Vehicle-mounted SAR imaging and motion compensation method based on contrast ratio

Abstract

The invention discloses a contrast-based vehicle-mounted SAR (synthetic aperture radar) imaging and motion compensation method, relates to the technical field of vehicle-mounted SAR imaging, and realizes vehicle-mounted SAR motion compensation without navigation equipment. The method comprises the following steps: acquiring vehicle-mounted SAR original data, and realizing vehicle-mounted SAR coarse imaging by using a vehicle-mounted Omega-K algorithm; designing a contrast-based point selection strategy and a WPGA kernel according to the definition of a standard deviation intensity contrast function; according to an equivalent bunching principle, carrying out sub-aperture division on the full-aperture vehicle-mounted SAR coarse-precision image; performing aperture-by-aperture iterative optimization by using the designed WPGA method to realize sub-aperture SAR image self-focusing; and finally, sub-aperture splicing is carried out to obtain a final vehicle-mounted SAR image. According to the method, the characteristics of the vehicle-mounted signal are analyzed, the self-focusing robustness can be improved by utilizing the contrast optimization WPGA method, large-width and high-precision vehicle-mounted SAR imaging is realized, and technical support is provided for the development of the vehicle-mounted SAR technology.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted SAR imaging technology, and particularly to a vehicle-mounted SAR imaging and motion compensation method based on contrast. Background Art

[0002] Nowadays, more and more vehicles are equipped with autonomous driving or driver assistance functions. As one of the core components of the Advanced Driver Assistance System (ADAS), automotive millimeter-wave radar has become the focus of attention of scholars and manufacturers due to its advantages such as all-weather, miniaturization, high integration, and key sensing capabilities. Automotive millimeter-wave radar operating in the W-band (76 to 81 GHz) is widely used to obtain radial distance, speed, and angular position measurements of targets in the vehicle environment. However, automotive radars from large-scale manufacturers must make trade-offs between angular resolution (usually higher than 1 degree), ranging, bandwidth, and field of view, which poses challenges for high-resolution environmental perception and mapping applications in autonomous driving.

[0003] Many studies have attempted to use Synthetic Aperture Radar (SAR) technology to improve azimuth resolution and enhance the accuracy of automotive environmental perception. The operation of SAR is compatible with the mechanical principles of automotive systems, and automotive SAR images similar to Light Detection and Ranging (LiDAR) Bird's Eye View (BEV) maps can also be used for target detection. The principle of SAR is to utilize the relative motion between the radar and the detected target to synthesize a virtual aperture along the motion trajectory. By comprehensively analyzing the frequency of the radar echo signal and accurately eliminating the range-Doppler coupling caused by the virtual aperture, the azimuth resolution is ultimately improved. The range resolution is only determined by the signal bandwidth, while the azimuth resolution (usually <1 degree) is determined by the synthetic angle.

[0004] In principle, SAR requires navigation accuracy better than the wavelength (4 mm for the W-band), but this requirement is for the relative motion within the synthetic aperture, whose range can reach dozens of centimeters. However, motion compensation is still crucial in automotive SAR imaging. The complexity of the automotive scene and the high mobility of the vehicle will cause the radar to be easily affected by bumps and vibrations during motion. This will introduce additional motion errors, ultimately leading to image defocusing.

[0005] Early motion compensation methods were based on external devices such as Global Navigation Satellite System (GNSS) and Inertial Measurement Unit (IMU) to obtain motion parameters (trajectory, velocity, etc.) that meet accuracy requirements. Subsequently, motion compensation and high-resolution imaging were performed based on these motion parameters. However, with the improvement of imaging resolution, the accuracy of current automotive navigation systems cannot meet the imaging requirements. Therefore, motion compensation remains a key issue in automotive SAR imaging. Autofocus methods have been verified on traditional SAR platforms (such as spaceborne and airborne). These methods can be divided into two major categories: parametric methods and non-parametric methods. Parametric methods usually model motion errors and perform parameter estimation and motion compensation based on model attributes. Non-parametric methods can also be divided into two categories: image optimization-based methods and strong scatterer-based methods. Among them, the standard Phase Gradient Autofocus (PGA), as an effective strong scatterer-based method, includes five main steps: 1) range cell selection; 2) cyclic shift; 3) windowing; 4) phase gradient estimation; 5) iterative phase correction.

[0006] Meanwhile, compared with traditional SAR imaging scenarios, there are many differences in automotive SAR imaging scenarios: 1) Automotive SAR systems usually operate in stripmap mode, using wide beams and Frequency Modulated Continuous Wave (FMCW) signals. This requires techniques such as sub-aperture (SA) division for imaging and autofocus to be applicable to automotive scenarios; 2) The automotive SAR imaging mode is near-field imaging, and the automotive environment is complex with strong clutter, which interferes more strongly with signals; 3) The targets in the strong point scenario of automotive SAR imaging are relatively concentrated and have a certain sparsity in the range dimension. Summary of the Invention

[0007] Object of the Invention: The object of the present invention is to provide a contrast-based vehicle-mounted SAR imaging and motion compensation method to improve the imaging stability of vehicle-mounted SAR systems in vehicle-mounted scenarios with isolated strong points and further optimize the quality of vehicle-mounted SAR images.

[0008] Technical Solution: To achieve the above object, the present invention adopts the following technical solution:

[0009] A contrast-based vehicle-mounted SAR imaging and motion compensation method, comprising:

[0010] Performing SAR imaging on the original FMCW signal of vehicle-mounted SAR using the Omega-K algorithm;

[0011] Dividing the full-aperture coarse-accuracy SAR image into sub-apertures;

[0012] According to the definition of the standard deviation intensity contrast function, the phase error is estimated for each sub-aperture by using the contrast-based WPGA method, and then the sub-aperture phase error is polynomially fitted. The phase error is compensated through the structural characteristics of the phase to obtain a high-precision SAR image of the sub-aperture;

[0013] Perform sub-aperture stitching to obtain a full-aperture high-precision SAR image.

[0014] Furthermore, in the SAR imaging of the original FMCW signal of the vehicle-mounted SAR using the Omega-K algorithm, after performing residual video phase compensation and intra-pulse motion compensation on the original FMCW signal of the SAR, the Omega-K algorithm is used for SAR imaging to obtain a full-aperture rough-precision SAR image.

[0015] Furthermore, using the Omega-K algorithm to perform SAR imaging on the signals compensated for residual video phase and intra-pulse motion to obtain a full-aperture rough-precision SAR image, including multiplying by a reference function multiplier and Stolt interpolation. The reference function multiplier H rfm (K r ,K x ) is expressed as where K r , K x and R 0 represent the range wavenumber, azimuth wavenumber, and reference range respectively, The Stolt interpolation kernel is expressed as where K y =K r -K rc represents the range wavenumber, and K rc =4π / λ represents the reference range wavenumber, and λ represents the signal wavelength.

[0016] Furthermore, in the sub-aperture division of the full-aperture rough-precision SAR image, the sub-aperture length is ηL sub , where η (1≤η≤2) is the overlap rate, and L sub satisfies: where L smin and L sref represent the minimum equivalent spotlight length and the reference equivalent spotlight length respectively, a = 4μtan(θ m / 2)ΔR, ΔR and θ m represent the range imaging range and the signal beam width respectively, and μ (μ≤0.05) represents the effective spotlight ratio.

[0017] Furthermore, in the polynomial fitting of the sub-aperture phase error, the sub-aperture phase error after polynomial fitting is expressed as where Krc represents the reference range wavenumber, and N p represent the coefficients and order of the polynomial fitting, respectively.

[0018] Furthermore, in the phase error compensation through the structural characteristics of the phase, according to the phase structural characteristics, the slant range error ΔR introduced by the Stolt interpolation stolt (K x ) is expressed as where k i represents the coefficient of the polynomial fitting; the sub-aperture phase error compensation function is expressed as where represents the azimuth motion error after polynomial fitting, represents the range motion error after polynomial fitting.

[0019] Furthermore, the standard deviation intensity contrast function C(m) of the vehicle-borne SAR characteristics is defined as

[0020]

[0021] where M and N are the number of range and azimuth sampling points, m and n represent the range and azimuth indices respectively, and I(m,n) is the element at the m-th row and n-th column of the selected sub-aperture SAR image.

[0022] Furthermore, the WPGA method based on contrast includes: contrast-based point selection, circular shift, windowing, calculating the contrast-based weight, estimating the phase error gradient using the WPGA kernel, accumulating the phase error, and iterative phase correction; among them, the contrast-based point selection process is: calculating C(m) for each range cell, setting the threshold σ, and selecting the range cell where C(m)≥σC max where C max is the maximum contrast; the contrast-based weight is The WPGA kernel is expressed as

[0023]

[0024] where arg[·] is the complex amplitude angle function, represents the estimated phase error, and I * (m,n - 1) represents the conjugate of the element at the m-th row and (n - 1)-th column of the selected sub-aperture image.

[0025] The present invention also provides a computer system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the method for vehicle-borne SAR imaging and motion compensation based on contrast.

[0026] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the method for vehicle-mounted SAR imaging and motion compensation based on contrast.

[0027] Beneficial effects: The present invention analyzes the characteristics of vehicle-mounted FMCW signals, designs a sub-aperture autofocus module, and applies it to a vehicle-mounted FMCW-SAR system. At the same time, for the vehicle-mounted imaging scenario with isolated strong points, more and more appropriate target range cells can be selected using contrast, and then the contrast is further used to weight the WPGA kernel, which can improve the robustness of the PGA method in this scenario, further optimize the SAR image quality, obtain wide-swath and high-precision vehicle-mounted SAR imaging, and provide technical support for the application of vehicle-mounted SAR technology. Description of the Drawings

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0029] Figure 1 It is a schematic flowchart of an embodiment of the present invention;

[0030] Figure 2 It is a schematic diagram of the transmitted waveform of a vehicle-mounted FMCW-SAR system;

[0031] Figure 3 It is a geometric model diagram of a vehicle-mounted SAR;

[0032] Figure 4 It is a schematic diagram of the equivalent spotlight mode of a vehicle-mounted SAR;

[0033] Figure 5 It is a schematic diagram of the sub-aperture division strategy of a vehicle-mounted SAR;

[0034] Figure 6 It is a schematic diagram of the sub-aperture division of a vehicle-mounted SAR;

[0035] Figure 7 It is a schematic diagram of the sub-aperture stitching of a vehicle-mounted SAR;

[0036] Figure 8 It is an example diagram of a vehicle-mounted SAR, where (a) is the RF system (TI AWR2243) of the vehicle-mounted SAR; (b) is the installation position of the vehicle-mounted SAR;

[0037] Figure 9 It is the optical image of the experimental scene and the vehicle-mounted SAR image with full-aperture coarse precision;

[0038] Figure 10 This is a comparison chart of experimental results. Among them, (a) is the original SAR image; (b) is the SAR image processed by the standard PGA method based on energy point selection; (c) is the SAR image processed by the standard PGA method based on contrast point selection; (d) is the SAR image processed by the WPGA method based on energy point selection and contrast weighted kernel; (e) is the SAR image processed by the method of the embodiment of the present invention.

[0039] Figure 11 This is the true width image of the metal street lamp in area A;

[0040] Figure 12 This is Figure 10 the contour map of area A in Figure 10 , where (a)-(e) respectively correspond to Figure 10 the contour maps of area A in (a)-(e) of

[0041] Figure 13 This is a comparison chart of experimental results. Among them, (a) is the SAR image processed by the PWE-PGA method based on contrast point selection; (b) is the SAR image processed by the WPGA method based on contrast point selection and variance weighted kernel;

[0042] Figure 14 This is Figure 13 the contour map of area A in Figure 13 , where (a)-(b) respectively correspond to Figure 13 This is Figure 10 the azimuthal sectional view of the metal street lamp in area A in (e) of Detailed implementation manners

[0043] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0044] The embodiments of the present invention provide a vehicle-mounted SAR imaging and motion compensation method based on contrast. The specific process of this method is as Figure 1 shown, and specifically includes the following steps:

[0045] Step 1, Vehicle-mounted SAR rough imaging. In this embodiment, after performing residual video phase compensation and intra-pulse motion compensation on the SAR raw frequency-modulated continuous wave (FMCW) signal, the Omega-K algorithm is used for SAR imaging; specifically including:

[0046] (1.a) Establish a system-level FMCW-SAR signal model. Figure 2 is the transmitted waveform of the vehicle-mounted FMCW-SAR system, and T p is the chirp cycle time. The signal frequency drops to the starting frequency f l within the idle time T s , and then, frequency modulation is performed within the frequency modulation duration T f . Since the analog-to-digital converter (ADC) requires a startup time, that is, the ADC effective startup time T i , therefore, the ADC sampling time T s should be less than T f , the effective signal bandwidth is B r , and the effective starting frequency of the FMCW signal is f c , which can be expressed as f c = f s + γT i , where γ = B r / T s is the chirp frequency modulation rate.

[0047] Therefore, the form of the received signal of the system-level FMCW-SAR is

[0048]

[0049] where A is the signal amplitude, τ and t a represent the fast time and slow time respectively, T is the signal transmission time width, σ(x n , y 0 , z 0 ) is the target complex reflection coefficient, (x n , y 0 , z 0 ) represents the coordinates of the target on the X, Y, and Z axes of the Cartesian coordinate system, rect[(τ - Δτ) / T - 1 / 2] and g(t a ) are the range dimension envelope of the received signal and the antenna pattern respectively, f c and γ represent the carrier frequency and frequency modulation rate of the transmitted signal respectively, and Δτ is the round-trip delay of the target; at this time, the signal is in the range wavenumber domain - azimuth time domain.

[0050] The geometric model of the vehicle-mounted SAR is as Figure 3 shown. Assuming that the vehicle is operating in the strip mode and moving a distance of vT along the X-axis a , where v is the vehicle reference speed and Ta The synthetic aperture time is \(T_a\), the radar reference altitude is \(H\), and the beam width is \(\theta\). m , and the reference slant range is \(R\). 0 , the reference point is \(P(x\) 0 ,y 0 ,z 0 ), the instantaneous slant angle is \(\theta\), and the SAR instantaneous position is \(C(x\) a ,y a ,z a ). Considering a target \(Q(x\) n ,y 0 ,z 0 ) at any point on the reference slant range, the instantaneous slant range \(R(R\) 0 ,X)\) is expressed as

[0051]

[0052] where X = vt a , \(\Delta x\), \(\Delta y\), \(\Delta z\) are motion errors, and \(\Delta R(R\) 0 ,X)\) is the slant range error introduced by the motion error. Therefore, the two-way time delay \(\Delta\tau\) is expressed as \(\Delta\tau = 2R(R\) 0 ,X) / c, where \(c\) is the speed of light.

[0053] (1.b) Perform residual video phase compensation and intra-pulse motion compensation on the received signal. The signal after residual video phase compensation is expressed as

[0054] s(K r ,X) = P(K r )·g(X)·exp[jK r R(R\) 0 ,X)], (3)

[0055] where r K c = 4\pi[f l +\gamma)\tau - T i )] / c represents the range wavenumber, \(P(K\) r ) and \(g(X)\) are the range wavenumber domain envelope and the azimuth antenna pattern; the intra-pulse motion compensation function \(H\) im (\tau,K\) x ) is expressed as:

[0056] H im (\tau,K\) x ) = exp(jK x v\tau), (4) where \(K\) x represents the azimuth wavenumber. The two-dimensional wavenumber spectrum of the compensated signal is expressed as

[0057]

[0058] Among them, P(K r ) represents the range wavenumber domain envelope in the range dimension, and G(K x ) is the wavenumber domain envelope in the azimuth dimension.

[0059] (1.c) Use the Omega-K algorithm to perform SAR imaging on the signal obtained in step (1.b) to obtain a full-aperture coarse-precision SAR image, including multiplying by a reference function multiplier and Stolt interpolation; specifically, the reference function multiplier H rfm (K r , K x ) is expressed as

[0060]

[0061] Here, directly multiply H rfm (K r , K x ) with the signal S(K r , K x ). Subsequently, perform Stolt interpolation for each range cell. The Stolt interpolation kernel is expressed as Among them, K y = K r - K rc represents the range wavenumber, and K rc represents the reference range wavenumber. Finally, perform a two-dimensional IFFT on the signal to achieve focusing and obtain a full-aperture coarse-precision SAR image.

[0062] Step 2, sub-aperture division. In this embodiment, use the equivalent spotlight principle to perform sub-aperture division on the full-aperture coarse-precision SAR image; specifically include:

[0063] (2.a) According to the sub-aperture imaging range and the ratio μ (μ ≤ 0.05) of the effective spotlight area, the schematic diagram of the vehicle-mounted SAR equivalent spotlight mode is as Figure 4 shown. S is the imaging area, R min and R max represent the minimum and maximum imaging distances. The orange area S spot is the equivalent spotlight area within the sub-aperture L m . L smin and L sref respectively represent the minimum and reference equivalent spotlight aperture lengths. The sub-aperture division strategy is as Figure 5 shown. Calculate the sub-aperture length range to obtain

[0064]

[0065] Among them, L sub is the sub-aperture length, a = 4μtan(θm / 2)ΔR, ΔR represents the distance imaging range;

[0066] (2.b) Select an appropriate value L within the sub-aperture length range sub , set the overlap ratio η (1 ≤ η ≤ 2), and divide the full-aperture coarse-precision SAR image described in step 1 into n a sub-apertures, with the sub-aperture length being ηL sub , obtaining n a sub-aperture coarse-precision SAR images. The schematic diagram of sub-aperture division is as shown in Figure 6 .

[0067] Step 3, sub-aperture SAR autofocus. Among them, according to the definition of the standard deviation intensity contrast function, use the contrast-based WPGA method to estimate the phase error for each sub-aperture, then perform polynomial fitting on the sub-aperture phase error, and compensate for the phase error through the structural characteristics of the phase to obtain a high-precision sub-aperture SAR image; specifically including:

[0068] (3.a) In the sub-aperture image described in step 2, use the contrast-based WPGA method to estimate the phase error and obtain the sub-aperture phase error;

[0069] The contrast-based WPGA method includes: contrast-based point selection, circular shift, windowing, calculating the contrast-based weight, using the WPGA kernel to estimate the phase error gradient, accumulating the phase error, and iterative phase correction;

[0070] Specifically, the contrast-based point selection strategy is expressed as C(m) ≥ σC max , where σ ∈ (0, 1] and C max represent the threshold and the maximum contrast respectively, and C(m) is defined as

[0071]

[0072] where M and N are the number of range and azimuth sampling points, m and n represent the range and azimuth indices respectively, and I(m, n) is the element at the m-th row and n-th column of the selected sub-aperture SAR image; here, select the range cells where C(m) ≥ σC max . In this embodiment, the minimum selection number is set to 50 and the maximum selection number is M / 2.

[0073] The contrast-based weight is expressed as

[0074]

[0075] The WPGA kernel is expressed as

[0076]

[0077] where arg[·] is the complex argument function, denotes the estimated phase error, I * (m,n - 1) represents the conjugate of the element in the m-th row and (n - 1)-th column of the selected sub-aperture image.

[0078] The obtained phase gradients are accumulated to obtain a phase error vector, and the central value is subtracted to obtain the sub-aperture phase error.

[0079] (3.b) Perform polynomial fitting on the sub-aperture phase error, and use the phase structure characteristics of the signal to compensate the sub-aperture phase error to obtain n a high-precision SAR images of sub-apertures;

[0080] The polynomial fitting of the sub-aperture phase error is expressed as

[0081]

[0082] where, is the sub-aperture phase error after polynomial fitting, and N p represent the coefficients and order of polynomial fitting respectively, K x i represents the i-th power of K x , ΔR res (K x ) is the residual slant range error; here, N p is set to 10;

[0083] The phase structure characteristics of the signal are expressed as

[0084]

[0085] where, ΔR stolt (K x ) is the slant range error introduced by Stolt interpolation, k i represents the coefficient of polynomial fitting;

[0086] The sub-aperture phase error compensation function H e (K y , K x ) is expressed as

[0087]

[0088] Here, denotes the azimuth motion error after polynomial fitting, denotes the range motion error after polynomial fitting. The process of phase error compensation is to perform azimuthal FFT on the sub-aperture image data described in step 2 and then multiply it by He (K y , K x ) are multiplied, and then the azimuth - dimension IFFT is performed to obtain a sub - aperture high - precision SAR image.

[0089] Step 4: According to the sub - aperture division principle in Step 2, sub - aperture stitching is performed to obtain a full - aperture high - precision SAR image. The schematic diagram of sub - aperture stitching is as shown in Figure 7 shown.

[0090] The effectiveness of the present invention can be further illustrated by the following experiments.

[0091] (I) Data and experimental scenario description

[0092] For the RF system of the vehicle - mounted SAR, the TI AWR2243 radar development board is selected, as shown in (a) of Figure 8 , and its installation position is as shown in (b) of Figure 8 . The vehicle collects data without navigation equipment such as an inertial navigation system. The experimental parameters of the vehicle - mounted SAR are shown in Table 1. The size of the original echo data of the vehicle - mounted SAR finally obtained is 1024×182000 (range×azimuth). The experimental scenario is selected as an office building scenario, and there are strong scatterers in the scenario.

[0093] Table 1 Experimental parameters of vehicle - mounted SAR

[0094]

[0095] (II) SAR rough imaging results

[0096] The optical image of the experimental scenario and the full - aperture rough - precision SAR image are as shown in Figure 9 . The size of the scenario is approximately 15m×100m (range×azimuth). Among them, the SAR image is divided into 9 red square areas, and each area corresponds to an optical image; it can be seen from Figure 9 that there are many details: Areas A and E are two metal street lamps, Area B is a sign behind the street lamp, Area C is the floor - to - ceiling window of the office building, and Area D is a van moving in the opposite direction. Although the Omega - K algorithm can perform full - aperture vehicle - mounted SAR imaging, its image still has defocusing, which is caused by motion errors and requires autofocus processing.

[0097] (III) SAR autofocus results

[0098] In this experiment, R min , μ, σ, η are set to 0.5m, 0.05, 0.6, and 1.5 respectively. After sub - aperture division, 167 sub - aperture rough - precision SAR images are obtained. The autofocus method is used to optimize the images, and after sub - aperture stitching, a full - aperture high - precision vehicle - mounted SAR image is obtained.

[0099] 1. Comparison of Point Selection Strategies

[0100] Here, the results of different point selection strategies are compared. The comparison methods include the point selection strategy based on energy and the point selection strategy based on variance. The comparison methods also include the standard PGA method and the method of the embodiment of the present invention; Figure 10 The comparison results are given, where, Figure 10 in, (a) is the original SAR image; (b) is the SAR image processed by the standard PGA method with point selection based on energy; (c) is the SAR image processed by the standard PGA method with point selection based on contrast; (d) is the SAR image processed by the WPGA method with point selection based on energy and contrast weighted kernel; (e) is the SAR image processed by the method of the embodiment of the present invention. It can be seen that the autofocus method is effective in the vehicle-mounted scenario, and among them, the performance of the method of the embodiment of the present invention is the best.

[0101] For a clearer comparison, analyze the area A in Figure 10 . Area A is a metal street lamp, and its actual width is 100 mm, as Figure 11 shown. The contour map of area A is as shown in (a)-(e) in Figure 12 , where, Figure 12 in, (a) is the contour map of area A in (a) of Figure 10 ; (b) is the contour map of area A in (b) of Figure 10 ; (c) is the contour map of area A in (c) of Figure 10 ; (d) is the contour map of area A in (d) of Figure 10 ; (e) is the contour map of area A in (e) of Figure 10 . It can be seen that obvious defocusing appears in the original image, and the method of the embodiment of the present invention is closest to the true value; Figure 12 (f) of Figure 10 is the azimuth sectional view of the metal street lamp in area A of

[0102] . The reason for the two peaks is that the shape of the street lamp is a rectangular column, and the radar cross section at the edge is larger. By comparison, it can be seen that the autofocus effect of the method of the embodiment of the present invention is the best.

[0103] Table 2 Performance Indicators of Autofocus Methods

[0104] Method Contrast Entropy Sub - graph calculation time Original image 7.0964 1.1677 \ Standard PGA method based on energy - selected points 8.6017 1.0157 0.4898s Standard PGA method based on contrast - selected points 9.6640 0.8635 0.6505s WPGA method based on energy - selected points and contrast - weighted kernel 9.2738 0.9777 0.5508s Method of the present invention 10.035 0.7886 0.6578s

[0105] 2. Comparison of PGA Kernels

[0106] The results of different PGA cores are compared here. The point selection strategy adopts the contrast-based point selection strategy, and the comparison methods include the PWE-PGA method and the variance-based WPGA method; Figure 13 Figures (a)-(b) show the comparison results. Among them, (a) is the SAR image processed by the PWE-PGA method with contrast-based point selection; (b) is the SAR image processed by the WPGA method with contrast-based point selection and variance-weighted kernel; Compared with Figure 10 Figure (e), it can be seen that the PGA core performance of the method of the embodiment of the present invention is superior to that of the PWE-PGA method and the variance-based WPGA method;

[0107] Similarly, the metal street lamp in area A is large, Figure 14 Figure (a) of Figure 13 is the contour map of area A in Figure (a) of Figure 14 Figure (b) of Figure 13 is the contour map of area A in Figure (b) of Figure 14 Figure (c) of Figure 13 Compared with Figure 10 The azimuth profile of the metal street lamp in area A in Figure (e) of

[0108] Table 3 Performance indicators of the PWE-PGA method and the variance-based WPGA method

[0109] Method Contrast Entropy Sub - graph calculation time PWE - PGA method based on contrast - selected points 8.8442 1.0302 0.4582s WPGA method based on contrast - selected points and variance - weighted kernel 9.2571 0.9220 0.4943s

[0110] In summary, the present invention provides a vehicle-mounted SAR imaging and motion compensation method based on contrast for vehicle-mounted imaging scenarios with isolated strong points. By using contrast, more and more appropriate target range cells can be selected, and then the contrast is further used to weight the WPGA core, which improves the algorithm robustness, optimizes the quality of vehicle-mounted SAR images, and can achieve large swath width and high-precision imaging of vehicle-mounted SAR.

[0111] The embodiment of the present invention also provides a computer system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the steps of the vehicle-mounted SAR imaging and motion compensation method based on contrast are implemented.

[0112] An embodiment of the present invention further provides a computer program product, including a computer program, which when executed by a processor implements the steps of the method for vehicle-mounted SAR imaging and motion compensation based on contrast as described above.

[0113] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, so that when the program codes are executed by the processor or controller, the steps of the method of the present invention are implemented. The program codes can be executed entirely on the machine, partially on the machine, executed partially on the machine and partially on a remote machine as an independent software package, or executed entirely on a remote machine or server. Those parts not detailed in the present invention are all well-known technologies to those skilled in the art.

[0114] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention through logical analysis, reasoning, or limited experiments based on the concept of the present invention on the basis of the prior art should fall within the protection scope determined by the claims.

Claims

1. A contrast-based vehicle-mounted SAR imaging and motion compensation method, characterized in that: include: The original FMCW signal of vehicle-mounted SAR is imaged using the Omega-K algorithm; Divide the full-aperture coarse-precision SAR image into sub-apertures; According to the definition of standard deviation intensity contrast function, the contrast-based WPGA method is used to estimate the phase error sub-aperture by sub-aperture, and then a polynomial fitting is performed on the sub-aperture phase error. The phase error is compensated by the structural characteristics of the phase to obtain a sub-aperture high-precision SAR image. Perform sub-aperture stitching to obtain full-aperture high-precision SAR images.

2. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 1, characterized in that: The SAR imaging using the Omega-K algorithm on the original FMCW signal of the vehicle-mounted SAR includes performing residual video phase compensation and intra-pulse motion compensation on the original SAR FMCW signal, and then performing SAR imaging using the Omega-K algorithm to obtain a full-aperture coarse-precision SAR image.

3. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 2, characterized in that: The Omega-K algorithm is used to perform SAR imaging on the residual video phase compensation and intra-pulse motion compensation signals to obtain a full-aperture coarse-precision SAR image, including multiplication with the reference function multiplier and Stolt interpolation. The reference function multiplier H rfm (K r ,K x ) is expressed as Among them, K r , K x and R0 represent the range wave number, azimuth wave number and reference distance respectively, The Stolt interpolation kernel is expressed as Among them, K y =K r -K rc Indicates the distance wave number, K rc =4π / λ represents the reference distance wave number, and λ represents the signal wavelength.

4. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 1, characterized in that: In the sub-aperture division of the full-aperture coarse-precision SAR image, the sub-aperture length is ηL sub , η(1≤η≤2) is the overlap rate, L sub satisfy: Among them, L smin and L sref They represent the minimum equivalent bunching length and the reference equivalent bunching length, respectively, a=4μtan(θ m / 2)ΔR, ΔR and θ m They represent the distance imaging range and signal beam width respectively, and μ (μ≤0.05) represents the effective beam focusing ratio.

5. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 1, characterized in that: In the polynomial fitting of the sub-aperture phase error, the sub-aperture phase error after the polynomial fitting is Expressed as Among them, K x represents the azimuthal wave number, K rc represents the reference distance wave number, and N p represent the coefficients and order of the polynomial fit respectively.

6. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 5, characterized in that: In the phase error compensation by the structural characteristics of the phase, according to the phase structural characteristics, the slant range error ΔR introduced by Stolt interpolation stolt (K x ) is expressed as Among them, k i represents the coefficient of polynomial fitting; the subaperture phase error compensation function is expressed as Among them, K y Indicates the distance wave number, K rc represents the reference distance wave number, represents the azimuth motion error after polynomial fitting, Represents the range motion error after polynomial fitting.

7. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 1, characterized in that: The standard deviation intensity contrast function C(m) of the vehicle-borne SAR characteristics is defined as Where M and N are the number of range and azimuth sampling points, m and n represent the range and azimuth indexes, respectively, and I(m,n) is the element in the mth row and nth column of the selected sub-aperture SAR image.

8. The contrast-based vehicle-mounted SAR imaging and motion compensation method according to claim 7, characterized in that: The contrast-based WPGA method includes: contrast-based point selection, circle shifting, windowing, calculation of contrast-based weights, estimation of phase error gradient using WPGA kernel, accumulation of phase error and iterative phase correction; the contrast-based point selection process is: calculation of C(m) of each distance unit, setting threshold σ, selecting C(m)≤σC max The distance unit, C max is the maximum contrast; the weight based on contrast is The WPGA kernel is represented as in, is the complex argument function, represents the estimated phase error, I * (m,n-1) represents the conjugate of the element in the mth row and n-1th column of the selected sub-aperture image.

9. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is executed by a processor, the steps of the contrast-based vehicle-mounted SAR imaging and motion compensation method according to any one of claims 1 to 8 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the contrast-based vehicle-mounted SAR imaging and motion compensation method according to any one of claims 1 to 8 are implemented.

Images