WO2023045431A1 - 逆合成孔径雷达成像方法、装置、电子设备及存储介质 - Google Patents

逆合成孔径雷达成像方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2023045431A1
WO2023045431A1 PCT/CN2022/099364 CN2022099364W WO2023045431A1 WO 2023045431 A1 WO2023045431 A1 WO 2023045431A1 CN 2022099364 W CN2022099364 W CN 2022099364W WO 2023045431 A1 WO2023045431 A1 WO 2023045431A1
Authority
WO
WIPO (PCT)
Prior art keywords
matrix
initial value
synthetic aperture
echo signal
aperture radar
Prior art date
Application number
PCT/CN2022/099364
Other languages
English (en)
French (fr)
Inventor
徐刚
张邦杰
张慧
黄岩
洪伟
郭坤鹏
冯友怀
Original Assignee
南京隼眼电子科技有限公司
东南大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京隼眼电子科技有限公司, 东南大学 filed Critical 南京隼眼电子科技有限公司
Priority to KR1020247007856A priority Critical patent/KR20240039054A/ko
Publication of WO2023045431A1 publication Critical patent/WO2023045431A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88Radar or analogous systems specially adapted for specific applications
    • G01S13/89Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/9004SAR image acquisition techniques
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88Radar or analogous systems specially adapted for specific applications
    • G01S13/89Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/904SAR modes
    • G01S13/9064Inverse SAR [ISAR]

Definitions

  • the present application relates to the technical field of radar signal processing, and in particular to an inverse synthetic aperture radar imaging method, device, electronic equipment and storage medium.
  • Inverse Synthetic Aperture Radar (Inverse Synthetic Aperture Radar, referred to as ISAR), as a main tool of microwave remote sensing, can provide high-resolution images of objects in the sky, space, ocean and other scenes, and has the ability to work all day and all day.
  • ISAR Inverse Synthetic Aperture Radar
  • the traditional ISAR imaging technology obtains the range-Doppler image by coherently accumulating echo data to obtain higher resolution and output signal-to-noise ratio (SNR).
  • SNR signal-to-noise ratio
  • higher resolution ISAR imaging requires the signal at the transmitter to have a wide frequency band, which requires more synthetic aperture time.
  • problems such as high system complexity, large amount of data, and complex target motion, and under non-ideal conditions such as the exchange of the working status of the multi-function radar or the maneuvering motion of the observed target, only part of the echo data can be obtained ( Incomplete echo data can be called sparse aperture echo data), so in this case it is difficult to achieve unambiguous imaging only by using traditional methods.
  • the embodiment of the present application provides an inverse synthetic aperture radar imaging method, device, electronic equipment, and storage medium to solve the problem of low ISAR imaging resolution at low sampling rates in the prior art, and to improve ISAR imaging at low sampling rates. resolution.
  • This application provides an inverse synthetic aperture radar imaging method, including:
  • the step of receiving the echo signal corresponding to the chirp signal transmitted by the radar, and preprocessing the echo signal to obtain the echo data includes:
  • the echo data is expressed as:
  • ⁇ k represents the product of the scattering coefficient of the k-th scattering point and the rectangular window
  • x k represents the abscissa of the k-th scattering point at the initial time
  • y k represents the ordinate of the k-th scattering point at the initial time
  • Indicates the frequency modulation slope
  • c indicates the propagation speed of electromagnetic waves
  • f c indicates the center frequency of the carrier
  • indicates the target speed
  • f s indicates the sampling frequency
  • PRF indicates the pulse repetition frequency
  • j indicates the imaginary number symbol
  • m 1, 2,...
  • M Indicates fast time series numbers in discrete form
  • m 1, 2, ...
  • N indicates slow time series numbers in discrete form
  • S(m, n) indicates sparsely sampled two-dimensional echo data.
  • the structured low-rank matrix is a two-layer Hankel matrix.
  • demodulated echo signal is expressed as:
  • ⁇ k , x k and y k respectively represent the scattering coefficient of the kth scattering point and the horizontal and vertical coordinates at the initial moment
  • T p represents the pulse width
  • represents the frequency modulation slope
  • R 0 represents the distance between the target rotation center and the radar
  • R ref represents the reference distance for de-chirping
  • f c represents the center frequency of the carrier
  • c represents the propagation speed of electromagnetic waves
  • represents the target rotational speed
  • t r represents the fast time
  • t m represents the slow time
  • j represents the imaginary number symbol.
  • the step of constructing a structured low-rank matrix according to the echo data includes:
  • a two-level Hankel matrix is constructed using all the columns of the echo data.
  • the two-layer Hankel matrix constructed using all the columns of the echo data is expressed as:
  • the step of decomposing the structured low-rank matrix to obtain the first initial value and the second initial value includes:
  • U and V H represent the decomposed matrix of the two-layer Hankel matrix.
  • the step of filling the structured low-rank matrix based on the first initial value and the second initial value to obtain the filled structured low-rank matrix includes:
  • the optimization problem with constraints is solved by using an alternating iterative multiplier method based on an augmented Lagrangian function to fill the structured low-rank matrix to obtain the filled structured low-rank matrix.
  • the step of obtaining an inverse SAR image based on the filled structured low-rank matrix includes:
  • Imaging processing is performed on the inverse transformation matrix by using a range-Doppler method to obtain the inverse synthetic aperture radar image.
  • U and V H represent the matrix after the decomposition of the two-layer Hankel matrix
  • S represents the matrix corresponding to the echo data under sparse sampling
  • the augmented Lagrange function is expressed as:
  • R represents the auxiliary variable
  • Ind C (X) represents the indicator function
  • the definition of Ind C (X) is:
  • P ⁇ ( ⁇ ) represents the projection operation to the region ⁇ .
  • the present application also provides an inverse synthetic aperture radar imaging device, including:
  • the echo signal preprocessing module is used to receive the echo signal corresponding to the chirp signal emitted by the radar, and preprocess the echo signal to obtain echo data;
  • a matrix generating module configured to construct a structured low-rank matrix according to the echo data
  • an initial value generating module configured to decompose the structured low-rank matrix to obtain a first initial value and a second initial value
  • a filling module configured to fill the structured low-rank matrix based on the first initial value and the second initial value, to obtain a filled structured low-rank matrix
  • An imaging module configured to obtain an inverse synthetic aperture radar image based on the filled structured low-rank matrix
  • the echo signal preprocessing module is also used for:
  • the echo data is expressed as:
  • ⁇ k represents the product of the scattering coefficient of the k-th scattering point and the rectangular window
  • x k represents the abscissa of the k-th scattering point at the initial time
  • y k represents the ordinate of the k-th scattering point at the initial time
  • Indicates the frequency modulation slope
  • c indicates the propagation speed of electromagnetic waves
  • f c indicates the center frequency of the carrier
  • indicates the target speed
  • f s indicates the sampling frequency
  • PRF indicates the pulse repetition frequency
  • j indicates the imaginary number symbol
  • m 1, 2,...
  • the present application also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and operable on the processor.
  • the processor executes the program, any of the above-mentioned The steps of the inverse synthetic aperture radar imaging method.
  • the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the inverse synthetic aperture radar imaging method described in any one of the above are realized.
  • the inverse synthetic aperture radar imaging method, device, electronic equipment, and storage medium utilize the low-rank characteristics of the inverse synthetic aperture radar echo data and the increasing effect of the Hankel matrix on the low-rank characteristics. Firstly, the sparse echoes are filled with the structured matrix to obtain a complete echo matrix, and the range Doppler method is used to obtain images.
  • an iterative solution method based on the alternating direction multiplier method is introduced.
  • This application does not rely on the singular value decomposition operation, greatly reduces the computational complexity, improves the solution efficiency, and solves the off-grid problem of the compressed sensing method, and still has good performance at low sparse sampling rates.
  • Fig. 1 is a schematic flow chart of the inverse synthetic aperture radar imaging method provided by the present application
  • FIG. 2 is a schematic flow chart of the relevant processing of echo signals provided by the present application.
  • Fig. 3 is the schematic flow chart of the construction two-layer Hankel matrix that the application provides;
  • Fig. 4 is a schematic flow chart of the decomposed Hankel matrix provided by the present application.
  • Fig. 5 is the structural representation that the application provides to generate the structured low-level matrix
  • Fig. 7 (a) is the schematic diagram of the location of sparse sampling
  • Figure 7(b) is a schematic diagram of the ISAR imaging method described in the present application.
  • Figure 7 (c) is a schematic diagram of the augmented Lagrangian matrix filling method
  • Figure 7(d) is a schematic diagram of the compressed sensing method
  • Fig. 7 (e) is the schematic diagram of traditional zero padding Fourier transform method
  • Figure 8(a) to Figure 8(b) are the results of the distance and orientation response of a scattering point under the sparse rate of 0.4 in this application;
  • Fig. 9 (a) ⁇ Fig. 9 (b) are the graphs of root mean square error and correlation coefficient under different sparse rates of the present application.
  • Fig. 10 is a schematic structural diagram of an inverse synthetic aperture radar imaging device provided by the present application.
  • FIG. 11 is a schematic structural diagram of an electronic device provided by the present application.
  • Low-rank matrix restoration is widely used in image processing for image restoration, such as denoising, deblurring, etc.
  • the data matrix of a clear natural image is often low-rank or approximately low-rank, but there are random errors with arbitrarily large amplitude but sparse distribution that destroy the low-rank of the original data.
  • Low-rank matrix restoration considers the degraded image as a set of low-dimensional data plus noise, so the data before degradation can be approximated by a low-rank matrix.
  • Low rank means that the rank of the matrix is small, and sparse means that the number of non-zero elements in the matrix is small. If a matrix is subjected to singular value decomposition and all its singular values are arranged into a vector, then the sparsity of this vector corresponds to the low rank of the matrix.
  • a grid-free sparse imaging method is matrix filling.
  • Matrix filling not only requires the echo data to have low-rank characteristics, but also needs to satisfy certain incoherence criteria.
  • the singular value vector of the matrix to be restored cannot have too high correlation with the orthonormal basis of its subspace, so that it is possible to restore the original echo data through partial observation.
  • Hankel matrix filling is further excavated to improve the performance of matrix filling.
  • This gridless method has its unique advantages in sparse imaging, so it has been widely used.
  • this application provides an inverse synthetic aperture radar imaging method.
  • a two-layer Hankel matrix is constructed to enhance the low-rank characteristics of sparse echoes, and then the low-rank first
  • the problem is transformed into an optimization problem constrained by the kernel norm by using the empirical information, and the alternate iterative multiplier method is used to solve it, which can effectively improve the sparse imaging performance.
  • FIG. 1 is a schematic flowchart of the inverse synthetic aperture radar imaging method provided by the present application, as shown in FIG. 1 .
  • An inverse synthetic aperture radar imaging method comprising:
  • Step 101 receiving an echo signal corresponding to a chirp signal transmitted by a radar, and performing preprocessing on the echo signal to obtain echo data.
  • Step 102 constructing a structured low-rank matrix according to the echo data.
  • Step 103 decompose the structured low-rank matrix to obtain a first initial value and a second initial value.
  • Step 104 Fill the structured low-rank matrix based on the first initial value and the second initial value to obtain a filled structured low-rank matrix.
  • Step 105 obtaining an inverse SAR image based on the filled structured low-rank matrix.
  • FIG. 2 is a schematic flowchart of the correlation processing of echo signals provided by the present application, as shown in FIG. 2 .
  • the step of receiving the echo signal corresponding to the chirp signal transmitted by the radar, and preprocessing the echo signal to obtain the echo data includes:
  • Step 201 performing dechirp processing on the echo signal to obtain a demodulated echo signal.
  • the echo signals of all K sampling points can be expressed as:
  • ⁇ k , x k and y k respectively represent the scattering coefficient of the kth scattering point and the horizontal and vertical coordinates at the initial moment
  • T p represents the pulse width
  • represents the frequency modulation slope
  • R 0 represents the distance between the target rotation center and the radar
  • R ref represents the reference distance for de-chirping
  • f c represents the center frequency of the carrier
  • c represents the electromagnetic wave propagation speed
  • represents the target speed
  • t r represents the fast time
  • t m represents the slow time
  • j represents the imaginary number symbol.
  • Step 202 performing translation compensation on the demodulated echo signal to obtain the echo signal after translation compensation.
  • the translation compensation includes two steps of envelope alignment and phase autofocus:
  • Envelope Alignment Range-dimensional imaging of dechirped echoes using Fourier transform. Due to the existence of the translational component, in the distance dimension image of each pulse, the same scattering point is not located in the same distance unit in different pulses, so it needs to be calibrated to the same distance unit.
  • phase self-focusing The phase error can be regarded as the model error of ISAR imaging, and the sparse representation of ISAR images can be realized by establishing a sparsely constrained optimization problem, and the estimation of the phase error can be realized during the imaging process.
  • the uneven spatial sampling of the image in the two-dimensional frequency domain can be obtained.
  • Step 203 performing sparse sampling processing on the echo signal after translation compensation to obtain the echo data.
  • Sparse sampling For example, 256 points in the pulse are sampled, and a total of 256 pulses are used to form a 256*256 two-dimensional echo matrix, and some elements in the matrix are randomly selected.
  • Low rank The number of eigenvalues/singular values of the matrix is much smaller than the dimension of the matrix. Singular value decomposition can be performed on the matrix, and whether it has low-rank characteristics can be judged by the number and distribution of its large singular values.
  • the echo data is expressed as:
  • ⁇ k represents the product of the scattering coefficient of the kth scattering point and the rectangular window
  • f s represents the sampling frequency
  • PRF represents the pulse repetition frequency
  • m 1, 2,...
  • M represents the fast time series number in discrete form
  • S(m,n) represents the two-dimensional echo data that has been sparsely sampled, that is, some elements in S are 0.
  • the formula (3) is the continuous form of the echo signal
  • the formula (4) is the discrete form of the echo signal (called echo data).
  • FIG. 3 is a schematic flow chart of constructing a two-layer Hankel matrix provided by the present application, as shown in FIG. 3 .
  • the step of constructing a structured low-rank matrix according to the echo data includes:
  • Step 301 constructing a Hankel matrix according to the nth column S(:,n) of the echo data.
  • the Hankel matrix constructed according to the nth column S(:,n) is expressed as:
  • Step 302 using all columns of the echo data to construct a two-layer Hankel matrix.
  • the two-layer Hankel matrix constructed using all columns of the echo data is expressed as:
  • the low-rank property of the Hankel matrix can be used for sparse ISAR imaging of targets composed of finite scattering points.
  • the Hankel matrix is constructed to make full use of the translation invariance of the signal for high-resolution spectrum estimation while reducing the influence of noise. Therefore, the two-layer Hankel structured method constructed in this application can enhance low-rank priors, and the derived results are more conducive to sparse ISAR imaging.
  • FIG. 4 is a schematic flow diagram of the Hankel matrix decomposition provided by the present application, as shown in FIG. 4 .
  • the step of decomposing the structured low-rank matrix to obtain the first initial value and the second initial value includes:
  • Step 401 decompose the two-layer Hankel matrix into the product of two parts, namely
  • step 402 rank estimation is performed using a low-rank matrix fitting method to obtain a first initial value U (0) and a second initial value V (0) .
  • U and V H represent the decomposed matrix of the two-layer Hankel matrix.
  • the two-layer Hankel matrix is a structured matrix.
  • FIG. 5 is a schematic structural diagram of generating a structured low-level matrix provided by the present application, as shown in FIG. 5 .
  • the step of filling the structured low-rank matrix based on the first initial value and the second initial value to obtain the filled structured low-rank matrix includes:
  • Step 501 construct an optimization problem with constraints based on the first initial value and the second initial value.
  • optimization problem with constraints is expressed as:
  • U and V H represent the matrix after decomposing the two-layer Hankel matrix
  • S represents the matrix corresponding to the echo data under sparse sampling, that is, the element values in the ⁇ region are consistent with the echo matrix X, while other regions All elements are 0.
  • 256*256 are all echoes, and some elements are known, and the area corresponding to these elements is ⁇ .
  • Step 502 Solve the optimization problem with constraints by using the alternate iterative multiplier method based on the augmented Lagrangian function, so as to fill the structured low-rank matrix, and obtain the filled structured low-rank matrix rank matrix.
  • the augmented Lagrangian function is expressed as:
  • R represents the auxiliary variable
  • Ind C (X) represents the indicator function
  • the definition of Ind C (X) is:
  • P ⁇ ( ⁇ ) represents the projection operation to the region ⁇ .
  • ADMM Alternate Iterative Multiplier Method
  • each iteration of ADMM is to find the point where the conjugate gradient is 0 for X, U, and V in turn. Because this function is convex with respect to the three variables X, U, and V, the extreme point is the most value point.
  • the iterative steps of the above ADMM are as follows:
  • FIG. 6 is a schematic flow chart of imaging using the range-Doppler method provided by the present application, as shown in FIG. 6 .
  • the inverse synthetic aperture radar image is obtained based on the filled structured low-rank matrix, including:
  • Step 601 performing inverse transformation on the filled structured low-rank matrix to obtain an inverse transformation matrix.
  • the structured matrix obtained by solving Perform inverse transformation processing to obtain the inverse transformation matrix X is the structured Hankel matrix, and then transforms back to the original two-dimensional echo matrix, and its elements are in one-to-one correspondence.
  • Step 602 Perform imaging processing on the inverse transformation matrix by using a range-Doppler method to obtain the inverse SAR image.
  • the range-Doppler method performs ISAR imaging processing on the inverse transformation matrix X in the dechirp mode, which actually performs two-dimensional Fourier transform, that is, 2D-FFT.
  • this application uses structured operations to enhance the low-rank characteristics of echoes, which can be applied to sparse inverse synthetic aperture radar imaging at low sampling rates.
  • the measured data is based on the Jacques 42 aircraft model.
  • the center frequency of the system is 5.52GHz
  • the bandwidth of the transmitted chirp signal is 500MHz
  • the pulse repetition frequency is 100Hz.
  • RMSE root mean square error
  • CORR correlation coefficient
  • S is the reference image corresponding to the full echo data, is the reconstructed image obtained by the sparse imaging method.
  • Figure 7(a) to Figure 7(e) are the sparse imaging results of different methods, Figure 7(a) is the position of sparse sampling; Figure 7(b) is the ISAR imaging method described in this application; Figure 7(c) is Augmented Lagrangian matrix filling method; Figure 7(d) is the compressed sensing method; Figure 7(e) is the traditional zero-filling Fourier transform method.
  • Figure 7(a) to Figure 7(e) are the sparse imaging results of different methods, and the imaging effects when the sparse sampling rate is 0.2, 0.4, and 0.6 from top to bottom.
  • Figure 8(a) and Figure 8(b) respectively show the range and azimuth impulse responses of a scattering point when the sparse rate is 0.4
  • Figure 9(a) and Figure 9(b) show the different methods Root mean square error and correlation coefficient at different sparsification rates.
  • the solid line (—) represents total echo data.
  • Figures 7(a) to 7(e) show the imaging results of different methods at different thinning rates.
  • the thinning rate ranges from 0.2 to 0.6 with an interval of 0.2. It still has excellent imaging performance at a fairly low sparsity rate, partly due to the enhancement of the low-rank property of the matrix by the structured operation.
  • the following describes the inverse synthetic aperture radar imaging device provided by the present application.
  • the inverse synthetic aperture radar imaging device described below and the above described inverse synthetic aperture radar imaging method can be referred to in correspondence.
  • FIG. 10 is a schematic structural diagram of the inverse synthetic aperture radar imaging device provided by the present application, as shown in FIG. 10 .
  • An inverse synthetic aperture radar imaging device 1000 includes an echo signal preprocessing module 1010 , a matrix generation module 1020 , an initial value generation module 1030 , a filling module 1040 and an imaging module 1050 . in,
  • the echo data module 1010 is configured to receive an echo signal corresponding to the chirp signal emitted by the radar, and preprocess the echo signal to obtain echo data.
  • the matrix generation module 1020 is configured to construct a structured low-rank matrix according to the echo data.
  • the initial value generating module 1030 is configured to decompose the structured low-rank matrix to obtain a first initial value and a second initial value.
  • a filling module 1040 configured to fill the structured low-rank matrix based on the first initial value and the second initial value, so as to obtain a filled structured low-rank matrix.
  • the imaging module 1050 is configured to obtain an inverse SAR image based on the filled structured low-rank matrix.
  • the echo signal preprocessing module 1010 is also configured to perform the following steps:
  • demodulated echo signal is expressed as:
  • ⁇ k , x k and y k respectively represent the scattering coefficient of the kth scattering point and the horizontal and vertical coordinates at the initial moment
  • T p represents the pulse width
  • represents the frequency modulation slope
  • R 0 represents the distance between the target rotation center and the radar
  • R ref represents the reference distance for de-chirping
  • f c represents the center frequency of the carrier
  • c represents the propagation speed of electromagnetic waves
  • represents the target rotational speed
  • t r represents the fast time
  • t m represents the slow time
  • j represents the imaginary number symbol.
  • the echo data is expressed as:
  • ⁇ k represents the product of the scattering coefficient of the kth scattering point and the rectangular window
  • f s represents the sampling frequency
  • PRF is the pulse repetition frequency
  • m 1, 2,...
  • M represents the fast time series number in discrete form
  • n 1,2,...
  • N represents the slow time series number in discrete form
  • S(m,n) is the two-dimensional echo data that has been sparsely sampled, that is, some elements in S are 0.
  • the matrix generating module 1020 is also configured to perform the following steps:
  • a two-level Hankel matrix is constructed using all the columns of the echo data.
  • the Hankel matrix constructed according to the nth column S(:,n) is expressed as:
  • the two-layer Hankel matrix constructed using all the columns of the echo data is expressed as:
  • the initial value generation module 1030 is also used to perform the following steps:
  • U and V H represent the decomposed matrix of the two-layer Hankel matrix.
  • the step of filling the structured low-rank matrix based on the first initial value and the second initial value to obtain the filled structured low-rank matrix includes:
  • the optimization problem with constraints is solved by using an alternate iterative multiplier method based on an augmented Lagrangian function to fill the structured low-rank matrix to obtain the filled structured low-rank matrix.
  • the filling module 1040 is also used to perform the following steps:
  • the optimization problem with constraints is solved by using an alternate iterative multiplier method based on an augmented Lagrangian function to fill the structured low-rank matrix to obtain the filled structured low-rank matrix.
  • the imaging module 1050 is also configured to perform the following steps:
  • Imaging processing is performed on the inverse transformation matrix by using a range-Doppler method to obtain the inverse synthetic aperture radar image.
  • optimization problem with constraints is expressed as:
  • U and V H represent the matrix after the decomposition of the two-layer Hankel matrix
  • S represents the matrix corresponding to the echo data under sparse sampling
  • the augmented Lagrange function is expressed as:
  • R represents the auxiliary variable
  • Ind C (X) represents the indicator function
  • the definition of Ind C (X) is:
  • P ⁇ ( ⁇ ) represents the projection operation to the region ⁇ .
  • the iterative steps of the alternate iterative multiplier method are expressed as:
  • the present application uses structured operations to enhance the low-rank characteristics of echoes, and is suitable for sparse inverse synthetic aperture radar imaging at low sampling rates.
  • Figure 11 illustrates a schematic diagram of the physical structure of an electronic device, as shown in Figure 11, the electronic device may include: a processor (processor) 1110, a communication interface (Communications Interface) 1120, a memory (memory) 1130 and a communication bus 1140, Wherein, the processor 1110 , the communication interface 1120 , and the memory 1130 communicate with each other through the communication bus 1140 .
  • the processor 1110 may call logic instructions in the memory 1130 to execute the above-mentioned inverse synthetic aperture radar imaging method, the method including:
  • An inverse synthetic aperture radar image is obtained based on the filled structured low-rank matrix.
  • the above-mentioned logic instructions in the memory 1130 may be implemented in the form of software function units and may be stored in a computer-readable storage medium when sold or used as an independent product.
  • the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), magnetic disk or wide disk and other media that can store program codes. .
  • the present application also provides a computer program product
  • the computer program product includes a computer program stored on a non-transitory computer-readable storage medium
  • the computer program includes program instructions, and when the program instructions are executed by a computer During execution, the computer can execute the inverse synthetic aperture radar imaging method provided by the above methods.
  • the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it is implemented to perform the inverse synthetic aperture radar imaging method provided above.
  • the device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. It can be understood and implemented by those skilled in the art without any creative efforts.
  • each implementation can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware.
  • the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic Disks, disks, etc., include several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments.

Landscapes

  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

一种逆合成孔径雷达成像方法、装置、电子设备及存储介质,方法包括:接收雷达发射的回波信号并对其进行预处理以得到回波数据;构建结构化低秩矩阵并将其进行分解以得到第一初值与第二初值,并对结构化低秩矩阵进行填充以最后得到逆合成孔径雷达图像。适用于低采样率下的稀疏逆合成孔径雷达成像。装置包括回波信号预处理模块,矩阵生成模块,初值生成模块,填充模块,成像模块。电子设备包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序。存储介质上存储有计算机程序。

Description

逆合成孔径雷达成像方法、装置、电子设备及存储介质
相关申请
本申请要求2021年09月26日申请的,申请号为202111125293.2,名称为“逆合成孔径雷达成像方法、装置、电子设备及存储介质”的中国专利申请的优先权,在此将其全文引入作为参考。
技术领域
本申请涉及雷达信号处理技术领域,尤其涉及一种逆合成孔径雷达成像方法、装置、电子设备及存储介质。
背景技术
逆合成孔径雷达(Inverse Synthetic Aperture Radar,简称ISAR)作为微波遥感的一种主要工具,能够提供天空、太空、海洋等场景目标的高分辨率图像,并且具备全天时、全天候工作的能力。
传统的ISAR成像技术时通过对回波数据进行相干积累,以获得较高的分辨率和输出信噪比(SNR),从而得到距离-多普勒图像。而较高分辨率ISAR成像则要求发射端信号具有宽频带,需要更多的合成孔径时间。但实际应用中会存在系统复杂度高、数据量大以及目标运动复杂等问题,并且多功能雷达的工作状态交换或被观测目标的机动运动等非理想情况下,只能获得部分回波数据(不完整的回波数据可称为稀疏孔径回波数据),所以此种情况下仅使用传统的方法很难实现不模糊的成像。
发明内容
本申请实施例提供一种逆合成孔径雷达成像方法、装置、电子设备及存储介质,用以解决现有技术中低采样率下ISAR成像分辨率较低的问题,实现低采样率下提高ISAR成像的分辨率。
本申请实施例采用下述技术方案:
本申请提供一种逆合成孔径雷达成像方法,包括:
接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据;
根据所述回波数据,构建结构化低秩矩阵;
将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值;
基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵;
基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像;
其中,所述接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据的步骤包括:
对所述回波信号进行解线性调频处理,得到解调后的回波信号;
对所述解调后的回波信号进行平动补偿,得到平动补偿后的回波信号;
对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据;
所述回波数据表示为:
Figure PCTCN2022099364-appb-000001
其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,x k表示第k个散射点在初始时刻的横坐标,y k表示第k个散射点在初始时刻的纵坐标,γ表示调频斜率,c表示电磁波传播速度,f c表示载波中心频率,ω表示目标转速,f s表示采样频率,PRF表示脉冲重复频率,j表示虚数符号,m=1,2,...,M表示离散形式的快时间序数,m=1,2,...,N表示离散形式的慢时间序数,S(m,n)表示经过稀疏采样的两维回波数据。
进一步地,所述结构化低秩矩阵是二层汉克尔矩阵。
进一步地,所述解调后的回波信号表示为:
Figure PCTCN2022099364-appb-000002
其中,σ k、x k和y k分别表示第k个散射点的散射系数以及在初始时刻的横纵坐标,T p表示脉冲宽度,γ表示调频斜率,R 0表示目标转动中心与雷达的距离,R ref表示解线性调频的参考距离,f c表示载波中心频率,c表示电磁波传播速度,ω表示目标转速,t r表示快时间,t m表示慢时间,j表示虚数符号。
进一步地,所述根据所述回波数据,构建结构化低秩矩阵的步骤包括:
根据所述回波数据的第n列S(:,n)构造汉克尔矩阵;
利用所述回波数据的所有列构造二层汉克尔矩阵。
进一步地,根据所述第n列S(:,n)构造出的汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000003
利用所述回波数据的所有列构造出的所述二层汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000004
其中,P和Q是束参数。
进一步地,所述将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值的步骤包括:
将所述二层汉克尔矩阵分解为两部分的乘积,即
Figure PCTCN2022099364-appb-000005
利用低秩矩阵拟合方法进行秩估计,得到第一初值U (0)和第二初值V (0)
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵。
进一步地,所述基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵的步骤包括:
基于所述第一初值与所述第二初值构建带有约束的优化问题;
利用基于增广拉格朗日函数的交替迭代乘子法对所述带有约束的优化问 题求解,以对所述结构化低秩矩阵进行填充,得到所述填充后的结构化低秩矩阵。
进一步地,所述基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像的步骤包括:
对所述填充后的结构化低秩矩阵进行逆变换,得到逆变换矩阵;
利用距离-多普勒方法对所述逆变换矩阵进行成像处理,得到所述逆合成孔径雷达图像。
进一步地,所述带有约束的优化问题表示为:
Figure PCTCN2022099364-appb-000006
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵,S表示稀疏采样下的回波数据对应的矩阵;
所述增广拉格朗日函数表示为:
Figure PCTCN2022099364-appb-000007
其中,R表示辅助变量,Ind C(X)表示指示函数,对Ind C(X)定义为:
Figure PCTCN2022099364-appb-000008
其中,P Ω(·)表示向区域Ω的投影操作。
进一步地,所述交替迭代乘子法的迭代步骤表示为:
Figure PCTCN2022099364-appb-000009
Figure PCTCN2022099364-appb-000010
Figure PCTCN2022099364-appb-000011
Figure PCTCN2022099364-appb-000012
i=i+1;
其中,
Figure PCTCN2022099364-appb-000013
对Ω的补集进行映射,
Figure PCTCN2022099364-appb-000014
表示逆结构化操作,I表示单位矩阵,表示所述填充后的结构化低秩矩阵。
本申请还提供一种逆合成孔径雷达成像装置,包括:
回波信号预处理模块,用于接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据;
矩阵生成模块,用于根据所述回波数据,构建结构化低秩矩阵;
初值生成模块,用于将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值;
填充模块,用于基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵;
成像模块,用于基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像;
所述回波信号预处理模块,还用于:
对所述回波信号进行解线性调频处理,得到解调后的回波信号;
对所述解调后的回波信号进行平动补偿,得到平动补偿后的回波信号;
对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据;
所述回波数据表示为:
Figure PCTCN2022099364-appb-000015
其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,x k表示第k个散射点在初始时刻的横坐标,y k表示第k个散射点在初始时刻的纵坐标,γ表示调频斜率,c表示电磁波传播速度,f c表示载波中心频率,ω表示目标转速,f s表示采样频率,PRF表示脉冲重复频率,j表示虚数符号,m=1,2,...,M表示离散形式的快时间序数,n=1,2,...,N表示离散形式的慢时间序数,S(m,n)表示经过稀疏采样的两维回波数据。
本申请还提供一种电子设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,所述处理器执行所述程序时实现如上任一项所述的逆合成孔径雷达成像方法的步骤。
本申请还提供一种非暂态计算机可读存储介质,其上存储有计算机程序,所述计算机程序被处理器执行时实现如上任一项所述的逆合成孔径雷达成像方法的步骤
本申请实施例采用的上述至少一个技术方案能够达到以下有益效果:
本申请实施例所提供的逆合成孔径雷达成像方法、装置、电子设备及存储介质,利用逆合成孔径雷达回波数据的低秩特性,以及汉克尔矩阵对低秩特性的增加作用,采用结构化操作以及矩阵填充方法,首先由稀疏回波进行结构化矩阵填充得到完整的回波矩阵,并利用距离多普勒方法得到图像。
具体的,在利用距离多普勒方法得到图像的求解中,为了较少由矩阵维度增加带来的高计算量,引入了基于交替方向乘子法的迭代求解方法。本申请不依赖于奇异值分解操作,极大减少了计算复杂度,提高了求解效率,并且解决了压缩感知方法的离网格问题,且在低稀疏采样率下仍有良好的性能表现。
附图说明
此处所说明的附图用来提供对本申请的进一步理解,构成本申请的一部分,本申请的示意性实施例及其说明用于解释本申请,并不构成对本申请的不当限定。在附图中:
图1是本申请提供的逆合成孔径雷达成像方法的流程示意图;
图2是本申请提供的对回波信号进行相关处理的流程示意图;
图3是本申请提供的构造二层汉克尔矩阵的流程示意图;
图4是本申请提供的分解汉克尔矩阵的流程示意图;
图5是本申请提供的生成结构化低轶矩阵的结构示意图;
图6是本申请提供的利用距离-多普勒方法成像的流程示意图;
图7(a)为稀疏采样的位置的示意图;
图7(b)为本申请所述ISAR成像方法的示意图;
图7(c)为增广拉格朗日矩阵填充方法的示意图;
图7(d)为压缩感知方法的示意图;
图7(e)为传统的补零傅里叶变换方法的示意图;
图8(a)~图8(b)是本申请在0.4稀疏率下某散射点距离和方位响应的结果图;
图9(a)~图9(b)是本申请在不同稀疏率下均方根误差和相关性系数的曲线图;
图10是本申请提供的逆合成孔径雷达成像装置的结构示意图;
图11是本申请提供的电子设备的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚,下面将结合本申请具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然,所描述的实施例仅是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
以下对本申请涉及的技术术语进行描述:
低秩矩阵恢复(LRMR)广泛用于图像处理用于图像恢复,比如去噪、去模糊等。一幅清晰的自然图像其数据矩阵往往是低秩或者近似低秩的,但存在随机幅值任意大但是分布稀疏的误差破坏了原有数据的低秩性。低秩矩阵恢复是将退化图像看做一组低维数据加上噪声形成的,因此退化前的数据就可以通过低秩矩阵来逼近。
低秩是指矩阵的秩较小,稀疏是指矩阵中非零元素的个数少。如果对矩阵进行奇异值分解,并把其所有奇异值排列为一个向量,那么这个向量的稀疏性便对应于该矩阵的低秩性。
秩,设在矩阵A中有一个不等于0的r阶子式D,且所有r+1阶子式(如果存在的话)全等于0,那么D称为矩阵A的最高阶非零子式,数r称为矩阵A的秩,记作R(A),并规定零矩阵的秩等于0。
为了解决现有技术关于稀疏成像的问题,压缩感知方法得到了广泛的应用。通常目标被认为是由有限个散射中心构成的,因为具有稀疏性。在稀疏性这一 先验信息的帮助下,压缩感知类方法在稀疏成像应用中有着良好的性能表现。但是,在这类方法中,不可避免地需要构造离散字典,通常为傅里叶变换字典,由网格划分带来的模型失配是压缩感知类方法的一大固有缺点。
示例性地,一种无网格的稀疏成像方法是矩阵填充,利用低秩特性,矩阵即使缺失部分元素,也可在某些条件下被精准地恢复。矩阵填充不仅要求回波数据具有低秩特性,而且需要满足一定的非相干性准则。具体来说,待恢复矩阵的奇异值矢量与其所在子空间的标准正交基不能有过高的相关性,才有可能通过部分观测恢复出原始回波数据。在此基础上,汉克尔矩阵填充被进一步挖掘,以提高矩阵填充的性能,这种无网格的方法在稀疏成像上有其独特优势,因此得到了广泛的应用。
因此,针对上述问题,本申请提供一种逆合成孔径雷达成像方法,通过充分利用结构化的矩阵填充技术,构造二层汉克尔矩阵增强稀疏回波的低秩特性,然后再利用低秩先验信息将问题转化为核范数约束的优化问题,并利用交替迭代乘子法进行求解,可有效地提高稀疏成像性能。
以下结合附图1-图11,详细说明本申请各实施例提供的技术方案。
图1是本申请提供的逆合成孔径雷达成像方法的流程示意图,如图1所示。一种逆合成孔径雷达成像方法,包括:
步骤101,接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据。
步骤102,根据所述回波数据,构建结构化低秩矩阵。
步骤103,将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值。
步骤104,基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵。
步骤105,基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像。
以下对上述步骤101~105进行具体描述。
图2是本申请提供的对回波信号进行相关处理的流程示意图,如图2所示。上述步骤101中,所述接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据的步骤包括:
步骤201,对所述回波信号进行解线性调频处理,得到解调后的回波信号。
假设雷达发射线性调频(LFM)信号,采样的所有K个散射点的回波信号可表示为:
Figure PCTCN2022099364-appb-000016
将上式(1)经过解线性调频处理,得到参考信号:
Figure PCTCN2022099364-appb-000017
对(1)中的s(t r,t m),和(2)中的s ref(t r,t m)作差频处理,得到所述解调后的回波信号表示为:
Figure PCTCN2022099364-appb-000018
其中,上述式(1)~(3)中,σ k、x k和y k分别表示第k个散射点的散射系数以及在初始时刻的横纵坐标,T p表示脉冲宽度,γ表示调频斜率,R 0表示目标转动中心与雷达的距离,R ref表示解线性调频的参考距离,f c表示载波中心频率,c表示电磁波传播速度,ω表示目标转速,t r表示快时间,t m表示慢时间,j表示虚数符号。
以下式子中同样的参数含义表示同上。
需要说明的是,上述对式(1)和式(2)的处理得到式(3),可以直接在模拟电路的混频器中得到,即从混频器中即可输出式(3)的回波信号。
步骤202,对所述解调后的回波信号进行平动补偿,得到平动补偿后的回 波信号。
由于雷达和目标之间有相对平动的情况,所以平动的影响需要从信号中移除,这个步骤称为平动补偿。所述平动补偿包括包络对齐和相位自聚焦两个步骤:
包络对齐:对解线性调频后的回波利用傅里叶变换进行距离维成像。由于平动分量的存在,每个脉冲的距离维图像中,同一个散射点在不同脉冲中并不位于同一个距离单元,需要将其校准到同一距离单元。
相位自聚焦:可以将相位误差看作ISAR成像的模型误差,通过建立稀疏约束的最优化问题实现对ISAR图像的稀疏表征,在成像过程中实现相位误差的估计。
平动补偿之后,就能够获得图像在二维频域内不均匀的空间采样。
步骤203,对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据。
稀疏采样:比如脉冲内256个点采样,共使用256个脉冲,构成一个256*256的两维回波矩阵,随机取矩阵中的部分元素。
低秩:矩阵的特征值/奇异值个数远小于矩阵的维度。可以对矩阵进行奇异值分解,通过其较大奇异值的个数及分布来判断其是否具有低秩特性。
可选的,所述回波数据表示为:
Figure PCTCN2022099364-appb-000019
其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,f s表示采样频率,PRF表示脉冲重复频率,m=1,2,...,M表示离散形式的快时间序数,n=1,2,...,N表示离散形式的慢时间序数,S(m,n)表示经过稀疏采样的两维回波数据,即 S中部分元素为0。
需要说明的是,式(3)是回波信号的连续形式,式(4)是回波信号的离散形式(称为回波数据),比如式(3)t r是连续的时间,但实际上本申请要对回波信 号进行采样,每隔1/f s的时间采一次,那么第m次对应的就是(m-1)/f s,同理,t m=(n-1)/PRF。因此经过解线性处理的回波信号,即式(3),舍去某些常数项后,可得到式(4)。
图3是本申请提供的构造二层汉克尔矩阵的流程示意图,如图3所示。上述步骤102中,所述根据所述回波数据,构建结构化低秩矩阵的步骤包括:
步骤301,根据所述回波数据的第n列S(:,n)构造汉克尔(Hankel)矩阵。
可选的,根据所述第n列S(:,n)构造出的汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000020
步骤302,利用所述回波数据的所有列构造二层汉克尔矩阵。
可选的,利用所述回波数据的所有列构造出的所述二层汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000021
其中,P和Q是束参数(pencil parameter)。
可以理解的是,利用汉克尔矩阵的低秩性质,可用于有限散射点组成的目标的稀疏ISAR成像。在矩阵束法中,构造汉克尔矩阵,充分利用信号的平移不变性进行高分辨率谱估计,同时减小了噪声的影响。因此,本申请构造所述二层汉克尔结构化方法可以增强低秩先验,而导出的结果更有利于稀疏ISAR成像。
图4是本申请提供的分解汉克尔矩阵的流程示意图,如图4所示。上述步骤103中,所述将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值的步骤包括:
步骤401,将所述二层汉克尔矩阵分解为两部分的乘积,即
Figure PCTCN2022099364-appb-000022
步骤402,利用低秩矩阵拟合方法进行秩估计,得到第一初值U (0)和第二初值V (0)
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵。所述二层汉克尔 矩阵
Figure PCTCN2022099364-appb-000023
为结构化后的矩阵。
图5是本申请提供的生成结构化低轶矩阵的结构示意图,如图5所示。上述步骤104中,所述基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵的步骤包括:
步骤501,基于所述第一初值与所述第二初值构建带有约束的优化问题。
可选的,所述带有约束的优化问题表示为:
Figure PCTCN2022099364-appb-000024
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵,S表示稀疏采样下的回波数据对应的矩阵,即在Ω区域内元素值与回波矩阵X一致,而其他区域元素均为0。
比如,上述步骤203中的示例,256*256是全部回波,现在知道其中某些元素,这些元素对应的区域为Ω。
步骤502,利用基于增广拉格朗日函数的交替迭代乘子法对所述带有约束的优化问题求解,以对所述结构化低秩矩阵进行填充,得到所述填充后的结构化低秩矩阵。
可选的,所述增广拉格朗日函数表示为:
Figure PCTCN2022099364-appb-000025
其中,R表示辅助变量,Ind C(X)表示指示函数,对Ind C(X)定义为:
Figure PCTCN2022099364-appb-000026
其中,P Ω(·)表示向区域Ω的投影操作。
可选的,利用所述交替迭代乘子法(简称ADMM)对上述优化问题进行求解的迭代步骤表示为:
Figure PCTCN2022099364-appb-000027
Figure PCTCN2022099364-appb-000028
Figure PCTCN2022099364-appb-000029
Figure PCTCN2022099364-appb-000030
i=i+1;  (14)
其中,
Figure PCTCN2022099364-appb-000031
表示对Ω的补集进行映射,
Figure PCTCN2022099364-appb-000032
表示逆结构化操作,即由所述二层汉克尔矩阵变换为原始矩阵,I表示单位矩阵,
Figure PCTCN2022099364-appb-000033
表示所述填充后的结构化低轶矩阵。
具体的,ADMM的每步迭代都是依次对X,U,V求共轭梯度为0的点。因为这个函数关于X,U,V三个变量都是凸的,所以极值点就是最值点。上述ADMM的迭代步骤如下:
针对上述式(10),结合式(9),得到式(15):
Figure PCTCN2022099364-appb-000034
对式(15)进行关于X共轭求梯度,令其为0,则得到式(16):
Figure PCTCN2022099364-appb-000035
所以,
Figure PCTCN2022099364-appb-000036
考虑到前一项Ind C(X),所以调整为:
Figure PCTCN2022099364-appb-000037
针对上述式(11),结合式(7),得到式(18):
Figure PCTCN2022099364-appb-000038
对式(18)进行关于U共轭求梯度,即:
Figure PCTCN2022099364-appb-000039
得到式(20):
Figure PCTCN2022099364-appb-000040
基于上述同样的方法,可实现对上述式(12)的求解。
最后,辅助变量R的更新为ADMM框架下的固定方法,即式(13):
Figure PCTCN2022099364-appb-000041
上述中,由于矩阵求共轭梯度是具有固定公式,故本申请对其具体的计算步骤不再赘述。
图6是是本申请提供的利用距离-多普勒方法成像的流程示意图,如图6所示。上述步骤105中,所述基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像,包括:
步骤601,对所述填充后的结构化低秩矩阵进行逆变换,得到逆变换矩阵。
具体的,对求解得到的结构化矩阵
Figure PCTCN2022099364-appb-000042
做逆变换处理,得到逆变换矩阵X。逆变换是由结构化后的Hankel矩阵,再变换回原来的两维回波矩阵,其元素是一一对应的。
步骤602,利用距离-多普勒方法对所述逆变换矩阵进行成像处理,得到所述逆合成孔径雷达图像。
可选的,距离-多普勒方法在解线性调频模式下,对所述逆变换矩阵X进行ISAR成像处理,实际是作两维傅里叶变换,即2D-FFT。
利用距离-多普勒方法对所述逆变换矩阵进行成像处理为本领域技术人员所知晓,故本申请对其变换不再赘述。
综上所述,本申请利用结构化操作增强了回波的低秩特性,可适用于低采 样率下的稀疏逆合成孔径雷达成像。
为了说明本申请对稀疏逆合成孔径雷达成像的有效性,通过基于实测数据的实验进行进一步的论证:
(1)实验设置
实测数据基于雅克42飞机模型,系统的中心频率为5.52GHz,发射线性调频信号的带宽为500MHz,脉冲重复频率为100Hz,系统工作方式为解线性调频,使用256个脉冲进行逆合成孔径成像。
为验证本申请的有效性,分别取不同稀疏率,比较其成像质量。并使用均方根误差(RMSE)和相关性系数(CORR)两个指标作为评价参考,其定义分别为:
Figure PCTCN2022099364-appb-000043
Figure PCTCN2022099364-appb-000044
其中,S是全回波数据对应的参考图像,
Figure PCTCN2022099364-appb-000045
是利用稀疏成像方法得到的重构图像。s 1=vec(S 1)和s 2=vec(S 2)分别代表向量化后的原始和重构图像。
(2)实验内容
基于MATLAB软件平台,分别取不同稀疏率,将本申请的均方根误差和相关性系数与稀疏回波直接成像、压缩感知方法、传统矩阵填充方法进行对比。
图7(a)~图7(e)为不同方法的稀疏成像结果,图7(a)为稀疏采样的位置;图7(b)为本申请所述ISAR成像方法;图7(c)为增广拉格朗日矩阵填充方法;图7(d)为压缩感知方法;图7(e)为传统的补零傅里叶变换方法。
图7(a)~图7(e)为不同方法的稀疏成像结果,从上到下按稀疏采样率为0.2、0.4、0.6时的成像效果。
基于L1范数的压缩感知方法,以及直接利用稀疏回波进行成像的结果。图8(a)和图8(b)分别给出了稀疏率为0.4时,其中某个散射点的距离向和方位 向脉冲响应,图9(a)和图9(b)分别为不同方法在不同稀疏率下的均方根误差和相关性系数。
如图8(a)、图8(b)所示,实线(—)表示全回波数据。虚线部分(---):
Figure PCTCN2022099364-appb-000046
表示本申请所述逆合成孔径雷达成像方法;
Figure PCTCN2022099364-appb-000047
表示传统的补零傅里叶变换方法;
Figure PCTCN2022099364-appb-000048
表示压缩感知方法;
Figure PCTCN2022099364-appb-000049
表示增广拉格朗日矩阵填充方法下距离和方位两个维度的响应曲线,距离维响应曲线(对应图8(a)),方位维响应曲线(对应图8(b))。
如图9(a)、图9(b)所示,
Figure PCTCN2022099364-appb-000050
表示本申请所述ISAR成像方法;
Figure PCTCN2022099364-appb-000051
表示传统的补零傅里叶变换方法;
Figure PCTCN2022099364-appb-000052
表示压缩感知方法;
Figure PCTCN2022099364-appb-000053
表示增广拉格朗日矩阵填充方法下的两个指标:均方根误差RMSE(对应9(a)),相关性系数CORR(对应9(b))。
(3)实验结果分析
图7(a)~图7(e)给出了不同方法在不同稀疏率下的成像结果,稀疏率从0.2到0.6,间隔为0.2,从图中可以看出,本申请所提出的方法即使在相当低的稀疏率下仍然有优秀的成像性能,一部分原因是结构化操作对矩阵低秩特性的增强。
图8(a)和图8(b)中的距离和方位向脉冲响应曲线表明,本申请的响应与全回波数据得到的图像最为接近。
图9(a)和图9(b)中的均方根误差以及相关性系数曲线可以看出,对于所有方法来说,均方根误差随着稀疏采样率的增加而减小,而相关性系数随稀疏采样率的增加而增加。其中,本申请所提出的结构化矩阵填充方法与参考图像相比,有着最小的误差和最高的一致性,该结论与图7(a)~图7(e)中直观的图像质量相符。
下面对本申请提供的逆合成孔径雷达成像装置进行描述,下文描述的逆合成孔径雷达成像装置与上文描述的逆合成孔径雷达成像方法可相互对应参照。
图10是本申请提供的逆合成孔径雷达成像装置的结构示意图,如图10所示。一种逆合成孔径雷达成像装置1000,包括回波信号预处理模块1010、矩阵生成模块1020、初值生成模块1030、填充模块1040以及成像模块1050。其中,
回波数据模块1010,用于接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据。
矩阵生成模块1020,用于根据所述回波数据,构建结构化低秩矩阵。
初值生成模块1030,用于将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值。
填充模块1040,用于基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵。
成像模块1050,用于基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像。
可选的,所述回波信号预处理模块1010,还用于执行如下步骤:
对所述回波信号进行解线性调频处理,得到解调后的回波信号;
对所述解调后的回波信号进行平动补偿,得到平动补偿后的回波信号;
对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据。
进一步地,所述解调后的回波信号表示为:
Figure PCTCN2022099364-appb-000054
其中,σ k、x k和y k分别表示第k个散射点的散射系数以及在初始时刻的横纵坐标,T p表示脉冲宽度,γ表示调频斜率,R 0表示目标转动中心与雷达的距离,R ref表示解线性调频的参考距离,f c表示载波中心频率,c表示电磁波传播速度,ω表示目标转速,t r表示快时间,t m表示慢时间,j表示虚数符号。
可选的,所述回波数据表示为:
Figure PCTCN2022099364-appb-000055
其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,f s表示采样频 率,PRF是脉冲重复频率,m=1,2,...,M表示离散形式的快时间序数,n=1,2,...,N表示离散形式的慢时间序数,S(m,n)为经过稀疏采样的两维回波数据,即S中部分元素为0。
可选的,矩阵生成模块1020,还用于执行如下步骤:
根据所述回波数据的第n列S(:,n)构造汉克尔矩阵;
利用所述回波数据的所有列构造二层汉克尔矩阵。
可选的,根据所述第n列S(:,n)构造出的汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000056
利用所述回波数据的所有列构造出的所述二层汉克尔矩阵表示为:
Figure PCTCN2022099364-appb-000057
其中,P和Q是束参数。
可选的,初值生成模块1030,还用于执行如下步骤:
将所述二层汉克尔矩阵分解为两部分的乘积,即
Figure PCTCN2022099364-appb-000058
利用低秩矩阵拟合方法进行秩估计,得到第一初值U (0)和第二初值V (0)
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵。
进一步地,所述基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵的步骤包括:
基于所述第一初值与所述第二初值构建带有约束的优化问题;
利用基于增广拉格朗日函数的交替迭代乘子法对所述带有约束的优化问题求解,以对所述结构化低秩矩阵进行填充,得到所述填充后的结构化低秩矩阵。
可选的,填充模块1040,还用于执行如下步骤:
基于所述第一初值与所述第二初值构建带有约束的优化问题;
利用基于增广拉格朗日函数的交替迭代乘子法对所述带有约束的优化问题求解,以对所述结构化低秩矩阵进行填充,得到所述填充后的结构化低秩矩阵。
可选的,成像模块1050,还用于执行如下步骤:
对所述填充后的结构化低秩矩阵进行逆变换,得到逆变换矩阵;
利用距离-多普勒方法对所述逆变换矩阵进行成像处理,得到所述逆合成孔径雷达图像。
可选的,所述带有约束的优化问题表示为:
Figure PCTCN2022099364-appb-000059
其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵,S表示稀疏采样下的回波数据对应的矩阵;
所述增广拉格朗日函数表示为:
Figure PCTCN2022099364-appb-000060
其中,R表示辅助变量,Ind C(X)表示指示函数,对Ind C(X)定义为:
Figure PCTCN2022099364-appb-000061
其中,P Ω(·)表示向区域Ω的投影操作。
可选的,所述交替迭代乘子法的迭代步骤表示为:
Figure PCTCN2022099364-appb-000062
Figure PCTCN2022099364-appb-000063
Figure PCTCN2022099364-appb-000064
Figure PCTCN2022099364-appb-000065
i=i+1;
其中
Figure PCTCN2022099364-appb-000066
表示对Ω的补集进行映射,
Figure PCTCN2022099364-appb-000067
表示逆结构化操作,I表示单位矩阵,
Figure PCTCN2022099364-appb-000068
示所述填充后的结构化低秩矩阵。
由此可见,本申请利用结构化操作增强了回波的低秩特性,适用于低采样率下的稀疏逆合成孔径雷达成像。
图11示例了一种电子设备的实体结构示意图,如图11所示,该电子设备可以包括:处理器(processor)1110、通信接口(Communications Interface)1120、存储器(memory)1130和通信总线1140,其中,处理器1110,通信接口1120,存储器1130通过通信总线1140完成相互间的通信。处理器1110可以调用存储器1130中的逻辑指令,以执行上述所述逆合成孔径雷达成像方法,所述方法包括:
接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据;
根据所述回波数据,构建结构化低秩矩阵;
将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值;
基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵;
基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像。
此外,上述的存储器1130中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-Only Memory)、随机存取存储器(RAM,RandomAccess Memory)、磁碟或者广盘等各种可以存储程序代码的介质。
另一方面,本申请还提供一种计算机程序产品,所述计算机程序产品包括 存储在非暂态计算机可读存储介质上的计算机程序,所述计算机程序包括程序指令,当所述程序指令被计算机执行时,计算机能够执行上述各方法所提供的所述逆合成孔径雷达成像方法。
又一方面,本申请还提供一种非暂态计算机可读存储介质,其上存储有计算机程序,该计算机程序被处理器执行时实现以执行上述各提供的所述逆合成孔径雷达成像方法。
以上所描述的装置实施例仅仅是示意性的,其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下,即可以理解并实施。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件。基于这样的理解,上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品可以存储在计算机可读存储介质中,如ROM/RAM、磁碟、广盘等,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行各个实施例或者实施例的某些部分所述的方法。
最后应说明的是:以上实施例仅用以说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。

Claims (13)

  1. 一种逆合成孔径雷达成像方法,其包括:
    接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据;
    根据所述回波数据,构建结构化低秩矩阵;
    将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值;
    基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵;
    基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像;
    其中,所述接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据的步骤包括:
    对所述回波信号进行解线性调频处理,得到解调后的回波信号;
    对所述解调后的回波信号进行平动补偿,得到平动补偿后的回波信号;
    对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据;
    所述回波数据表示为:
    Figure PCTCN2022099364-appb-100001
    其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,x k表示第k个散射点在初始时刻的横坐标,y k表示第k个散射点在初始时刻的纵坐标,γ表示调频斜率,c表示电磁波传播速度,f c表示载波中心频率,ω表示目标转速,f s表示采样频率,PRF表示脉冲重复频率,j表示虚数符号,m=1,2,...,M表示离散形式的快时间序数,n=1,2,...,N表示离散形式的慢时间序数,S(m,n)表示经过稀疏采样的两维回波数据。
  2. 根据权利要求1所述的逆合成孔径雷达成像方法,其中所述结构化低秩矩阵是二层汉克尔矩阵。
  3. 根据权利要求1所述的逆合成孔径雷达成像方法,其中所述解调后的 回波信号表示为:
    Figure PCTCN2022099364-appb-100002
    其中,σ k、x k和y k分别表示第k个散射点的散射系数以及在初始时刻的横纵坐标,T p表示脉冲宽度,γ表示调频斜率,R 0表示目标转动中心与雷达的距离,R ref表示解线性调频的参考距离,f c表示载波中心频率,c表示电磁波传播速度,ω表示目标转速,t r表示快时间,t m表示慢时间,j表示虚数符号。
  4. 根据权利要求2所述的逆合成孔径雷达成像方法,其中所述根据所述回波数据,构建结构化低秩矩阵的步骤包括:
    根据所述回波数据的第n列S(:,n)构造汉克尔矩阵;
    利用所述回波数据的所有列构造二层汉克尔矩阵。
  5. 根据权利要求4所述的逆合成孔径雷达成像方法,其中根据所述第n列S(:,n)构造出的汉克尔矩阵表示为:
    Figure PCTCN2022099364-appb-100003
    利用所述回波数据的所有列构造出的所述二层汉克尔矩阵表示为:
    Figure PCTCN2022099364-appb-100004
    其中,P和Q是束参数。
  6. 根据权利要求2所述的逆合成孔径雷达成像方法,其中所述将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值的步骤包括:
    将所述二层汉克尔矩阵分解为两部分的乘积,即
    Figure PCTCN2022099364-appb-100005
    利用低秩矩阵拟合方法进行秩估计,得到第一初值U (0)和第二初值V (0)
    其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵。
  7. 根据权利要求2所述的逆合成孔径雷达成像方法,其中所述基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵的步骤包括:
    基于所述第一初值与所述第二初值构建带有约束的优化问题;
    利用基于增广拉格朗日函数的交替迭代乘子法对所述带有约束的优化问题求解,以对所述结构化低秩矩阵进行填充,得到所述填充后的结构化低秩矩阵。
  8. 根据权利要求7所述的逆合成孔径雷达成像方法,其中所述基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像的步骤包括:
    对所述填充后的结构化低秩矩阵进行逆变换,得到逆变换矩阵;
    利用距离-多普勒方法对所述逆变换矩阵进行成像处理,得到所述逆合成孔径雷达图像。
  9. 根据权利要求8所述的逆合成孔径雷达成像方法,其中所述带有约束的优化问题表示为:
    Figure PCTCN2022099364-appb-100006
    Figure PCTCN2022099364-appb-100007
    Figure PCTCN2022099364-appb-100008
    其中,U、V H表示所述二层汉克尔矩阵分解后的矩阵,S表示稀疏采样下的回波数据对应的矩阵;
    所述增广拉格朗日函数表示为:
    Figure PCTCN2022099364-appb-100009
    其中,R表示辅助变量,Ind C(X)表示指示函数,对Ind C(X)定义为:
    Figure PCTCN2022099364-appb-100010
    其中,P Ω(·)表示向区域Ω的投影操作。
  10. 根据权利要求9所述的逆合成孔径雷达成像方法,其中所述交替迭代乘子法的迭代步骤表示为:
    Figure PCTCN2022099364-appb-100011
    Figure PCTCN2022099364-appb-100012
    Figure PCTCN2022099364-appb-100013
    Figure PCTCN2022099364-appb-100014
    i=i+1;
    其中,
    Figure PCTCN2022099364-appb-100015
    表示对Ω的补集进行映射,
    Figure PCTCN2022099364-appb-100016
    表示逆结构化操作,I表示单位矩阵,
    Figure PCTCN2022099364-appb-100017
    表示所述填充后的结构化低秩矩阵。
  11. 一种逆合成孔径雷达成像装置,其包括:
    回波信号预处理模块,用于接收雷达发射的线性调频信号所对应的回波信号,并对所述回波信号进行预处理,以得到回波数据;
    矩阵生成模块,用于根据所述回波数据,构建结构化低秩矩阵;
    初值生成模块,用于将所述结构化低秩矩阵进行分解,以得到第一初值与第二初值;
    填充模块,用于基于所述第一初值和所述第二初值对所述结构化低秩矩阵进行填充,以得到填充后的结构化低秩矩阵;
    成像模块,用于基于所述填充后的结构化低秩矩阵得到逆合成孔径雷达图像;
    所述回波信号预处理模块,还用于:
    对所述回波信号进行解线性调频处理,得到解调后的回波信号;
    对所述解调后的回波信号进行平动补偿,得到平动补偿后的回波信号;
    对所述平动补偿后的回波信号进行稀疏采样处理,得到所述回波数据;
    所述回波数据表示为:
    Figure PCTCN2022099364-appb-100018
    其中,δ k表示第k个散射点的散射系数与矩形窗的乘积,x k表示第k个散射点在初始时刻的横坐标,y k表示第k个散射点在初始时刻的纵坐标,γ表示调频斜率,c表示电磁波传播速度,f c表示载波中心频率,ω表示目标转速,f s表示采样频率,PRF表示脉冲重复频率,j表示虚数符号,m=1,2,...,M表示离散形式的快时间序数,n=1,2,...,N表示离散形式的慢时间序数,S(m,n)表示经过稀疏采样的两维回波数据。
  12. 一种电子设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,其中所述处理器执行所述程序时实现如权利要求1至10任一项所述的逆合成孔径雷达成像方法的步骤。
  13. 一种非暂态计算机可读存储介质,其上存储有计算机程序,其中所述计算机程序被处理器执行时实现如权利要求1至10任一项所述的逆合成孔径雷达成像方法的步骤。
PCT/CN2022/099364 2021-09-26 2022-06-17 逆合成孔径雷达成像方法、装置、电子设备及存储介质 WO2023045431A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020247007856A KR20240039054A (ko) 2021-09-26 2022-06-17 역 합성 개구 레이더 이미징 방법, 장치, 전자 장치 및 저장 매체

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111125293.2 2021-09-26
CN202111125293.2A CN113567985B (zh) 2021-09-26 2021-09-26 逆合成孔径雷达成像方法、装置、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023045431A1 true WO2023045431A1 (zh) 2023-03-30

Family

ID=78174338

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/099364 WO2023045431A1 (zh) 2021-09-26 2022-06-17 逆合成孔径雷达成像方法、装置、电子设备及存储介质

Country Status (3)

Country Link
KR (1) KR20240039054A (zh)
CN (1) CN113567985B (zh)
WO (1) WO2023045431A1 (zh)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116449369A (zh) * 2023-06-16 2023-07-18 四川杰诺创科技有限公司 一种基于多范数约束的逆合成孔径雷达成像方法
CN116626647A (zh) * 2023-07-26 2023-08-22 清华大学 面向通用化波形的微动目标回波仿真方法及装置
CN116626646A (zh) * 2023-07-21 2023-08-22 西安电子科技大学 基于时频非均采样的雷达目标无网格化损失相参积累方法
CN116643249A (zh) * 2023-07-27 2023-08-25 西安交通大学 Gfrp内部分层缺陷的毫米波成像可视化定量检测方法
CN116990817A (zh) * 2023-09-26 2023-11-03 北京无线电测量研究所 一种雷达前视无网格重构sar成像方法和装置
CN117250604A (zh) * 2023-11-17 2023-12-19 中国海洋大学 一种目标反射信号与浅海混响的分离方法
CN117706554A (zh) * 2024-02-05 2024-03-15 西安电子科技大学 一种isar回波实时接收的成像处理方法及系统
CN117761694A (zh) * 2023-12-25 2024-03-26 中国人民解放军战略支援部队航天工程大学士官学校 一种压缩感知雷达超分辨成像方法、系统及设备
CN117761636A (zh) * 2024-02-22 2024-03-26 中国科学院空天信息创新研究院 基于稀疏低秩恢复的双基sar相位同步干扰抑制方法

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109633647A (zh) * 2019-01-21 2019-04-16 中国人民解放军陆军工程大学 一种双基地isar稀疏孔径成像方法
WO2019080521A1 (zh) * 2017-10-25 2019-05-02 清华大学 太赫兹isar成像方法及系统
US10310074B1 (en) * 2014-03-27 2019-06-04 Hrl Laboratories, Llc System and method for denoising synthetic aperture radar (SAR) images via sparse and low-rank (SLR) decomposition and using SAR images to image a complex scene
CN110174651A (zh) * 2019-05-16 2019-08-27 河海大学 基于低秩Hankel矩阵补全的气象雷达风电场杂波抑制方法
CN110244303A (zh) * 2019-07-12 2019-09-17 中国人民解放军国防科技大学 基于sbl-admm的稀疏孔径isar成像方法
CN111274525A (zh) * 2020-01-19 2020-06-12 东南大学 一种基于多线性增广拉格朗日乘子法的张量数据恢复方法
CN111812648A (zh) * 2020-07-22 2020-10-23 东南大学 多通道合成孔径雷达rpca幅相联合目标检测方法与装置
CN112684445A (zh) * 2020-12-02 2021-04-20 中国人民解放军国防科技大学 基于md-admm的mimo-isar三维成像方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103513234B (zh) * 2012-06-19 2015-11-04 中国科学院电子学研究所 一种基于矩阵恢复的运动目标快速检测方法及系统
CN107843875A (zh) * 2016-09-19 2018-03-27 南京理工大学 基于奇异值分解降噪的贝叶斯压缩感知雷达数据融合方法
CN109444880A (zh) * 2018-10-23 2019-03-08 重庆大学 一种基于多特征低秩表示融合的sar目标识别方法
CN112946644B (zh) * 2021-01-28 2022-04-19 中国人民解放军国防科技大学 基于最小化卷积加权l1范数的稀疏孔径ISAR成像方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10310074B1 (en) * 2014-03-27 2019-06-04 Hrl Laboratories, Llc System and method for denoising synthetic aperture radar (SAR) images via sparse and low-rank (SLR) decomposition and using SAR images to image a complex scene
WO2019080521A1 (zh) * 2017-10-25 2019-05-02 清华大学 太赫兹isar成像方法及系统
CN109633647A (zh) * 2019-01-21 2019-04-16 中国人民解放军陆军工程大学 一种双基地isar稀疏孔径成像方法
CN110174651A (zh) * 2019-05-16 2019-08-27 河海大学 基于低秩Hankel矩阵补全的气象雷达风电场杂波抑制方法
CN110244303A (zh) * 2019-07-12 2019-09-17 中国人民解放军国防科技大学 基于sbl-admm的稀疏孔径isar成像方法
CN111274525A (zh) * 2020-01-19 2020-06-12 东南大学 一种基于多线性增广拉格朗日乘子法的张量数据恢复方法
CN111812648A (zh) * 2020-07-22 2020-10-23 东南大学 多通道合成孔径雷达rpca幅相联合目标检测方法与装置
CN112684445A (zh) * 2020-12-02 2021-04-20 中国人民解放军国防科技大学 基于md-admm的mimo-isar三维成像方法

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
"Micro-Doppler Effect Analysis and Micro-motion Feature Extraction Technology of Ballistic Targets", 31 December 2018, NATIONAL DEFENSE INDUSTRY PRESS, CN, ISBN: 978-7-118-11668-7, article FENG, CUNQIAN : "Target Echo Model under LFM System", pages: 26 - 28, XP009544910 *
CHEN YUXIN; CHI YUEJIE: "Robust Spectral Compressed Sensing via Structured Matrix Completion", IEEE TRANSACTIONS ON INFORMATION THEORY, vol. 60, no. 10, 1 October 2014 (2014-10-01), USA, pages 6576 - 6601, XP011558693, ISSN: 0018-9448, DOI: 10.1109/TIT.2014.2343623 *
HUANG YAN, ZHAO BO ,TAO MINGLIANG ,CHEN ZHANYE , HONG WEI: "Review of Synthetic Aperture Radar Interference Suppression", JOURNAL OF RADARS, vol. 9, no. 1, 5 March 2020 (2020-03-05), pages 86 - 106, XP093054438, ISSN: 2095-283X, DOI: 10.12000/JR19113 *
QIU WEI; ZHOU JIANXIONG; FU QIANG: "Jointly Using Low-Rank and Sparsity Priors for Sparse Inverse Synthetic Aperture Radar Imaging", IEEE TRANSACTIONS ON IMAGE PROCESSING, vol. 29, 1 January 1900 (1900-01-01), USA, pages 100 - 115, XP011745851, ISSN: 1057-7149, DOI: 10.1109/TIP.2019.2927458 *
USEVICH KONSTANTIN; COMON PIERRE: "Hankel Low-Rank Matrix Completion: Performance of the Nuclear Norm Relaxation", IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING, vol. 10, no. 4, 1 June 2016 (2016-06-01), US , pages 637 - 646, XP011610082, ISSN: 1932-4553, DOI: 10.1109/JSTSP.2016.2535182 *

Cited By (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116449369A (zh) * 2023-06-16 2023-07-18 四川杰诺创科技有限公司 一种基于多范数约束的逆合成孔径雷达成像方法
CN116449369B (zh) * 2023-06-16 2023-08-15 四川杰诺创科技有限公司 一种基于多范数约束的逆合成孔径雷达成像方法
CN116626646A (zh) * 2023-07-21 2023-08-22 西安电子科技大学 基于时频非均采样的雷达目标无网格化损失相参积累方法
CN116626646B (zh) * 2023-07-21 2023-09-22 西安电子科技大学 基于时频非均采样的雷达目标无网格化损失相参积累方法
CN116626647A (zh) * 2023-07-26 2023-08-22 清华大学 面向通用化波形的微动目标回波仿真方法及装置
CN116626647B (zh) * 2023-07-26 2023-10-03 清华大学 面向通用化波形的微动目标回波仿真方法及装置
CN116643249A (zh) * 2023-07-27 2023-08-25 西安交通大学 Gfrp内部分层缺陷的毫米波成像可视化定量检测方法
CN116643249B (zh) * 2023-07-27 2023-10-10 西安交通大学 Gfrp内部分层缺陷的毫米波成像可视化定量检测方法
CN116990817A (zh) * 2023-09-26 2023-11-03 北京无线电测量研究所 一种雷达前视无网格重构sar成像方法和装置
CN116990817B (zh) * 2023-09-26 2023-12-22 北京无线电测量研究所 一种雷达前视无网格重构sar成像方法和装置
CN117250604A (zh) * 2023-11-17 2023-12-19 中国海洋大学 一种目标反射信号与浅海混响的分离方法
CN117250604B (zh) * 2023-11-17 2024-02-13 中国海洋大学 一种目标反射信号与浅海混响的分离方法
CN117761694A (zh) * 2023-12-25 2024-03-26 中国人民解放军战略支援部队航天工程大学士官学校 一种压缩感知雷达超分辨成像方法、系统及设备
CN117706554A (zh) * 2024-02-05 2024-03-15 西安电子科技大学 一种isar回波实时接收的成像处理方法及系统
CN117706554B (zh) * 2024-02-05 2024-05-03 西安电子科技大学 一种isar回波实时接收的成像处理方法及系统
CN117761636A (zh) * 2024-02-22 2024-03-26 中国科学院空天信息创新研究院 基于稀疏低秩恢复的双基sar相位同步干扰抑制方法
CN117761636B (zh) * 2024-02-22 2024-04-30 中国科学院空天信息创新研究院 基于稀疏低秩恢复的双基sar相位同步干扰抑制方法

Also Published As

Publication number Publication date
CN113567985B (zh) 2021-12-07
CN113567985A (zh) 2021-10-29
KR20240039054A (ko) 2024-03-26

Similar Documents

Publication Publication Date Title
WO2023045431A1 (zh) 逆合成孔径雷达成像方法、装置、电子设备及存储介质
CN107462887B (zh) 基于压缩感知的宽幅星载合成孔径雷达成像方法
Xu et al. 3D geometry and motion estimations of maneuvering targets for interferometric ISAR with sparse aperture
Fang et al. Fast compressed sensing SAR imaging based on approximated observation
Yang et al. Segmented reconstruction for compressed sensing SAR imaging
US8493262B2 (en) Synthetic aperture radar image formation system and method
CN110275166A (zh) 基于admm的快速稀疏孔径isar自聚焦与成像方法
Zhang et al. Computationally efficient sparse aperture ISAR autofocusing and imaging based on fast ADMM
CN110068805B (zh) 基于变分贝叶斯推论的高速目标hrrp重构方法
CN106680815B (zh) 基于张量稀疏表示的mimo雷达成像方法
CN109507666B (zh) 基于离网变分贝叶斯算法的isar稀疏频带成像方法
Li et al. ISAR imaging of nonuniformly rotating target based on the multicomponent CPS model under low SNR environment
Andersson et al. Fast Fourier methods for synthetic aperture radar imaging
CN112099007A (zh) 适用于非理想天线方向图的方位向多通道sar模糊抑制方法
Yang et al. High-resolution and wide-swath SAR imaging via Poisson disk sampling and iterative shrinkage thresholding
CN112147608A (zh) 一种快速高斯网格化非均匀fft穿墙成像雷达bp方法
CN108318891B (zh) 一种基于改进sva和cs的sal数据旁瓣的压低方法
Kang et al. SAR image reconstruction via incremental imaging with compressive sensing
CN105929397A (zh) 基于lq 正则化的偏置相位中心天线成像方法
CN108562901B (zh) 基于最大信杂噪比准则的isar高分辨成像方法
Liu et al. Parameter design and imaging method of spaceborne azimuth interrupted FMCW SAR
Rao et al. Comparison of parametric sparse recovery methods for ISAR image formation
Quan et al. An efficient general algorithm for SAR imaging: Complex approximate message passing combined with backprojection
CN113608218A (zh) 一种基于后向投影原理的频域干涉相位稀疏重构方法
CN116027293A (zh) 一种扫描雷达快速稀疏角超分辨方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22871492

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 20247007856

Country of ref document: KR

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE