A two-dimensional high-resolution imaging method for radar targets based on rotating antenna

Abstract

The present invention provides a two-dimensional high-resolution imaging method for radar targets based on a rotating antenna. The method comprises the following steps: first, constructing a rotating antenna observation model, performing Fresnel approximation processing on the distance between the radar and the target; performing "Dechirp" processing on the echo data, and then performing Fourier transform on the obtained single-frequency pulse signal to obtain the corresponding sinc-shaped narrow pulse in the distance frequency domain, and extracting the peak of the narrow pulse to obtain the distance information of the target; second, extracting the signal S of each distance unit unit (t m ), perform time-frequency analysis on the signal and substitute the approximate distance into S unit (t m ) to obtain the Doppler shift expression. The target's elevation angle is estimated based on the time-frequency plot. This elevation angle is used to construct a sparse dictionary, and the target's azimuth is reconstructed using the orthogonal matching pursuit (OMP) algorithm. This imaging method boasts a simple model design, low cost, and low algorithmic complexity. It can achieve high-resolution two-dimensional imaging of targets at all elevation and azimuth angles, and can achieve stable and effective imaging of targets in low signal-to-noise ratio scenarios, demonstrating excellent robustness.

Description

Technical Field

[0001] The present invention relates to signal and information processing technology, and in particular to a two-dimensional high-resolution imaging method of a radar target based on a rotating antenna. Background Art

[0002] Radar imaging has the characteristics of all-day and all-weather operation, and has been widely used in the field of target detection and recognition. It is a hot area of radar technology research.

[0003] Inverse Synthetic Aperture Radar (ISAR) imaging is a traditional two-dimensional imaging method that relies on the relative motion between the target and the radar to image the moving target. Z. Xingyu et al. proposed a near-field target ISAR imaging method based on coordinate conversion and image interpolation in "Approach for ISAR imaging of near-field targets based on coordinate conversion and image interpolation" (Journal of Systems Engineering and Electronics, 2021, 64(9):3850-3857). The simplified uniform linear motion model is applied to near-field target imaging, and the near-field distortion correction method of the range-Doppler (RD) algorithm is combined to overcome the complexity of the traditional ISAR near-field imaging algorithm and obtain a better near-field target ISAR focused image than the backprojection (BP) algorithm. In recent years, vortex electromagnetic waves have been used for radar imaging. Guo Guirong et al. proposed in “Radar Target Imaging Based on Electromagnetic Vortex” (Journal of National University of Defense Technology, 2013, 035(006): 71-76) that electromagnetic vortexes have the potential to image radar targets in azimuth. This research provides a reference and reference for the design of new radar systems and the development of target recognition technology. Liu Kang et al. proposed in “Research Progress of Vortex Electromagnetic Waves and Their Applications in Radar” (Acta Electronica Sinica, 2018, 46(9): 2283-2290) that the spiral phase wavefront distribution of vortex electromagnetic waves and the orthogonality between different orbital angular momentum eigenstates can be used to perform staring imaging of the target without relative motion between the target and the radar.

[0004] However, ISAR imaging requires a high number of pulses and motion compensation, making it difficult to image and placing high demands on radar performance. Furthermore, because ISAR imaging requires relative motion between the radar and the target, it cannot image stationary targets. Electromagnetic vortex imaging also requires a high number of orbital angular momentum modes. However, current technology is unable to generate pure, high-modal vortex electromagnetic waves, significantly limiting the radar's ability to accurately resolve target azimuths. Summary of the Invention

[0005] The present invention aims to overcome the shortcomings of the aforementioned imaging methods by proposing a two-dimensional, high-resolution imaging method for radar targets based on a rotating antenna. This imaging method obtains Doppler information by rotating the antenna in a circular motion, enabling high-resolution azimuth imaging of stationary targets. Combined with range information obtained through "Dechirp" processing, this method enables high-resolution two-dimensional imaging of stationary targets. The model is simple in design, requiring only two antennas: a fixed transmitter and a rotating receiver for echo data acquisition. This approach places low demands on radar performance and reduces imaging costs.

[0006] The present invention is achieved in the following ways:

[0007] Step 1: Construct a rotating antenna observation model and perform Fresnel approximation on the distance between the radar and the target. Perform "Dechirp" processing on the echo data and then perform Fourier transform on the single-frequency pulse signal after "Dechirp" processing to obtain the corresponding sinc-shaped narrow pulse in the range frequency domain. By extracting the peak of the narrow pulse, the target distance information can be obtained.

[0008] Step 2: Extract the signal S of each distance unit unit (t m ), perform time-frequency analysis on the signal; substitute the approximate distance into S unit (t m ), obtain the Doppler shift expression, estimate the target's pitch angle according to the time-frequency diagram, construct a sparse dictionary based on the pitch angle, and reconstruct the target azimuth using the orthogonal matching pursuit (OMP) algorithm.

[0009] The step 1 specifically includes the following steps:

[0010] Step 1) Construct a rotating antenna observation model and perform Fresnel approximation on the distance between the radar and the target to obtain the approximate distance r p (t m ), r p (t m ) is determined by the distance from the target to the center of the circle, the radius of the circle, the target pitch angle and the azimuth angle, and the echo delay τ is calculated. p (t m );

[0011] Step 2) Echo signal s echo (t m ,t) and the reference signal s ref (t) conjugate multiplication to obtain a single-frequency pulse signal, where t represents the fast time and t m Indicates slow time;

[0012] Step 3) Perform Fast Fourier Transform (FFT) on the single-frequency pulse signal to obtain the corresponding sinc-shaped narrow pulse in the frequency domain, and then compare it with the phase compensation function S de (f r ) and remove the residual video phase (RVP) and envelope tilt term to obtain the echo signal S in the range frequency domain echo (t m ,f r ), where f r represents the distance frequency;

[0013] Step 4) r =-γτ p (t m ) to extract S echo (t m ,f r ) peak, obtain the distance information of the scattering point;

[0014] The step 2 specifically includes the following steps:

[0015] Step 1) Extract the signal S of each distance unit unit (t m ), for S unit (t m ) to perform time-frequency analysis and obtain the time-frequency diagram of the signal, that is, the Doppler frequency shift H p (t m ) changes with slow time, and performs mathematical morphological image processing on the time-frequency graph to extract its skeleton features.

[0016] Step 2) Substitute the approximate distance into the distance unit signal S unit (t m ), obtain the phase φ(t m ), for φ(t m ) Take the derivative to get the Doppler frequency shift expression H at the scattering point p (t m ), in this model, H p (t m ) is a constant amplitude with phase changing with slow time t m Changing sine function, H p (t m) contains the pitch angle information of the target, so the pitch angle of the scattering point can be estimated by using the maximum value of the time-frequency diagram;

[0017] Step 3) Build an azimuth dictionary The estimated pitch angle and azimuth angle dictionary is used to construct the sparse dictionary matrix D. Given the known observation matrix y = S unit (t m ) and dictionary matrix D, the azimuth is reconstructed using the OMP algorithm The optimization objective of the OMP algorithm can be expressed as: in is the optimization objective, X is a column vector, and ||||2 represents the matrix norm;

[0018] Step 4) The index position corresponding to the non-zero element in X is the azimuth In the azimuth dictionary The corresponding position in the image can be reconstructed, so the target azimuth can be reconstructed. Combined with the distance information obtained in step 1, a high-resolution image of the target's range-azimuth can be obtained.

[0019] The beneficial effects of the present invention are: on the one hand, it overcomes the shortcomings of existing ISAR imaging methods that require relative motion between the radar and the target, cannot image stationary targets, and need to accumulate a large number of pulses, which makes imaging difficult and requires high performance of the radar. On the other hand, it avoids the problem that high-resolution electromagnetic vortex imaging has high requirements for the number of orbital angular momentum modes, but pure high-modal vortex electromagnetic waves cannot be obtained under existing technical conditions. In response to the specific problems of the above-mentioned ISAR imaging and electromagnetic vortex imaging, a two-dimensional high-resolution imaging method of radar targets based on a rotating antenna is proposed. It can generate Doppler information through the rotation of the antenna itself, thereby performing two-dimensional high-resolution imaging of stationary targets. The imaging method model is simple in design, low in cost, and low in algorithm complexity. After simulation verification, the imaging method can perform high-resolution two-dimensional imaging of targets at all pitch angles and all angles, and can perform stable and effective imaging of targets in scenarios where the signal-to-noise ratio is greater than -24dB, demonstrating good noise resistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flow chart of the method of the present invention;

[0021] Figure 2 It is the geometric relationship diagram between radar and target;

[0022] Figure 3(a) shows the ratio changing with target distance, Figure 3(b) shows the ratio changing with pitch angle, and Figure 3(c) shows the ratio changing with azimuth angle.

[0023] Figure 4(a) shows the variation of the approximate error with target distance, Figure 4(b) shows the variation of the approximate error with pitch angle, and Figure 4(c) shows the variation of the approximate error with azimuth angle.

[0024] Figure 5 is the scattering point model diagram;

[0025] Figure 6(a) is the signal time-frequency diagram when the distance unit is 425m, Figure 6(b) is the Gaussian mask result when the distance unit is 425m, Figure 6(c) is the binarization result, and Figure 6(d) is the skeleton extraction result.

[0026] Figure 7(a) shows the range imaging result, Figure 7(b) shows the azimuth imaging result, and Figure 7(c) shows the two-dimensional imaging result when the signal-to-noise ratio is 0 dB.

[0027] Figure 8 This is the azimuth profile diagram using the matched filter imaging method under this model;

[0028] FIG9(a) is a diagram showing how the imaging quality changes with target distance, FIG9(b) is a diagram showing how the imaging quality changes with pitch angle, FIG9(c) is a diagram showing how the imaging quality of the proposed method changes when the pitch angle is 0.8655 rad, and FIG9(d) is a diagram showing how the imaging quality changes with azimuth angle;

[0029] Figure 10(a) is a diagram showing the change in imaging quality with signal-to-noise ratio, Figure 10(b) is a diagram showing the imaging results of the proposed inventive method when the signal-to-noise ratio is -24dB, and Figure 10(c) is a diagram showing the imaging results of the proposed inventive method when the signal-to-noise ratio is -19dB. DETAILED DESCRIPTION

[0030] The present invention will be further described below with reference to the accompanying drawings and examples of the present invention.

[0031] like Figure 1 As shown, the present invention is implemented through the following steps: constructing a rotating antenna observation model, and using the Fresnel method to approximate the distance between the target and the radar under far-field conditions. The antenna located at the center of the circle transmits a linear frequency modulation signal, and the antenna performing uniform circular motion on the circle receives the target echo signal. First, the echo signal is subjected to "Dechirp" processing, and then a fast Fourier transform is performed in the fast time domain to obtain the sinc function. The distance information is obtained through the peak of the sinc function, and the signal of each distance unit is extracted. Time-frequency analysis is performed on it to obtain the amplitude of its Doppler frequency shift. The pitch angle of the target is estimated by the amplitude, and a sparse dictionary is constructed using the pitch angle. The signal of the distance unit is used as the observation value, and the azimuth angle of the target is reconstructed using the OMP algorithm, thereby achieving two-dimensional high-resolution imaging of the target. The specific instructions are as follows:

[0032] Step 1: Build a rotating antenna observation model to obtain the target echo signal, and reconstruct the target distance through range-directed matched filtering.

[0033] like Figure 2 The figure shows the observation model of the rotating antenna. There is a scattering point P in the free space, which can be expressed in the polar coordinate system as where r p is the distance from the scattering point to the origin of the coordinate system, θ p is the scattering point elevation angle, is the scattering point azimuth, In the space rectangular coordinate system, it can be expressed as P(x p ,y p ,z p ), the projection point of the scattering point P on the XOY plane is M. The radar takes point (a,0,0) as the initial point and the origin O as the center of the circle, and performs counterclockwise uniform circular motion with a radius of a. The frequency of the radar circular motion is f. The radar position changes with the slow time t m The changing process Q(t m ) can be expressed as

[0034] Q(t m )=(acos(2πft m ),asin(2πft m ),0) (1)

[0035] The distance r between the radar and the scattering point p (t m ) can be expressed as

[0036] r p (t m )=||(x p ,y p ,z p )-Q(t m )|| (2)

[0037] In the formula, || || represents the modulus of the vector.

[0038] Using the Fresnel approximation, (2) can be approximated in polar coordinates as

[0039]

[0040] The distance between the radar and the scattering point r p (t m ) is the Taylor expansion of

[0041]

[0042] The condition for satisfying (3) is that the third term of the Taylor expansion can be ignored, that is,

[0043]

[0044] Where, λ is the signal wavelength. So we can get

[0045]

[0046] Radar transmits linear frequency modulation signal

[0047]

[0048] During the duration of a pulse, the scattering point and the radar can be regarded as relatively stationary, so the received echo s echo (t m ,t) can be expressed as

[0049]

[0050] Where: t is the fast time, σ p is the scattering coefficient, rect(·) is the rectangular window function, T p is the pulse duration, f c is the carrier frequency, γ is the modulation frequency, and the echo delay j is the imaginary unit, c is the speed of light. When there are n scattering points, the echo s echo (t m ,t) can be rewritten as

[0051]

[0052] Assume that the reference signal s ref (t) is:

[0053] s ref (t)=exp(-j2π(f c t+0.5γt 2 )) (10)

[0054] Perform "Dechirp" processing on (9), S echo (t m ,f r ) is obtained:

[0055]

[0056] Where: f r is the distance frequency, FFT{·} is the fast Fourier transform operation, and the phase compensation function S de (f r)=exp(-j3πf r 2 / γ), the purpose is to remove the RVP and envelope tilt terms. Substituting (3) into (11) we can get

[0057]

[0058] Its peak is at f r =-γτ p (t m ), the echo delay τ p (t m ) contains the target distance information. At this point, the target distance is reconstructed.

[0059] Step 2: Perform time-frequency analysis on the signals of each range unit, estimate the target pitch angle, build a sparse dictionary, and use the OMP algorithm to reconstruct the target azimuth

[0060] Assume that there are L scattering points in a single distance unit, and the signal S of a single distance unit unit (t m ) can be expressed as

[0061]

[0062] make The Doppler shift of a single scattering point is H p (t m )for

[0063]

[0064] Since the pitch angle range of the radar beam is small, the Doppler frequency shift of different scattering points will result in sine functions with different initial phases and similar amplitudes. The OMP algorithm can be applied to (13) to estimate the azimuth angles of different scattering points. In order to construct the dictionary required by the OMP algorithm, the pitch angles of the scattering points must be estimated first. Applying the short-time Fourier transform to (13) to obtain the time-frequency diagram of a single distance unit, and then performing skeleton extraction on the time-frequency diagram, a relatively accurate peak point can be obtained, thereby estimating the pitch angle of the scattering point. The specific steps of skeleton extraction are: first, smooth the time-frequency map with a Gaussian spatial mask, then convert the Gaussian mask result into a binary image, and then extract the skeleton of the binary image to obtain a more accurate time-frequency map, which greatly improves the estimation accuracy of the pitch angle. pmax |, estimate the pitch angle of the scattering point

[0065]

[0066] Where: arcsin(·) is the inverse sine function.

[0067] Reusing the pitch angle estimate Construct dictionary D.

[0068] D=[d1,d2,…,d m ] (16)

[0069] Where: column vector Azimuth Dictionary m represents the number of equal divisions of the azimuth angle within the range of (-π, π).

[0070] The optimization objective of the OMP algorithm can be expressed as

[0071]

[0072] Where: || ||2 is the matrix norm, X is the column vector. unit (t m ) and dictionary D, the azimuth is reconstructed using the OMP algorithm So far, the two-dimensional imaging results in the range and azimuth directions based on the rotating antenna have been obtained.

[0073] Example: Simulation experiment of two-dimensional high-resolution imaging of radar targets based on rotating antenna

[0074] Simulation experiment: To verify the effectiveness of the proposed method, a 3D imaging simulation was performed using a point target as a model. The transmitted signal was a linear frequency modulation signal. The simulation parameters are shown in Table 1. The position parameters are shown in Table 2.

[0075] Table 1 Simulation parameter settings

[0076]

[0077] Table 2 Point target position parameters

[0078]

[0079]

[0080] Simulation 1: In order to analyze the impact of the approximate error of the distance between the radar and the target on the imaging quality, the following experiment is conducted. The Taylor expansion of the distance between the radar and the target can be approximated to the second order, which needs to meet the conditions of formula (6). Now we explore the change process of the ratio on the left side of formula (6) with the target distance, azimuth and pitch angle. The results are shown in Figures 3(a), (b) and (c) respectively. It can be seen that the maximum ratio is less than 0.06, which is much less than 8. Therefore, the second-order approximation of the distance is reasonable. The following analysis shows the change process of the approximate error with the target distance, azimuth and pitch angle. The results are shown in Figures 4(a), (b) and (c) respectively. It can be seen that the maximum approximate error is less than 3.5×10 -3 m, which is much smaller than the radar wavelength of 0.03 m. The approximation error has little effect on the phase, so the approximation method has little effect on azimuth imaging. In Simulation 4, we will further analyze the impact of various parameters on imaging quality.

[0081] Simulation 2: To verify the effectiveness of the proposed method, the following simulation experiment is conducted: 17 point targets are set at the positions shown in Table 2, with a scattering coefficient of 1. These 17 point targets are imaged using the proposed method.

[0082] Point target location Figure 5 As shown in Figure 6, the signal of the distance unit at 425m is extracted. There are three scattering points with different initial phases in this distance unit. Its time-frequency diagram is obtained, and the skeleton of the time-frequency diagram is extracted. The result is shown in Figure 6. According to the skeleton of the obtained time-frequency diagram, the maximum value of the Doppler frequency shift |H pmax | is 370 Hz, and substituting into (15) we can obtain: The difference between the estimated value and the true value (0.3rad) is extremely small, so it can be considered that the method of estimating the pitch angle based on the time-frequency diagram is effective. The imaging result in the range domain using the method proposed in the present invention is shown in Figure 7(a). The 17 point targets are distributed in 9 range units. Due to the effect of pulse accumulation, the range image has a prominent peak and can be accurately reconstructed. The imaging result in the azimuth domain is shown in Figure 7(b). The 17 point targets are distributed in 9 azimuth units. It can be seen from the figure that the reconstructed azimuth has a very small deviation from the ideal result and does not affect the azimuth imaging result of the target. The imaging result of the method proposed in the present invention at 0dB is shown in Figure 7(c). The distance and azimuth of the 17 point targets are accurately reconstructed.

[0083] Simulation 3: Imaging using the compressed sensing method will reduce the signal sidelobes, but will not affect the imaging resolution. In this section, in order to analyze the azimuth resolution of the method proposed in this invention, a point target model is set up, and the azimuth profile of the target under the traditional imaging method is obtained, and its 3dB bandwidth is obtained, which is the azimuth resolution. The resolution of the method proposed in this invention is compared with that of the electromagnetic vortex imaging method to prove the effectiveness of the method of this invention. The scattering point position is set to (400m, 0.3rad, 1.5rad) and the scattering coefficient is 1.

[0084] The azimuth profile under traditional imaging methods is as follows Figure 8 As shown in the figure, it can be seen that the azimuth angle resolution of the method of the present invention is 0.06rad, while the azimuth angle resolution of the electromagnetic vortex imaging method is Where Δα is the modal number range of the vortex electromagnetic wave. To achieve an azimuthal resolution of 0.06 rad, the required modal number range is Δα = 105. The modal number of pure vortex electromagnetic waves that can currently be generated is far less than 105, so the imaging method proposed in this invention has a good application advantage.

[0085] Simulation 4: In this section, the influence of various parameters on the imaging quality in the proposed method is analyzed. The peak signal-to-noise ratio (PSNR) is introduced as a criterion for evaluating imaging quality. The larger the PSNR value, the better the imaging quality, and vice versa. The definition of PSNR is as follows: Assume two m×n monochrome images I and K. If image I is a noise approximation of image K, then their mean square error (MSE) and PSNR can be defined as

[0086]

[0087]

[0088] As shown in Figure 9(a), the imaging quality changes with the target distance. It can be seen that when the target distance is less than 50m, the imaging quality changes dramatically. After the target distance is greater than 50m, the imaging quality tends to be stable, indicating that the imaging method proposed in the present invention can effectively image far-field targets. Figure 9(b) is a diagram showing the imaging quality changes with the azimuth angle. It can be seen that the change in the peak signal-to-noise ratio is only 0.0025dB, and the imaging quality is minimally affected by the change in azimuth angle. When the azimuth angle is 0 to -π, the imaging result is a mirror image result when the azimuth angle is 0 to π. Therefore, the imaging method proposed in the present invention can effectively image azimuth angles within the range of -π to π. Figure 9(c) is a diagram showing the imaging quality changes with the pitch angle. It can be seen that In the range of pitch angle, the peak signal-to-noise ratio varies randomly without a fixed trend. When the peak signal-to-noise ratio is the smallest, the corresponding pitch angle is 0.8655 rad. Figure 9(d) shows the imaging result of the imaging method proposed in the present invention when the pitch angle is 0.8655 rad. It can be seen that although there are some noise spots in the image, the outline of the target area can still be clearly distinguished. Therefore, the imaging method proposed in the present invention has a good effect on the pitch angle. All scattering points within the range can be effectively imaged. In summary, the imaging method proposed in the present invention can perform two-dimensional high-resolution imaging of targets at all azimuths and elevation angles under far-field conditions. The simulation results verify the effectiveness of this method.

[0089] Simulation 5: In this section, we analyze the robustness of our proposed method. Figure 10(a) shows how imaging quality changes with signal-to-noise ratio (SNR). It can be seen that as the SNR increases, the imaging quality gradually improves and stabilizes when the SNR exceeds -19 dB. Figure 10(b) shows the imaging results of our proposed method at a SNR of -24 dB. It can be seen that some noise spots appear at this time, but the outline of the target area can still be clearly distinguished. Figure 10(c) shows the imaging results of our proposed method at a SNR of -19 dB. It can be seen that SNRs above -19 dB have little effect on the imaging quality, and the two-dimensional image of the target is accurately reconstructed. The reason why this method can image at such low SNRs is that the noise is uncorrelated, while the signal emitted by the radar is correlated. When the received echo signal is Fourier transformed, only the signal related to the target position information is enhanced, which improves the noise immunity of this method. The simulation results verify the robustness of this method.

[0090] The proposed method for two-dimensional high-resolution imaging of radar targets using a rotating antenna constructs a new observation model based on a rotating antenna. This method features a simple design, low data processing complexity, and low radar performance requirements, thus reducing imaging costs. Using compressed sensing, it effectively suppresses azimuth sidelobes, improving azimuth resolution and enhancing imaging quality. Simulation results demonstrate that the proposed imaging method can effectively image targets at all elevation and azimuth angles while maintaining high imaging quality.

Claims

1. A method for two-dimensional high-resolution imaging of radar targets based on a rotating antenna, comprising the following steps: Step 1: Construct a rotating antenna observation model and approximate the distance between the radar and the target using the Fresnel method to obtain an approximate distance. Obtain the true echo signal from the scattering point and perform "Dechirp" processing on the echo signal. Then, perform a fast Fourier transform in the range direction to obtain a sinc-shaped narrow pulse. The range image is extracted using the peak of the sinc function. Step 2: Extract the echo signal within each range unit and substitute the approximate distance into the echo signal. On this basis, derive the expression of Doppler shift. Perform a short-time Fourier transform on the echo signal within the range unit to obtain the slow time-frequency image of the echo signal. Combined with the Doppler shift expression, estimate the target pitch angle. Use the target pitch angle to construct a sparse dictionary. Use the echo signal within the range unit as the observation value and use the OMP algorithm to reconstruct the target azimuth. At this point, a high-resolution image of the target's range-azimuth can be obtained.

2. The method for two-dimensional high-resolution imaging of radar targets based on a rotating antenna according to claim 1, characterized in that: The step 1 specifically includes the following steps: Step 1) Construct a rotating antenna observation model and perform Fresnel approximation on the distance between the radar and the target to obtain the approximate distance r p (t m ), r p (t m ) is determined by the distance from the target to the center of the circle, the radius of the circle, the target pitch angle and the azimuth angle, and the echo delay τ is calculated. p (t m ); Step 2) Echo signal s echo (t m ,t) and the reference signal s ref (t) conjugate multiplication to obtain a single-frequency pulse signal, where t represents the fast time and t m Indicates slow time; Step 3) Perform Fast Fourier Transform (FFT) on the single-frequency pulse signal to obtain the corresponding sinc-shaped narrow pulse in the frequency domain, and then compare it with the phase compensation function S de (f r ) and remove the residual video phase (RVP) and envelope tilt term to obtain the echo signal S in the range frequency domain echo (t m ,f r ), where f r represents the distance frequency; Step 4) r =-γτ p (t m ) to extract S echo (t m ,f r ) peak, and obtain the distance information of the scattering point, γ is the modulation frequency.

3. The method for two-dimensional high-resolution imaging of radar targets based on a rotating antenna according to claim 1, characterized in that: The step 2 specifically includes the following steps: Step 1) Extract the signal S of each distance unit unit (t m ), for S unit (t m ) is processed by short-time Fourier transform to obtain the time-frequency diagram of the signal, that is, the Doppler frequency shift H p (t m ) changes with slow time, and performs mathematical morphological image processing on the time-frequency graph to extract its skeleton features; Step 2) Substitute the approximate distance into the distance unit signal S unit (t m ), obtain the phase φ(t m ), for φ(t m ) Take the derivative to get the Doppler frequency shift expression H at the scattering point p (t m ), in this model, H p (t m ) is constant in amplitude and its phase changes with the slow time t m Changing sine function, H p (t m ) contains the pitch angle information of the target, so the pitch angle of the scattering point can be estimated by using the maximum value of the time-frequency diagram; Step 3) Build an azimuth dictionary The estimated pitch angle and azimuth angle dictionary is used to construct the sparse dictionary matrix D. Given the known observation matrix y = S unit (t m ) and the sparse dictionary matrix D, the OMP algorithm is used to reconstruct the azimuth The optimization objective of the OMP algorithm can be expressed as: in is the optimization objective, X is a column vector, and ||||2 represents the matrix norm; Step 4) The index position corresponding to the non-zero element in X is the azimuth In the azimuth dictionary The corresponding position in the image can be reconstructed, so the target azimuth can be reconstructed. Combined with the distance information obtained in step 1, a high-resolution image of the target's range-azimuth can be obtained.

Images