Forward-looking scanning radar frequency domain rapid imaging method based on Gaussian prior

By using Gaussian prior and Fourier transform in forward-view scanning radar, the time domain deconvolution problem is converted into frequency domain, and the problems of high computational complexity and low efficiency in the prior art are solved, and efficient radar is realized Imaging.

CN118915061AActive Publication Date: 2024-11-08UNIV OF ELECTRONICS SCI & TECH OF CHINA

Patent Information

Application Number
CN202410934750.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-12
Publication Date
2024-11-08
Estimated Expiration
2044-07-12

AI Technical Summary

Technical Problem

When the prior art improves the azimuth resolution of forward-view scanning radar, the calculation complexity is high, the imaging efficiency is low, and it is difficult to achieve efficient radar imaging.

Method used

Using the Gaussian prior-based frequency domain fast imaging method, the time domain is smoothed by adding a smooth filter operator, and the time domain deconvolution problem is converted into the frequency domain by using Fourier transform, and the frequency domain multiplication property is reduced. Computational complexity.

Benefits of technology

The imaging quality and efficiency are improved, the computational complexity is reduced, and the imaging process has higher efficiency without losing quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118915061A_ABST
    Figure CN118915061A_ABST
Patent Text Reader

Abstract

The invention discloses a foresight scanning radar frequency domain fast imaging method based on Gaussian priori, which comprises the following steps: firstly obtaining echo data and preprocessing, then carrying out Bayesian reasoning based on a Gaussian priori Bayesian framework, then constructing a time domain deconvolution problem, adding a smooth filtering operator in an equivalent regularization process, and finally carrying out frequency domain fast imaging on a foresight scanning radar. The deconvolution problem of the time domain is converted to the frequency domain through Fourier transform, and rapid frequency domain solving is carried out. According to the method, a smooth filtering operator is added when a deconvolution problem in a time domain is processed, noise and details of an image can be smoothed, meanwhile, the basic structure of the image is reserved, the imaging quality is improved, Fourier transform is converted into a frequency domain, and the time domain convolution of Fourier transform is equal to the property of frequency domain multiplication, so that the image quality is improved. Inverse operation is successfully converted into division operation, the calculation complexity is greatly reduced, and higher imaging efficiency is achieved under the condition that the imaging quality is not lost.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of radar imaging, and in particular relates to a frequency domain rapid imaging method of a forward-looking scanning radar based on Gaussian prior. Background Art

[0002] High-resolution imaging of the forward-looking area of ​​airborne scanning radar has urgent application needs in the fields of blind landing navigation and sea detection. The antenna aperture of real aperture radar is limited, so the azimuth resolution is low. In order to improve the azimuth resolution, the paper "Yin Zhang, Changlin Li, Deqing Mao, Yulin Huang, and Jianyu Yang, 'Bayesian superresolution method of forward-looking imaging with generalized gaussian constraint, 'in IGARSS2018-2018IEEE International Geoscience and Remote Sensing Symposium, 2018, pp.51285131" proposed an adjustable angle super-resolution method to achieve high azimuth resolution in the forward-looking area in scanning radar imaging, and discussed how to select appropriate scale parameters to obtain better super-resolution performance in different scenes, but the method has low computational efficiency. The paper "Yin Zhang, Jiahao Shen, Xingyu Tuo, Haiguang Yang, Yongchao Zhang, Yulin Huang, and Jianyu Yang, 'Scanning radar forward-looking superresolution imaging based on the weibull distribution for a sea-surface target,' IEEE Transactions on Geoscience and Remote Sensing, vol.60, pp.1-11, 2022" proposes a super-resolution imaging method based on Weibull distribution. The method introduces generalized Gaussian distribution and Weibull distribution to represent the statistical distribution functions of target prior information and sea clutter, respectively, and derives the corresponding objective function under the maximum a posteriori (MAP) criterion. In order to solve the nonlinear problem of the objective function, the paper uses the Newton-Raphson iterative method to solve the problem. Simulation and experimental data evaluation show that compared with other existing sea surface target super-resolution methods, the proposed method has better super-resolution imaging performance, but the computational complexity is high and the computational efficiency is low.The paper "Yasser Radouane Haddadi and Boualem Mansouri, 'Ultrasound medical image deconvolution using l2 regularization method and artificial bee colony optimization algorithm,' in 2022 7th International Conference on Image and Signal Processing and their Applications (ISPA), 2022, pp. 1-6" proposed a new method for ultrasound medical image deconvolution based on regularization parameter estimation based on L2 regularization method and artificial bee colony optimization algorithm. The method studied the influence of regularization parameter selection on the deconvolution process and evaluated the regularization parameter estimation, which confirmed that the method improved the efficient performance of ultrasound medical image deconvolution operation, but the computational efficiency was low.

[0003] Although the above method can avoid the pathological characteristics caused by direct deconvolution reconstruction of the target and improve the azimuth resolution of the forward-looking scanning radar, the imaging calculation complexity is high and the imaging efficiency is low. Summary of the invention

[0004] To solve the above technical problems, the present invention provides a frequency domain rapid imaging method for a forward-looking scanning radar based on Gaussian prior. When processing the deconvolution problem in the time domain, a smoothing filter operator is added to smooth the noise and details of the image while retaining the basic structure of the image and improving the imaging quality. The time domain is converted into the frequency domain through Fourier transform, and the property of Fourier transform that time domain convolution is equal to frequency domain multiplication is utilized to successfully convert the inverse operation into a division operation, thereby greatly reducing the computational complexity.

[0005] The technical solution adopted by the present invention is: a frequency domain rapid imaging method of forward-looking scanning radar based on Gaussian prior, and the specific steps are as follows:

[0006] Step 1: echo data acquisition and preprocessing;

[0007] A forward-looking scanning radar azimuth echo convolution model is constructed. The flight altitude of the carrier platform is set to H, the movement direction is along the Y-axis direction of the three-dimensional coordinate system, the speed is V, the radar beam scans counterclockwise at a speed of ω', the beam pitch angle is α, and the carrier platform is located at point A directly above the coordinate origin at time zero.

[0008] Assume that a target P in space is located at the center of the beam at time zero, its distance from the carrier platform is R0, and its horizontal azimuth is The spatial azimuth angle is θ0, and it can be known from the spatial geometric relationship that

[0009] At time t, the aircraft platform moves from point A to point D. At this time, the horizontal azimuth angle of the target P relative to the aircraft platform is The spatial azimuth angle is θ, and the distance R(t) between the carrier platform and the target is expressed as follows:

[0010]

[0011] The Taylor series expansion of the distance history expression at t = 0 is as follows:

[0012]

[0013] The distance history expression can be approximated as follows:

[0014] R(t)≈R0-Vcosθ0t(3)

[0015] Assume that the radar transmission signal is a linear frequency modulation signal S(τ), the expression is as follows:

[0016]

[0017] Among them, τ represents the distance time variable, T r represents the pulse width of the linear frequency modulation signal, f0 represents the carrier frequency, K r represents the frequency modulation slope, and rect[·] represents the window function in the distance to time domain, as follows:

[0018]

[0019] For any point target P in the imaging scene, the echo signal received by the forward scanning radar is expressed as follows after down-conversion:

[0020]

[0021] Where σ0 represents the target scattering coefficient, ψ(t) represents the antenna pattern modulation function, t represents the azimuth time variable, and τ d =2R(t) / c represents the round-trip echo delay, and c represents the speed of light.

[0022] The antenna pattern modulation function is expressed as represents the transposed antenna pattern function, and

[0023] Assuming the slant range of any point target in the imaging scene is R, and the spatial azimuth is θ, we can get:

[0024]

[0025] Substituting equations (7) and (8) into the signal expression, we can obtain:

[0026]

[0027] Among them, B represents the bandwidth of the transmitted signal, λ represents the wavelength, the last term is the Doppler phase caused by the platform motion, σ represents the target scattering coefficient σ0 and the two-way echo delay τ d The product of the additional phase caused by is expressed as follows:

[0028]

[0029] Assume the target unit impulse response function is h(R,θ), the expression is as follows:

[0030]

[0031] The echo signal after pulse pressure and movement correction is expressed in a two-dimensional convolution form, as follows:

[0032]

[0033] in, The symbol represents a two-dimensional convolution operation, δ(·) represents the impulse function, and ignoring the Doppler phase in equation (12), a more simplified form of the point target echo is obtained, which is expressed as follows:

[0034]

[0035] in, represents the target scattering distribution function.

[0036] The target echo y is expressed as the convolution form of the antenna measurement matrix h′ and the target scattering coefficient x, and the expression is as follows:

[0037] y=h′*x (14)

[0038] Where * represents the convolution operator. The Richardson-Lucy deconvolution algorithm is used to solve x in equation (14).

[0039] A frequency domain solution strategy is proposed to estimate x, and equation (14) is rewritten as follows:

[0040] y=hx (15)

[0041] Where h is an N×N matrix, representing the antenna pattern matrix, y is an N×1 vector, x is an N×1 vector, and N represents the number of azimuth sampling points. Formula (15) is expressed in frequency domain as follows:

[0042] Y(ω)=H(ω)X(ω) (16)

[0043] Among them, Y(ω), H(ω) and X(ω) are the frequency domain representations of y, h and x, and ω represents the coordinates in the frequency domain.

[0044] If the echo is affected by noise, equation (16) can be expressed as follows:

[0045] Y(ω)=H(ω)X(ω)+N(ω) (17)

[0046] Where N(ω) represents the frequency domain representation of the noise.

[0047] If the antenna pattern matrix h is a full-rank matrix, the target scattering coefficient x is determined by solving the inverse function. When x is calculated in the frequency domain, the antenna pattern matrix is ​​low-rank, and the high-frequency component H(ω) is 0 or close to 0. The echo is affected by noise, and the expression of X(ω) is as follows:

[0048]

[0049] Step 2: Perform Bayesian reasoning based on the Gaussian prior Bayesian framework;

[0050] The goal is to determine the most likely solution for x based on y, as follows:

[0051]

[0052] Where p(·) represents the probability distribution, represents the maximum a posteriori estimate of x.

[0053] Assuming that the noise follows a Gaussian distribution with a mean of 0 and a variance of ξ, the likelihood function is derived, and the expression is as follows:

[0054]

[0055] For x prior, use a set of filters g k To characterize, the expression is as follows:

[0056]

[0057] Among them, β represents the weight, i represents the number of sampling points, f(·) represents the filter function, g i,k represents the kth filter centered at i.

[0058] Select the filter, i.e. the horizontal derivative g x = [1, -1] and the vertical derivative g y =[1,-1] T filter.

[0059] Where T stands for transpose.

[0060] G x , g y Substituting into formula (21), the expression is as follows:

[0061]

[0062] Substituting equations (20) and (21) into (19), we get the maximum a posteriori expression of x as follows:

[0063]

[0064] Then the negative logarithm expression of formula (23) is as follows:

[0065]

[0066] Where w = βξ 2 , the estimated value of x is obtained by minimizing formula (24).

[0067] Step 3: Based on step 2, construct the time domain deconvolution problem;

[0068] Assuming the prior is a Gaussian prior, the differential equation (24) is obtained as follows:

[0069]

[0070] Let equation (25) be zero, and we get the following expression:

[0071]

[0072] Among them, C gk Indicates g i,k When I represents the identity matrix, and formula (26) is equivalent to L2 regularization, and its solution expression is as follows:

[0073]

[0074] Formula (27) is the azimuth solution of the distance unit. ADMM is used to optimize the solution process and obtain The estimated value expression of is as follows:

[0075]

[0076] Where H, X and Y represent the antenna matrix, target matrix and echo matrix respectively. Z and U represent the introduced relaxation variables, and ρ represents the Lagrangian parameter. represents the shrinkage parameter, (k) represents the kth iteration, and the shrinkage operator The expression is as follows:

[0077]

[0078] Step 4: Convert the deconvolution problem in the time domain to the frequency domain through Fourier transform and perform fast frequency domain solution;

[0079] The vector form in equation (26) is converted into a matrix form to speed up the solution. The expression is as follows:

[0080]

[0081] The optimal solution is derived by solving a set of sparse linear equations A'X=B', which is expressed as follows:

[0082]

[0083] Then A'X=B' can be expressed as the following expression:

[0084]

[0085] Solving A'X=B' in the frequency domain, the frequency domain expression obtained is as follows:

[0086]

[0087] Among them, G k (ω) represents C gk The Fourier transform of .

[0088] Solving equation (33), we get the expression of X(ω) as follows:

[0089]

[0090] Through inverse transformation, the scattering coefficient of the target is solved, and the expression is as follows:

[0091]

[0092] Here, IFFT(·) represents the inverse Fourier transform operator.

[0093] Beneficial effects of the present invention: The method of the present invention first obtains echo data and performs preprocessing, then performs Bayesian reasoning based on the Gaussian prior Bayesian framework, and then constructs the time domain deconvolution problem, adds a smoothing filter operator in the equivalent regularization process, and converts the time domain deconvolution problem to the frequency domain through Fourier transform to perform a fast frequency domain solution. The method of the present invention adds a smoothing filter operator when processing the deconvolution problem in the time domain, which can smooth the noise and details of the image while retaining the basic structure of the image, improving the imaging quality, and converting it into the frequency domain through Fourier transform. The property that the Fourier transform time domain convolution is equal to the frequency domain multiplication is used to successfully convert the inverse operation into a division operation, which greatly reduces the computational complexity and has higher imaging efficiency without losing imaging quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] Figure 1 The present invention is a flow chart of a method for rapid frequency domain imaging of a forward-looking scanning radar based on Gaussian prior.

[0095] Figure 2 This is a convolution model diagram of the azimuth echo of the forward-looking scanning radar in an embodiment of the present invention.

[0096] Figure 3 This is a diagram of the simulation imaging result in an embodiment of the present invention.

[0097] Figure 4 1 is an imaging time diagram of different algorithms in an embodiment of the present invention. DETAILED DESCRIPTION

[0098] The present invention adopts simulation experiments to demonstrate the effectiveness of the method of the present invention. All steps and conclusions of the method of the present invention are verified on the Matlab2019b simulation platform. In order for relevant personnel in this field to understand the content of the invention, the method of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0099] like Figure 1 As shown, a flow chart of a method for fast frequency domain imaging of a forward-looking scanning radar based on Gaussian prior of the present invention is shown, and the specific steps are as follows:

[0100] Step 1: echo data acquisition and preprocessing;

[0101] Construct a forward-looking scanning radar azimuth echo convolution model. The forward-looking scanning radar azimuth echo convolution model is as follows: Figure 2 As shown in the figure, the flight height of the carrier platform is H, the movement direction is along the Y-axis direction of the three-dimensional coordinate system, the speed is V, the radar beam scans counterclockwise at a speed of ω', the beam pitch angle is α, and the carrier platform is located at point A directly above the origin of the coordinate system at time zero. The radar system simulation parameters in this example are shown in Table 1.

[0102] Table 1

[0103] Simulation parameters Numeric Carrier frequency 10GHz bandwidth 60MHz Main lobe beamwidth 3° Pulse repetition frequency 4000Hz Scan speed 30° / s Scan range ±5°

[0104] A target P in space is located at the center of the beam at time zero, with a distance R0 relative to the carrier platform and a horizontal azimuth angle of The spatial azimuth angle is θ0, and it can be known from the spatial geometric relationship that

[0105] At time t, the aircraft platform moves from point A to point D. At this time, the horizontal azimuth angle of the target P relative to the aircraft platform is The spatial azimuth angle is θ, and the distance R(t) between the carrier platform and the target is expressed as follows:

[0106]

[0107] The Taylor series expansion of the distance history expression at t = 0 is as follows:

[0108]

[0109] Because the radar antenna beam scans quickly, it takes a short time to scan a target. In addition, the forward-looking scanning radar generally has a long range. In practical applications, the influence of the quadratic term in equation (2) is often ignored. The distance history expression can be approximated as follows:

[0110] R(t)≈R0-Vcosθ0t(3)

[0111] Assume that the radar transmission signal is a linear frequency modulation signal S(τ), the expression is as follows:

[0112]

[0113] Among them, τ represents the distance time variable, T r represents the pulse width of the linear frequency modulation signal, f0 represents the carrier frequency, K r represents the frequency modulation slope, and rect[·] represents the window function in the distance to time domain, as follows:

[0114]

[0115] For any point target P in the imaging scene, the echo signal received by the forward scanning radar is expressed as follows after down-conversion:

[0116]

[0117] Where σ0 represents the target scattering coefficient, ψ(t) represents the antenna pattern modulation function, t represents the azimuth time variable, and τ d =2R(t) / c represents the round-trip echo delay, and c represents the speed of light.

[0118] The antenna pattern modulation function is expressed as represents the transposed antenna pattern function, and

[0119] From formula (6), it can be seen that the echo is a two-dimensional time variable matrix about variables τ and t. For ease of understanding, it can be compared to the "stop-go-stop" mode of echo reception in SAR imaging. During the scanning process of the forward-looking scanning radar, when the antenna beam scans to a certain angle, it can be considered that the carrier platform and the antenna beam remain stationary at this time, and a linear frequency modulation pulse signal is emitted. The signal propagates to the irradiated area at the speed of light and then reflects back to be received by the receiver. After down-conversion and channelization, a complex baseband echo signal is obtained and stored as a column of the data matrix. Then the platform continues to move forward. When the antenna beam scans to the next angle position, it remains stationary at this position, and then emits and receives pulse signals, which are stored in the next column of the data matrix. By analogy, a two-dimensional echo data matrix is ​​finally obtained. According to the generation process of the echo matrix, it can be known that the column vector of the matrix reflects the target distance information, which is called the range direction, that is, the τ direction; the row vector of the matrix reflects the target azimuth angle information, which is called the azimuth direction, that is, the t direction.

[0120] Assuming the slant range of any point target in the imaging scene is R, and the spatial azimuth is θ, we can get:

[0121]

[0122] Substituting equations (7) and (8) into the signal expression, we can obtain:

[0123]

[0124] Among them, B represents the bandwidth of the transmitted signal, λ represents the wavelength, the last term is the Doppler phase caused by the platform motion, σ represents the target scattering coefficient σ0 and the two-way echo delay τ d The product of the additional phase caused by is expressed as follows:

[0125]

[0126] Assume that the point target unit impulse response function is h(R,θ), and the expression is as follows:

[0127]

[0128] The echo signal after pulse pressure and movement correction is expressed in a two-dimensional convolution form, as follows:

[0129]

[0130] in, The symbol represents a two-dimensional convolution operation, δ represents the impulse function, and ignoring the Doppler phase in equation (12), a more simplified form of the point target echo is obtained, which is expressed as follows:

[0131]

[0132] in, represents the target scattering distribution function.

[0133] The target echo y is expressed as the convolution form of the antenna measurement matrix h′ and the target scattering coefficient x, and the expression is as follows:

[0134] y=h′*x (14)

[0135] Where * represents the convolution operator. The Richardson-Lucy deconvolution algorithm is used to solve x in equation (14). However, this algorithm can only solve problems in the real domain, but not in the complex domain. When the platform moves, the imaging accuracy is low.

[0136] A frequency domain solution strategy is proposed to estimate x, and equation (14) is rewritten as follows:

[0137] y=hx (15)

[0138] Where h is an N×N matrix, representing the antenna pattern matrix, y is an N×1 vector, x is an N×1 vector, and N represents the number of azimuth sampling points. Formula (15) is expressed in frequency domain as follows:

[0139] Y(ω)=H(ω)X(ω) (16)

[0140] Among them, Y(ω), H(ω) and X(ω) are the frequency domain representations of y, h and x, and ω represents the coordinates in the frequency domain.

[0141] If the echo is affected by noise, equation (16) can be expressed as follows:

[0142] Y(ω)=H(ω)X(ω)+N(ω) (17)

[0143] Where N(ω) represents the frequency domain representation of the noise.

[0144] If the antenna pattern matrix h is a full rank matrix, the target scattering coefficient x is determined by solving the inverse function. It is just as easy to calculate x in the frequency domain. However, the antenna pattern matrix is ​​low rank, which causes the high-frequency component H(ω) to be 0 or close to 0. The target belongs to the low-frequency component, while the noise fills the entire spectrum. In the high-frequency part, the noise will be amplified, which greatly affects the quality of imaging. The echo is affected by the noise, and the expression of X(ω) is as follows:

[0145]

[0146] Step 2: Perform Bayesian reasoning based on the Gaussian prior Bayesian framework;

[0147] In order to solve the above problem, the use of prior information in the deconvolution process is conducive to the recovery of the target scattering coefficient x. That is, the goal is to determine the most likely solution of x based on y.

[0148]

[0149] Where p(·) represents the probability distribution, represents the maximum a posteriori estimate of x.

[0150] Assuming that the noise follows a Gaussian distribution with a mean of 0 and a variance of ξ, the likelihood function is derived, and the expression is as follows:

[0151]

[0152] For x prior, use a set of filters g k To characterize, the expression is as follows:

[0153]

[0154] Among them, β represents the weight, i represents the number of sampling points, f(·) represents the filter function, g i,k represents the kth filter centered at i.

[0155] The simplest filter choice is the horizontal derivative g x = [1, -1] and the vertical derivative g y =[1,-1] T filter.

[0156] Where T stands for transpose.

[0157] G x , g y Substituting into formula (21), the expression is as follows:

[0158]

[0159] The above priors show that images tend to be smooth and their derivatives are close to zero.

[0160] Substituting equations (20) and (21) into (19), we get the maximum a posteriori expression of x as follows:

[0161]

[0162] In order to simplify the calculation, the negative logarithm expression of formula (23) is as follows:

[0163]

[0164] Where w = βξ 2 , the estimated value of x is obtained by minimizing formula (24).

[0165] Step 3: Based on step 2, construct the time domain deconvolution problem;

[0166] The prior is set to be a Gaussian prior. In fact, most images satisfy the Gaussian prior property, that is, the function f(z) is |z 2 Differentiating equation (24) yields the following expression:

[0167]

[0168] Let equation (25) be zero, and we get the following equation:

[0169]

[0170] Among them, C gk Indicates g i,k When I represents the identity matrix, and formula (26) is equivalent to L2 regularization, and its solution expression is as follows:

[0171]

[0172] Formula (27) is the azimuth solution of the distance unit, which is inefficient. ADMM is used to optimize the solution process and obtain The estimated value expression of is as follows:

[0173]

[0174] Where H, X and Y represent the antenna matrix, target matrix and echo matrix respectively. Z and U represent the introduced relaxation variables, and ρ represents the Lagrangian parameter. represents the shrinkage parameter, (k) represents the kth iteration, and the shrinkage operator The expression is as follows:

[0175]

[0176] Step 4: Convert the deconvolution problem in the time domain to the frequency domain through Fourier transform and perform fast frequency domain solution;

[0177] The vector form in equation (26) is converted into a matrix form to speed up the solution. The expression is as follows:

[0178]

[0179] Formula (28) gives the iterative solution method of the entire echo matrix, but the inverse operation cannot be avoided. Then the optimal solution is derived by solving a set of sparse linear equations A'X=B', which is expressed as follows:

[0180]

[0181] Then A'X=B' can be expressed as the following expression:

[0182]

[0183] The simplest and most effective way to solve A'X=B' is in the frequency domain, and the frequency domain expression obtained is as follows:

[0184]

[0185] Among them, G k (ω) represents The Fourier transform of .

[0186] Solving equation (33), we get the expression of X(ω) as follows:

[0187]

[0188] Through inverse transformation, the scattering coefficient of the target is solved, and the expression is as follows:

[0189]

[0190] Here, IFFT(·) represents the inverse Fourier transform operator.

[0191] In order to prove the effectiveness of the method of the present invention, a simulation experiment is carried out on the Matlab2019b platform.

[0192] The most computationally complex part of the algorithm of the present invention is the Fourier transform and the inverse Fourier transform, but it still proves to be very effective compared to other operations (such as inversion). Table 2 compares the computational complexity of the algorithm proposed by the method of the present invention with other L2 regularization related algorithms. It can be seen that the complexity of the method of the present invention is much lower than that of other processing methods.

[0193] Table 2

[0194] method Complexity L2 Method <![CDATA[M(O(N 3 +N 2 ))]]> Batch L2 Methods <![CDATA[max{O(N 3 ),O(MN 2 )}]]> Method of the present invention O(MNlog(MN))

[0195] Where M represents the number of sampling points in the distance direction.

[0196] Figure 3 This is the simulation imaging result diagram in this example. Figure 3 (a) is the target scene, Figure 3 (b) is the echo. The imaging results of the method of the present invention are compared with those obtained by the L2 correlation method and the Lucy method. Figure 3 (c) and Figure 3 As shown in (d), both Lucy’s method and existing L2 methods cannot distinguish between two objects that are very close to each other. Figure 3 (e) and Figure 3As shown in (f), the batch L2 method and the method of the present invention can effectively separate these targets. The results show that compared with the existing L2 method and Lucy method, the method of the present invention has excellent resolution performance. In addition, the method of the present invention shows similar performance to the batch L2 method. Table 3 is a comparison of the imaging efficiency of different algorithms. By comparing the processing time shown in Table 3, it can be seen that the processing efficiency of the method of the present invention is more than 30 times that of other methods.

[0197] Table 3

[0198] method Lucy Existing L2 L2 batching Method of the present invention Time(s) 1.4031 14.3993 1.7733 0.0460

[0199] In order to demonstrate the effectiveness of the method of the present invention, the distance and azimuth dimensions are respectively set to a fixed value of 300 in this embodiment. Then, the imaging time of various methods is compared by increasing the size of another dimension.

[0200] Figure 4 is the imaging time diagram of different algorithms in this example, Figure 4 (a) is the imaging time of different methods (changing the number of sampling points in azimuth), Figure 4 (b) is the imaging time of different methods (changing the number of sampling points in the distance direction). When processing large scenes, the method of the present invention is obviously more effective than the existing L2 method and batch L2 method. Although these methods may require up to 100 seconds of processing time, the method of the present invention only requires 0.1 seconds. On the other hand, the Lucy method has good processing efficiency. However, the processing efficiency of the method of the present invention is more than 100 times higher than when processing large scenes.

[0201] In summary, under the Bayesian framework, the Bayesian derivation of Gaussian prior distribution and Gaussian likelihood distribution is equivalent to the L2 regularization optimization problem. The method of the present invention is based on the Gaussian prior Bayesian framework. A smoothing filter operator is added in the equivalent regularization process, which can smooth the noise and details of the image while retaining the basic structure of the image and improving the imaging quality. The deconvolution problem in the time domain is converted to the frequency domain by Fourier transform. By using the property that the time domain convolution is equal to the frequency domain multiplication after Fourier transform, the inversion operation is successfully converted to a division operation, which greatly reduces the computational complexity. In the entire imaging process, the main computational complexity of the method of the present invention comes from the Fourier transform and the inverse Fourier transform. This computational cost is extremely low compared to the inversion, and can ensure the real-time application of the imaging algorithm. Compared with methods such as L2 regularization and Lucy-Richardson (a more extensive deconvolution method), the method of the present invention has higher imaging efficiency without losing imaging quality.

[0202] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for fast frequency domain imaging of forward-looking scanning radar based on Gaussian prior, the specific steps are as follows: Step 1: echo data acquisition and preprocessing; Construct a forward-looking scanning radar azimuth echo convolution model, set the flight altitude of the carrier platform to H, the moving direction along the Y-axis of the three-dimensional coordinate system, the speed to V, the radar beam scans counterclockwise at a speed of ω', the beam pitch angle is α, and the carrier platform is located at point A directly above the origin of the coordinate system at time zero; Assume that a target P in space is located at the center of the beam at time zero, its distance from the carrier platform is R0, and its horizontal azimuth is The spatial azimuth angle is θ0, and it can be known from the spatial geometric relationship that At time t, the aircraft platform moves from point A to point D. At this time, the horizontal azimuth angle of the target P relative to the aircraft platform is The spatial azimuth angle is θ, and the distance R(t) between the carrier platform and the target is expressed as follows: The Taylor series expansion of the distance history expression at t = 0 is as follows: The distance history expression can be approximated as follows: R(t)≈R0-Vcosθ0t (3) Assume that the radar transmission signal is a linear frequency modulation signal S(τ), the expression is as follows: in, τ represents the distance time variable, T r represents the pulse width of the linear frequency modulation signal, f0 represents the carrier frequency, K r represents the frequency modulation slope, and rect[·] represents the window function in the distance to time domain, as follows: For any point target P in the imaging scene, the echo signal received by the forward scanning radar is expressed as follows after down-conversion: Where σ0 represents the target scattering coefficient, ψ(t) represents the antenna pattern modulation function, t represents the azimuth time variable, and τ d =2R(t) / c represents the two-way echo delay, c represents the speed of light; The antenna pattern modulation function is expressed as represents the transposed antenna pattern function, and Assuming the slant range of any point target in the imaging scene is R, and the spatial azimuth is θ, we can get: Substituting equations (7) and (8) into the signal expression, we can obtain: Among them, B represents the bandwidth of the transmitted signal, λ represents the wavelength, the last term is the Doppler phase caused by the platform motion, σ represents the target scattering coefficient σ0 and the two-way echo delay τ d The product of the additional phase caused by is expressed as follows: Assume the target unit impulse response function is h(R,θ), the expression is as follows: The echo signal after pulse pressure and movement correction is expressed in a two-dimensional convolution form, as follows: in, The symbol represents a two-dimensional convolution operation, δ(·) represents the impulse function, and ignoring the Doppler phase in equation (12), a more simplified form of the point target echo is obtained, which is expressed as follows: in, represents the target scattering distribution function; The target echo y is expressed as the convolution form of the antenna measurement matrix h′ and the target scattering coefficient x, and the expression is as follows: y=h′*x (14) Where * represents the convolution operator; the Richardson-Lucy deconvolution algorithm is used to solve x in equation (14); A frequency domain solution strategy is proposed to estimate x, and equation (14) is rewritten as follows: y=hx (15) Where h is an N×N matrix, representing the antenna pattern matrix, y is an N×1 vector, x is an N×1 vector, and N represents the number of azimuth sampling points. Formula (15) is expressed in frequency domain as follows: Y(ω)=H(ω)X(ω) (16) Among them, Y(ω), H(ω) and X(ω) are the frequency domain representations of y, h and x, and ω represents the coordinates in the frequency domain; If the echo is affected by noise, equation (16) can be expressed as follows: Y(ω)=H(ω)X(ω)+N(ω) (17) Where N(ω) represents the frequency domain representation of noise; If the antenna pattern matrix h is a full-rank matrix, the target scattering coefficient x is determined by solving the inverse function; when x is calculated in the frequency domain, the antenna pattern matrix is ​​low-rank, and the high-frequency component H(ω) is 0 or close to 0; the echo is affected by noise, and the expression of X(ω) is as follows: Step 2: Perform Bayesian reasoning based on the Gaussian prior Bayesian framework; The goal is to determine the most likely solution for x based on y, as follows: Where p(·) represents the probability distribution, represents the maximum a posteriori estimate of x; Assuming that the noise follows a Gaussian distribution with a mean of 0 and a variance of ξ, the likelihood function is derived, and the expression is as follows: For x prior, use a set of filters g k To characterize, the expression is as follows: Among them, β represents the weight, i represents the number of sampling points, f(·) represents the filter function, g i,k represents the kth filter centered at i; Select the filter, i.e. the horizontal derivative g x = [1, -1] and the vertical derivative g y =[1,-1] T filter; Where T represents transpose; G x , g y Substituting into formula (21), the expression is as follows: Substituting equations (20) and (21) into (19), we get the maximum a posteriori expression of x as follows: Then the negative logarithm expression of formula (23) is as follows: Where w = βξ 2 , obtain the estimated value of x by minimizing formula (24); Step 3: Based on step 2, construct the time domain deconvolution problem; Assuming the prior is a Gaussian prior, the differential equation (24) is obtained as follows: Let equation (25) be zero, and we get the following expression: Among them, C gk Indicates g i,k The linear operator of I represents the identity matrix, and formula (26) is equivalent to L2 regularization, and its solution expression is as follows: Formula (27) is the azimuth solution of the distance unit. ADMM is used to optimize the solution process and obtain The estimated value expression of is as follows: X (k) =(H T H+wI) -1 H T Y+ρ((Z)−(U)) U (k) =U (k-1) +X (k) -Z (k) Where H, X and Y represent the antenna matrix, target matrix and echo matrix respectively; Z and U represent the introduced relaxation variables, ρ represents the Lagrangian parameter, represents the shrinkage parameter, (k) represents the kth iteration, and the shrinkage operator The expression is as follows: Step 4: Convert the deconvolution problem in the time domain to the frequency domain through Fourier transform and perform fast frequency domain solution; The vector form in equation (26) is converted into a matrix form to speed up the solution. The expression is as follows: The optimal solution is derived by solving a set of sparse linear equations A'X=B', which is expressed as follows: B'=H T Y Then A'X=B' can be expressed as the following expression: Solving A'X=B' in the frequency domain, the frequency domain expression obtained is as follows: Among them, G k (ω) represents C gk Fourier transform of Solving equation (33), we get the expression of X(ω) as follows: Through inverse transformation, the scattering coefficient of the target is solved, and the expression is as follows: Here, IFFT(·) represents the inverse Fourier transform operator.

Citation Information

Patent Citations

  • Method for realizing angular super-resolution imaging of forward-looking sea surface targets in sea clutter background

    CN104950306A

  • Deconvolution method for realizing scanning radar azimuth super-resolution imaging

    CN104977582A

  • Scanning radar forward-looking imaging orientation super-resolution method

    CN110109097A

  • Rapid super-resolution method for radar foresight imaging

    CN115372968A

  • ITR-DTV radar foresight super-resolution imaging method based on Renyi entropy

    CN115902886A

Cited By

  • Forward-looking super-resolution rapid imaging method and system for airborne scanning radar

    CN122506560A