ Δs p(h)exp(-j2πξ' m,K h)dh, m≠K (12)

[0092] where h = s·sinθ represents the elevation perpendicular to the reference ground. represents the autocorrelation function, r K represents the slant range of the K - th image, and θ K represents the depression angle of the K - th image.

[0093] Then, the interferograms are grouped according to their corresponding master images, and then the deskewing process is performed on each group of interferograms. The k-th group of data stacks uses the k-th SAR image as the master image, which includes M - 1 elements. According to formula (12), its mapping relationship can be expressed as

[0094] η k = Ψ k p + ε k (13)

[0095] where is the (M - 1)-dimensional observation vector.

[0096] [ψ k (h1), ψ k (h2),..., ψ k (h N )] is the (M - 1)×N-dimensional mapping matrix.

[0097] is the mapping vector corresponding to p(h i ).

[0098] p = [p(h1), p(h2),..., p(h N )] T is the N-dimensional spatial power spectrum vector.

[0099] ε k is the noise vector.

[0100] All the deskewed interferograms constitute the tomographic SAR data stack (i.e., the observed data) η under the multi-master image model.

[0101] Since all SAR images can be used as the master image, M groups of data stacks can be obtained. Concatenate all M groups of data stacks and the mapping matrix column by column:

[0102]

[0103] Then the mapping relationship between the observed data and the spatial power spectrum under the multi-master image model is obtained:

[0104] η = Ψp + ε (15)

[0105] Step 4: Perform height dimension estimation based on the tomographic SAR data stack and the mapping relationship to obtain the tomographic SAR three-dimensional imaging result.

[0106] The elevation estimation problem of tomographic SAR is essentially a spectral estimation problem for height dimension imaging. Therefore, based on the optimized tomographic SAR data stack and the mapping relationship, the compressive sensing method can be used to perform spectral estimation processing for the height dimension.

[0107] Another embodiment of the present application is a multi-master image tomography SAR baseline design and three-dimensional imaging device based on minimum redundancy, including a baseline design module, an image acquisition module, an image processing module, and a three-dimensional imaging module;

[0108] The baseline design module designs the baseline of the multi-master image model based on the "minimum redundancy" criterion;

[0109] The image acquisition module is used to collect data according to the designed baseline to obtain M SAR images;

[0110] The image processing module performs interference processing on the data based on the multi-master image model to obtain a tomography SAR data stack and a mapping relationship;

[0111] The three-dimensional imaging module performs height dimension estimation based on the tomography SAR data stack and the mapping relationship to obtain a tomography SAR three-dimensional imaging result.

[0112] Embodiment

[0113] The effectiveness of a multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy proposed by the present invention is verified by computer simulation.

[0114] The computer simulation is based on point targets, and a total of 81 groups of point targets are set. Each group contains two stacked scatterers, where the elevation of one scatterer is fixed at 0 m, and the elevation of the other scatterer linearly varies within the range of -20 m to +20 m.

[0115] To verify the proposed multi-master image tomography SAR model and baseline design method, we designed two track distributions. One is a uniform distribution, as shown in Figure 2 (a), and the other is a non-uniform distribution designed according to minimum redundancy, as shown in Figure 2 (b). Both track distributions have 7 tracks, and the track distribution ranges are the same, so they have the same Rayleigh resolution.

[0116] Under the two distributed tracks, processing is respectively performed based on the single-master image tomography SAR model (SM model) and the multi-master image tomography SAR model (MM model), and their equivalent sampling in the elevation direction is as shown in Figure 3 . As can be seen from Figure 3 , there is more redundancy in the equivalent sampling of the MM model with a uniform track distribution in the elevation direction, while the redundancy of the MM model with a minimum redundancy track in the equivalent sampling in the elevation direction is lower.

[0117] The minimum redundancy array requires that its equivalent virtual array is a non-porous array, and here the equivalent virtual array of the minimum redundancy track ( Figure 3(d) is not poreless. The purpose of this design is to make the ranges of uniformly distributed tracks and non-uniformly distributed tracks the same, so as to facilitate the performance comparison between the two.

[0118] The ISTA algorithm of the compressive sensing type is used for elevation estimation. As Figure 4 shown, there are four methods of simulation results for comparison: the ISTA method based on the SM model under the uniform baseline (ULB-SM-ISTA), the ISTA method based on the SM model under the minimum redundancy baseline (MR-SM-ISTA), the ISTA method based on the MM model under the uniform baseline (ULB-MM-ISTA), and the ISTA method based on the MM model under the minimum redundancy baseline (MRB-MM-ISTA). According to Figure 4 (d), it can be seen that the MM model under the non-uniform baseline has the advantages of suppressing elevation ambiguity and sidelobe level. This also verifies the effectiveness of the proposed MM model and the baseline design strategy based on minimum redundancy.

[0119] The performance of the elevation estimation results of the above four methods is evaluated. The integrated sidelobe ratio and the standard deviation of the estimated height of the point target and the true height are used as the evaluation criteria. The evaluation results are shown in Table 1.

[0120] Table 1 Performance evaluation results of four methods

[0121]

[0122] In summary, computer simulation proves the effectiveness of the method proposed in the present invention.

[0123] The above is only a preferred embodiment of the present invention, and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy, characterized in that The method steps include: Step 1: Design the baseline of the multi-master image model based on the "minimum redundancy" criterion; Step 2: Collect data according to the designed baseline to obtain M SAR images; Step 3: Perform interferometric processing on the data based on the multi-master image model to obtain a tomographic SAR data stack and a mapping relationship; Step 4: Perform height dimension estimation based on the tomographic SAR data stack and the mapping relationship to obtain the tomographic SAR three-dimensional imaging result.

2. The method for designing the baseline of multi-master image tomography SAR based on minimum redundancy and three-dimensional imaging according to claim 1, characterized in that The specific process of designing the baseline of the multi-master image model based on the "minimum redundancy" criterion is as follows: Determine the elevation sampling range according to the flight range of the SAR payload platform and baseline decorrelation; Search for the possible distribution forms of elevation sampling within the elevation sampling range, and record the distribution form with the lowest redundancy, which is the baseline of the multi-master image model.

3. The method for designing a multi-master image tomography SAR baseline and three-dimensional imaging based on minimum redundancy according to claim 2, wherein Let the normalized sampling distribution form in the elevation direction be \(b = \{b_1, b_2, \ldots, b\) M \}, where \(b\) i are all integers, and the elevation sampling distribution form is the corresponding multiple \(b\times b_0\) of the unit baseline length \(b_0\). Search for the possible distribution forms of high-level sampling within the elevation sampling range, and record the distribution form with the lowest redundancy, that is, the baseline of the multi-master image model.

4. The multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy according to claim 3, wherein The minimum redundancy of b is: The quantitative representation of the redundancy is: b = min{R(b)} where M(M - 1) is the maximum number of pairwise combinations of M array elements, and H represents the degrees of freedom of the array.

5. The method for designing the baseline of multi-master image tomography SAR based on minimum redundancy and three-dimensional imaging according to claim 1, characterized in that The specific process of Step 3 is as follows: First, register the M collected SAR images. Take each of the registered M SAR images as the master image and perform interferometric processing with the remaining (M - 1) images to obtain M*(M - 1) interferograms; Second, divide the interferograms belonging to the same master image into a group, and then perform de-skewing processing on each group of interferograms. All the de-skewed interferograms constitute the tomographic SAR data stack under the multi-master image model; Finally, obtain the mapping relationship between the tomographic SAR data stack and the spatial power spectrum under the multi-master image model.

6. The multi-master image tomography SAR baseline design and three-dimensional imaging method based on minimum redundancy according to claim 5, characterized in that The mapping relationship between the tomographic SAR data stack and the spatial power spectrum is: η = Ψp + ε where η represents the tomographic SAR data stack, η i represents the tomographic SAR data stack with the i-th image as the main image, Ψ represents the mapping matrix, p represents the spatial power spectrum vector, and ε is the noise vector.

7. The method for designing a baseline of a multi-master image tomography SAR based on minimum redundancy and three-dimensional imaging according to claim 6, characterized in that Taking the K-th image as the main image, based on the interference images of the K-th image and the remaining (M - 1) images, calculate the autocorrelation function η(ξ' m,K ); η(ξ' m,K ) = ∫ Δs p(h) exp(-j2πξ' m,K h) dh, m ≠ K where h = s·sinθ represents the elevation perpendicular to the reference ground, r K represents the slant range of the K-th image, and θ K represents the depression angle of the K-th image, p(h) represents the spatial power spectrum of the scattering coefficient, and b m,K represents the baseline between the m-th and K-th images; Then, taking the Kth image as the master image, the mapping relationship based on the interferometric image between the Kth image and the remaining (M - 1) images is: η k = Ψ k p + ε k Among them, is an (M - 1)-dimensional observation vector; [ψ k (h1), ψ k (h2), …, ψ k (h N )] is an (M - 1)×N dimensional mapping matrix; is the mapping vector corresponding to p(h i ); p = [p(h1), p(h2),..., p(h N )] T is the N - dimensional space power spectrum vector; h i is the elevation perpendicular to the reference ground, and ε k is the noise vector.

8. A multi-master image tomography SAR baseline design and three-dimensional imaging device based on minimum redundancy, characterized in that It includes a baseline design module, an image acquisition module, an image processing module, and a three-dimensional imaging module; The baseline design module designs the baseline of the multi-master image model based on the "minimum redundancy" criterion; The image acquisition module is used to collect data according to the designed baseline to obtain M SAR images; The image processing module performs interferometric processing on the data based on the multi-master image model to obtain a tomographic SAR data stack and a mapping relationship; The three-dimensional imaging module performs height dimension estimation based on the tomographic SAR data stack and the mapping relationship to obtain the tomographic SAR three-dimensional imaging result.

Images