Step frequency imaging method based on speed matched filtering and false scattering point elimination

Abstract

The invention discloses a step frequency imaging method based on speed matched filtering and false scattering point elimination, and the method comprises the steps: firstly, obtaining a distance unit corresponding to the maximum amplitude in a non-coherent processing channel and a speed matched filtering result, and calculating a target compensation speed range; traversing each speed value in the compensation speed range, and generating a target one-dimensional range profile corresponding to different speeds according to the sub-pulse data; taking the ratio of the kurtosis of the range profile to the envelope entropy as a cost function, and taking the range profile corresponding to the maximum cost function as a range profile after speed compensation; and finally, calculating a scattering point correlation coefficient by using the distance image after speed compensation and the scattering point sample, performing false scattering point judgment according to the scattering point correlation coefficient and an amplitude ratio, and obtaining a high-resolution one-dimensional distance image of the target through false scattering point elimination. According to the invention, accurate estimation of the target speed is realized, the imaging accuracy of the step frequency system is improved, and the measurement and calculation precision of detection target parameters is improved.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a step-frequency imaging method based on velocity-matched filtering and false scattering point removal. Background Technology

[0002] The target size, shape, and scattering point distribution information acquired by radar are inherently related to the threat level of the target. High-resolution range-dimensional imaging of targets using radar is one of the important means of obtaining detailed target information. In radar transmission signals, broadband signals are the primary signal form for obtaining high-resolution range-dimensional imaging. Simultaneously, detection range is also a crucial indicator of a radar system; the greater the detection range, the greater the required transmission energy. This limits the pulse width to a certain level of transmitter peak power. Reconciling the contradiction between bandwidth and pulse width, and selecting a suitable product signal with a large time-bandwidth ratio, is the key issue enabling radar to simultaneously achieve high resolution and long detection range.

[0003] Linear frequency modulated (LFM) signals are typical examples of large bandwidth-time product signals. These signals possess both a large frequency bandwidth and a good pulse duration, enabling high distance resolution through pulse compression techniques. However, in practical engineering implementations, the frequency bandwidth of a signal cannot be infinitely expanded. Therefore, utilizing stepped-frequency signals and employing pulse synthesis algorithms to achieve distance resolution equivalent to that of large-bandwidth signals is an important engineering implementation method.

[0004] A stepped-frequency pulse signal transmits a pulse group in which the carrier frequency of each pulse changes linearly at a fixed frequency. Stepped-frequency pulse signals are Doppler-sensitive signals. When the radial velocity of the target is not zero, the target's stepped-frequency synthesized range image will be shifted or distorted. Therefore, velocity compensation of the target is required before stepped-frequency imaging. Deviations in the velocity estimation of the target will lead to distortion of the target's one-dimensional range image. Furthermore, the imaging results of stepped-frequency volumetric radar are affected by parameter selection, resulting in redundancy and the appearance of false scattering points on the range image, severely impacting the estimation of parameters such as the number and size of target scattering points. Summary of the Invention

[0005] The purpose of this invention is to provide a step-frequency imaging method based on velocity matched filtering and false scattering point removal, thereby improving the accuracy of target velocity estimation, avoiding the distance image distortion phenomenon caused by velocity estimation deviation in traditional step-frequency imaging algorithms, improving the accuracy of step-frequency system imaging, and improving the measurement accuracy of parameters such as target size.

[0006] The technical solution to achieve the purpose of this invention is: a step-frequency imaging method based on velocity matched filtering and spurious scattering point removal, comprising the following steps:

[0007] Step 1: Calculate the upper limit of subpulse velocity compensation based on the parameters of the subpulse of the radar's transmitted step-frequency signal;

[0008] Step 2: Divide the velocity estimation matching channel and calculate the velocity estimation matching coefficient based on the upper limit of the absolute value of the sub-pulse velocity compensation;

[0009] Step 3: Using the data after pulse compression of the sub-pulse, search for the distance cell corresponding to the maximum amplitude through modulus calculation, non-coherent accumulation and constant false alarm rate detection.

[0010] Step 4: Multiply the velocity estimation matching coefficient with the data of the sub-pulse after pulse compression to obtain the velocity matching filter result. Calculate the target compensation velocity estimation range based on the velocity matching filter result and the distance cell corresponding to the maximum amplitude.

[0011] Step 5: Traverse each velocity value within the target compensation velocity range, and perform velocity compensation, IFFT, modulus calculation, range image extraction and stitching on the data after pulse compression of the sub-pulse to obtain high-resolution one-dimensional range images corresponding to different compensation velocities.

[0012] Step 6: Calculate the kurtosis and envelope entropy of the high-resolution one-dimensional range image corresponding to each compensation velocity. Using kurtosis and envelope entropy as cost functions, search for the maximum value of the objective function within the compensation velocity range. Obtain the high-resolution one-dimensional range image after target velocity compensation based on the maximum value index.

[0013] Step 7: Use the high-resolution one-dimensional range image after target velocity compensation and the false scattering point samples to detect and judge false scattering points, remove false scattering points, and obtain the final high-resolution one-dimensional range image.

[0014] A computer device includes a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the step-frequency imaging method based on velocity matched filtering and false scattering point removal.

[0015] Compared with the prior art, the present invention has the following significant advantages: (1) It uses the parameter information of the radar step-frequency imaging system to calculate the upper limit of velocity compensation and velocity matching coefficient, calculates the target compensation velocity range based on the velocity matching filtering result, and uses the ratio of range image kurtosis and envelope entropy as the cost function to achieve accurate estimation of target velocity within the compensation velocity range, thus avoiding the range image distortion phenomenon caused by velocity estimation deviation in the traditional step-frequency imaging algorithm; (2) It uses the velocity-compensated range image and scattering point samples to calculate the scattering point correlation coefficient, and makes a false scattering point judgment based on the scattering point correlation coefficient and amplitude ratio threshold, eliminating false scattering points caused by parameter selection, increasing the accuracy of step-frequency system imaging, and improving the accuracy of the measurement of parameters such as the size of the detected target. Attached Figure Description

[0016] Figure 1 This is a flowchart of the step-frequency imaging method based on velocity matched filtering and false scattering point removal of the present invention.

[0017] Figure 2 This is a graph showing the detection results of the non-coherent channel under simulation conditions in this embodiment of the invention, as well as the velocity filtering results of each matched channel in the distance unit where the target is located.

[0018] Figure 3 This is a graph showing the peak entropy ratio and the high-resolution one-dimensional range image result after target velocity compensation for different compensation velocities in this embodiment of the invention.

[0019] Figure 4 This is a curve showing the correlation coefficient results of scattering point samples and scattering point samples from the simulation data in this embodiment of the invention.

[0020] Figure 5 This is a comparison curve of the imaging results of the method of the present invention and the traditional step-frequency imaging algorithm on the simulation data in the embodiments of the present invention.

[0021] Figure 6 This is a curve showing the correlation coefficient between the scattering point samples and the scattering point samples of the measured data in this embodiment of the invention.

[0022] Figure 7 This is a comparison curve of the imaging results of the method of the present invention and the traditional step-frequency imaging algorithm on the measured data in the embodiments of the present invention. Detailed Implementation

[0023] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0024] This invention proposes a step-frequency imaging method based on velocity-matched filtering and false scattering point removal, mainly addressing the range image distortion problem caused by velocity compensation errors and false scattering points in step-frequency radar. The method includes: calculating the velocity compensation upper limit and velocity matching coefficient based on radar detection parameters; calculating the target compensation velocity range based on the range cell corresponding to the maximum amplitude obtained from the non-coherent processing channel and the velocity matching filtering result; traversing each velocity value within the compensation velocity range; generating one-dimensional range images of the target corresponding to different velocities based on sub-pulse data; using the ratio of kurtosis to envelope entropy of the range image as a cost function; selecting the range image corresponding to the maximum cost function as the velocity-compensated range image; calculating the scattering point correlation coefficient using the velocity-compensated range image and scattering point samples; determining false scattering points based on the scattering point correlation coefficient and amplitude ratio; and obtaining a high-resolution one-dimensional range image of the target after false scattering point removal.

[0025] Combination Figure 1 This invention discloses a step-frequency imaging method based on velocity matched filtering and spurious scattering point removal, comprising the following steps:

[0026] Step 1: Calculate the upper limit of subpulse velocity compensation based on the parameters of the subpulse of the radar's transmitted step-frequency signal;

[0027] Step 2: Divide the velocity estimation matching channel and calculate the velocity estimation matching coefficient based on the upper limit of the absolute value of the sub-pulse velocity compensation;

[0028] Step 3: Using the data after pulse compression of the sub-pulse, search for the distance cell corresponding to the maximum amplitude through modulus calculation, non-coherent accumulation and constant false alarm rate detection.

[0029] Step 4: Multiply the velocity estimation matching coefficient with the data of the sub-pulse after pulse compression to obtain the velocity matching filter result. Calculate the target compensation velocity estimation range based on the velocity matching filter result and the distance cell corresponding to the maximum amplitude.

[0030] Step 5: Traverse each velocity value within the target compensation velocity range, and perform velocity compensation, IFFT, modulus calculation, range image extraction and stitching on the data after pulse compression of the sub-pulse to obtain high-resolution one-dimensional range images corresponding to different compensation velocities.

[0031] Step 6: Calculate the kurtosis and envelope entropy of the high-resolution one-dimensional range image corresponding to each compensation velocity. Using kurtosis and envelope entropy as cost functions, search for the maximum value of the objective function within the compensation velocity range. Obtain the high-resolution one-dimensional range image after target velocity compensation based on the maximum value index.

[0032] Step 7: Use the high-resolution one-dimensional range image after target velocity compensation and the false scattering point samples to detect and judge false scattering points, remove false scattering points, and obtain the final high-resolution one-dimensional range image.

[0033] As a specific example, step 1 is as follows:

[0034] Step 1.1: Obtain the repetition period T(1)~T(N) and transmission frequency f(1)~f(N) of the N sub-pulses of the radar transmitted step frequency signal;

[0035] Step 1.2: Calculate the average repetition period T avr and mean repetition frequency f avr The calculation formula is:

[0036]

[0037]

[0038] Step 1.3: Calculate the upper limit of sub-pulse velocity compensation v cmax The calculation formula is:

[0039]

[0040] Where c = 3 * 10 8 For the speed of light, v cmax The unit is meters per second.

[0041] As a specific example, step 2 is as follows:

[0042] The number of speed estimation matching channels is divided into N, based on the upper limit of the absolute value v of the sub-pulse speed compensation. cmax The speed calculation estimates the matching coefficient W = [w1; w2; ... w N The speed estimation matching coefficient w of the p-th matching channel. p The calculation formula is:

[0043]

[0044] w p The dimension of is 1×N, and the dimension of W is N×N, where the formula for calculating fd(p,k) is:

[0045]

[0046] Where f(k) is the transmission frequency of the k-th sub-pulse in the step-frequency signal, and c = 3 * 10 8 Let p be the speed of light, 1 ≤ p ≤ N, 1 ≤ k ≤ N-1, and the formula for calculating v(p) is:

[0047] .

[0048] As a specific example, step 3 is as follows:

[0049] Using the M range cell I / Q data obtained after pulse compression of N sub-pulses, the range cell t corresponding to the maximum amplitude is searched through modulus calculation, non-coherent accumulation, and constant false alarm rate detection.

[0050] As a specific example, step 4 is as follows:

[0051] Step 4.1: Using the M range cell I / Q data obtained after pulse compression processing of the N sub-pulses, form a matrix S according to the pulse-range cell, i.e., S = [s1; s2; ... s... N ], where s n Let S represent the I / Q data of the nth sub-pulse, where 1 ≤ n ≤ N, and the dimension of matrix S is N × M.

[0052] Step 4.2: Multiply the speed estimation matching coefficient W with the matrix S, i.e., H = W * S, to obtain the speed matching filter matrix H, where the dimension of matrix H is N × M.

[0053] Step 4.3: Calculate the velocity estimation matching channel Chn corresponding to the maximum amplitude of the t-th distance unit in the velocity matched filter matrix H. The calculation formula is as follows:

[0054]

[0055] Where [B, C] = max[A] means performing a maximum value search on the elements in vector A, storing the maximum value in B, and storing the index of the maximum value in C;

[0056] Step 4.4: Based on the speed estimation matching channel Chn, calculate the speed estimation matching channel range ChnRng. The calculation formula is as follows:

[0057]

[0058] Step 4.5: Calculate the lower limit of the target compensation velocity estimation range based on ChnRng. and upper limit The calculation formula is:

[0059]

[0060]

[0061] Where [B]=minval(A) means performing a minimum value search on the elements in set A, and storing the minimum value found in B; [B]=maxval(A) means performing a maximum value search on the elements in set A, and storing the maximum value found in B. The symbol represents the floor function; The symbol represents the floor function;

[0062] Step 4.6: Calculate the target compensation speed estimation range The calculation formula is:

[0063]

[0064] Among them, v cRng The unit is meters per second; v cStep For speed compensation step, a value of 0.1 meters per second is used; v cRng The dimension is 1×J, that is, v cRng =[v1;v2; ... v J ].

[0065] As a specific example, step 5 is as follows:

[0066] Step 5.1: Traverse the target compensation speed range v cRng For each velocity value within the sub-pulse, velocity compensation is performed on the data S after pulse compression processing to obtain SC, i.e., SC = [sc1; sc2; ... sc J ], sc j This represents the result after applying speed compensation to S using the j-th speed, i.e., sc. j =[sc j1 ;sc j2 ; ··· sc jN ], sc jn This represents the velocity compensation result for the j-th velocity and the n-th sub-pulse. Since each sub-pulse has M distance units, sc jn The dimension is 1×M, sc jn The calculation formula is as follows:

[0067]

[0068] Where c = 3 * 10 8 Let B be the speed of light, exp(·) represent the natural exponential function, and B step This represents the step bandwidth; i represents the imaginary unit, i.e. ;

[0069] Step 5.2: SC is processed by IFFT and modulo to obtain SF, i.e., SF = [sf1; sf2; ... sf... J], sf j sf represents the result of taking the modulus after performing an N-point IFFT on the j-th compensation velocity. j The dimension is N×M, sf j The calculation formula is:

[0070]

[0071] Here, IFFT(A,B) represents the inverse Fourier transform function with B points applied to data A. Represents the modulus function for complex numbers;

[0072] Step 5.3: Traverse M range cells, extract and stitch the SF to obtain J high-resolution one-dimensional range images HP corresponding to compensated velocities, i.e., HP = [hp1; hp2; ... hp... J ], hp j hp represents the high-resolution one-dimensional range image data corresponding to the j-th compensated velocity. j The dimension is 1×MC, where MC represents the distance dimension of the high-resolution one-dimensional range image HP. The formula for calculating MC is:

[0073]

[0074] Among them, B step For step bandwidth, B sub For sub-pulse bandwidth, The symbol represents the floor function, and the dimension of HP is J×MC.

[0075] As a specific example, the high-resolution one-dimensional range image data hp corresponding to the j-th compensated velocity mentioned in step 5.3 j The calculation process is as follows:

[0076] Step 5.3.1: Calculate the center distance R of the imaging region. cen The formula for calculating the fine resolution distance element ΔR is as follows:

[0077]

[0078] Among them, R bgn This represents the position of the first distance unit in the imaging area, in meters, c = 3 * 10. 8 For the speed of light, f s B represents the distance sampling rate. step For step bandwidth;

[0079] Step 5.3.2: Calculate the starting distance index IND of the high-resolution one-dimensional range image HP. bgn ,Right now

[0080] IND bgn=[ind bgn (1);ind bgn (2); ··· ind bgn (M)],ind bgn The formula for calculating (m) is as follows:

[0081]

[0082] Step 5.3.3: Calculate the end distance index IND of the high-resolution one-dimensional range image HP. end ,Right now:

[0083] IND end =[ind end (1);ind end (2); ··· ind end (M)],ind end The formula for calculating (m) is as follows:

[0084]

[0085] Among them, B sub Where is the sub-pulse bandwidth, and Coe is the index calculation constant, with a value of 0.25. The symbol represents the floor function;

[0086] Step 5.3.4, according to IND bgn Calculate the starting distance index P of the data SF, i.e., P = [p(1); p(2); ... p(M)], where p(i) is calculated using the following formula:

[0087]

[0088] Step 5.3.5, according to IND end Calculate the end distance index Q of the data SF, i.e., Q = [q(1), q(2), ... q(M)], where q(i) is calculated using the following formula:

[0089]

[0090] Where mod(A,B) represents the modulo function of A with respect to B;

[0091] Step 5.3.6, according to sf j P, Q, IND bgn and IND end Calculate HP j The calculation formula is:

[0092] .

[0093] As a specific example, step 6 is as follows:

[0094] Step 6.1: Calculate the high-resolution one-dimensional range image hp corresponding to each compensated velocity. j The kurtosis K(j) and envelope entropy E(j), j=1,2,…,J;

[0095] The formula for calculating kurtosis K(j) is:

[0096]

[0097] in ;

[0098] The formula for calculating the envelope entropy E(j) is:

[0099]

[0100] Among them, hp j (i) represents the high-resolution one-dimensional range image hp corresponding to the j-th compensated velocity. j The data in the i-th distance cell;

[0101] Step 6.2: Denote the ratio of kurtosis K to envelope entropy E as the peak-entropy ratio, and use the peak-entropy ratio as the objective function. Within the compensation speed range v cRng The maximum value of the objective function f(K,E) is found in the search function, and the formula is as follows:

[0102]

[0103] Let the index corresponding to the maximum value be denoted as IMAX, and let HP be denoted as MAX. imax HPO is a high-resolution one-dimensional range image after target velocity compensation.

[0104] As a specific example, step 7 is as follows:

[0105] Step 7.1: Write the distance cell indices corresponding to values ​​greater than the threshold Thr in the HPO data into set V, where the threshold Thr = 0.125 * HPO. max HPO max To find the maximum value of HPO, using the set V and the high-resolution one-dimensional range image HPO, sample data SP for spurious scattering point detection is selected from HPO, using the formula:

[0106]

[0107] The length of the sample data SP is M sp ;

[0108] Step 7.2: Using the spurious scattering point sample data SP and the one-dimensional range image data HPO, calculate the spurious scattering point correlation coefficient CR. The dimension of CR is 1×M. cr , that is, CR=[cr(1);cr(2);···cr(M cr )], where M cr =MC-M sp +1, the formula for calculating cr(u) is as follows:

[0109]

[0110] in, Represents the covariance function. Represents the standard deviation function;

[0111] Step 7.3: Place the u values ​​corresponding to cr(u) greater than 0.7 in the correlation coefficient CR of spurious scattering points into set R, M. R The number of elements in set R;

[0112] Step 7.4: Add the elements in set R that satisfy the relative amplitude ratio of the spurious scattering points to set D, with the following condition:

[0113]

[0114] Step 7.5: Range HPO data Points within the range are identified as false scattering points and are removed, i.e.:

[0115]

[0116] Where R ii Let i be the i-th element in set R. D jj Let jj be the j-th element in set D. M D The number of elements in set D;

[0117] Step 7.6: The high-resolution one-dimensional range image HPO obtained after removing false scattering points is the final output high-resolution one-dimensional range image HPO.

[0118] The present invention also provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the step-frequency imaging method based on velocity matched filtering and false scattering point removal.

[0119] Example

[0120] To verify the effectiveness of the step-frequency imaging method based on velocity matched filtering and false scattering point removal provided by this invention, this embodiment conducted simulation experiments and measured data analysis. The simulation experiment examples and measured data analysis are further explained below.

[0121] Simulation Experiment Example: Simulation parameters are as follows: The simulated target moves radially relative to the radar, with a target velocity of 100 m / s, three scattering points, a target size of 2 meters, a radar floor noise mean of 42 dB, and an echo duration of 20 μs and a signal bandwidth of B. sub The signal is a 40MHz zero-IF linear frequency modulated signal with a step frequency range of [3:B]. step [3.46], unit is GHz, B step Equal to 0.02GHz, the number of step-frequency sub-pulses N is 24, the signal-to-noise ratio of a single pulse target is 30dB, and the range sampling rate f s With a frequency of 50MHz, a pulse repetition period of 100μs, the target composite bandwidth is 20*24=480MHz, and the target fine resolution element is approximately 150 / 480≈0.31 meters.

[0122] Calculate the sub-pulse velocity compensation upper limit v according to the formula in step 1. cmax Equals 464.4 m / s;

[0123] Following step 2, the number of speed estimation matching channels is divided into 24, and the speed estimation matching coefficient W is calculated.

[0124] Following step 3, each sub-pulse undergoes pulse compression, noncoherent accumulation, and constant false alarm rate (CFAR) detection. The results of the noncoherent accumulation and CFAR detection processing of the 24 sub-pulse echo signals are as follows: Figure 2 As shown in (a) in the figure, by Figure 2 As shown in (a), the distance unit t corresponding to the maximum amplitude is 100;

[0125] Following step 4, velocity-matched filtering is performed to obtain the filtering results H for 24 velocity-matched channels. According to the formula... The velocity estimation matching channel Chn corresponding to the maximum amplitude of the t-th distance unit in the velocity matched filtering result H is equal to 6, such as... Figure 2 As shown in (b), based on Chn equaling 6, the range of the matching channel ChnRng is estimated to be [5, 7] based on the calculation speed. Then, based on ChnRng and v cmax The lower limit of the target compensation speed estimation range is obtained. and upper limit 96 m / s and 136 m / s respectively, v cStep The value is taken as 0.1 m / s, and then according to the formula... We can obtain J = 401;

[0126] Follow step 5 to traverse the target compensation speed range v cRng For each velocity value within the sub-pulse, velocity compensation, IFFT, modulus calculation, range image extraction and stitching are performed on the data after pulse compression processing to generate 401 high-resolution one-dimensional range images (HP).

[0127] Following step 6, calculate the kurtosis and envelope entropy of HP corresponding to the 401 velocities. Use the ratio of kurtosis to envelope entropy, i.e., the peak-entropy ratio, as the objective function within the compensation velocity range v. cRng The search for the maximum value of the objective function yields the peak entropy ratio results for different compensation speeds, as shown below. Figure 3 As shown in (a) in the figure, by Figure 3 From (a) in the diagram, we know that the velocity that maximizes the peak entropy ratio to the objective function is 100 m / s, within the range v. cRng The index IMAX corresponding to a medium speed value of 100 m / s is 41. HP 41 As the high-resolution one-dimensional range profile HPO after target velocity compensation, the imaging window size of HPO was set to 37 range units for easy data observation. The high-resolution normalized one-dimensional range profile HPO after target velocity compensation is shown below. Figure 3 As shown in (b);

[0128] According to the formula in step 7

[0129]

[0130] Based on the threshold Thr, the sample data S=HPO[14:22] for detecting false scattering points is selected, and the length M of the sample data SP is... sp The value is 9. Using the false scattering point sample data S and the one-dimensional range image data HPO, the false scattering point correlation coefficient r is calculated according to the formula. The results of the false scattering point detection sample S and the false scattering point correlation coefficient CR are as follows: Figure 4 (a) and Figure 4 As shown in (b), the correlation coefficient of false scattering points of the data in the 4th and 24th distance units in the imaging window is greater than 0.7, and the amplitude ratio of the maximum value of the false scattering point sample to the maximum value of the detection area, i.e., HPO[4:12], is 20*log10(1 / 0.09783) approximately equal to 20.2dB, which satisfies the judgment condition that the amplitude ratio of false scattering points is greater than 12dB. Therefore, distance unit index 4 and distance unit index 24 are put into set D, resulting in D={4, 24}. Then, according to the formula...

[0131]

[0132] The spurious scattering point data is zeroed out to obtain the final high-resolution one-dimensional range profile (HPO). The HPO before and after spurious scattering point removal are shown below. Figure 5 (a) and Figure 5 As shown in (b) of the diagram. Figure 5 It can be seen that before the false scattering point removal, the target size is approximately (31-5)*0.31≈8 meters, which deviates significantly from the true target size of 2 meters. After processing by the method of this invention, the false scattering points of the target's high-resolution one-dimensional range image HPO are suppressed, i.e. Figure 5 The data corresponding to the two rectangular boxes in (a) show that the target size is approximately (21-15)*0.31≈1.9 meters, which is close to the true target value of 2 meters. This indicates that the method provided by the present invention is beneficial for obtaining accurate target size and other information. Simulation experiments show that the method provided by the present invention can correctly measure the radial size of the target.

[0133] Example of measured data: To verify the engineering applicability of the method provided by this invention, stepped-frequency imaging analysis was performed on measured data of a radar using the method provided by this invention. The target moves radially relative to the radar, the true value of the target's radial length is 14 meters, and the length of the range cell after imaging is 0.5 meters. The target's false scattering point samples and the correlation coefficient of the false scattering points are as follows. Figure 6 As shown, by Figure 6 It can be seen that the correlation coefficient of the spurious scattering point in the 14th range cell within the imaging window exceeds the threshold of 0.7. Further judgment is then made based on the relative amplitude ratio of the spurious scattering points. Figure 7 From (a) in the diagram, we know that the amplitude ratio between the 50th range cell and the 16th range cell is approximately 20*log10(1 / 0.172)≈15.3dB, which satisfies the criterion that the relative amplitude ratio of false scattering points is greater than 12dB. Therefore, for... Figure 7 The imaging results of the rectangular frame portion were identified as spurious scattering points, and their corresponding data were zeroed out. The one-dimensional high-resolution range image after removing spurious scattering points is as follows: Figure 7 As shown in (b) above, by Figure 7 It can be seen that before the false scattering points are removed, the radial length of the target is (66-14)*0.5 equals 26 meters. The measured radial length deviates significantly from the true length, with a measurement error of (26.5-14) / 14=89.3%. After the false scattering points are removed using the method provided in this embodiment of the invention, the radial length of the target is (66-39)*0.5 equals 13.5 meters, which is close to the true radial length of the target. The measurement error is (13.5-14) / 14=3.6%. The measured data shows that the method provided in this invention can accurately measure the radial dimension of the target.

[0134] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A step-frequency imaging method based on velocity-matched filtering and spurious scattering point removal, characterized in that, Includes the following steps: Step 1: Calculate the upper limit of subpulse velocity compensation based on the parameters of the subpulse of the radar's transmitted step-frequency signal; Step 2: Divide the velocity estimation matching channel and calculate the velocity estimation matching coefficient based on the upper limit of the absolute value of the sub-pulse velocity compensation; Step 3: Using the data after pulse compression of the sub-pulse, search for the distance cell corresponding to the maximum amplitude through modulus calculation, non-coherent accumulation and constant false alarm rate detection. Step 4: Multiply the velocity estimation matching coefficient with the data of the sub-pulse after pulse compression to obtain the velocity matching filter result. Calculate the target compensation velocity estimation range based on the velocity matching filter result and the distance cell corresponding to the maximum amplitude. Step 5: Traverse each velocity value within the target compensation velocity range, and perform velocity compensation, IFFT, modulus calculation, range image extraction and stitching on the data after pulse compression of the sub-pulse to obtain high-resolution one-dimensional range images corresponding to different compensation velocities. Step 6: Calculate the kurtosis and envelope entropy of the high-resolution one-dimensional range image corresponding to each compensation velocity. Using kurtosis and envelope entropy as cost functions, search for the maximum value of the objective function within the compensation velocity range. Obtain the high-resolution one-dimensional range image after target velocity compensation based on the maximum value index. Step 7: Use the high-resolution one-dimensional range image after target velocity compensation and the false scattering point samples to detect and judge false scattering points, remove false scattering points, and obtain the final high-resolution one-dimensional range image.

2. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 1, characterized in that, Step 1 is as follows: Step 1.1: Obtain the repetition period T(1)~T(N) and transmission frequency f(1)~f(N) of the N sub-pulses of the radar transmitted step frequency signal; Step 1.2: Calculate the average repetition period T avr and mean repetition frequency f avr The calculation formula is: ； ； Step 1.3: Calculate the upper limit of sub-pulse velocity compensation v cmax The calculation formula is: ； Where c = 3 * 10 8 For the speed of light, v cmax The unit is meters per second.

3. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 2, characterized in that, Step 2 is as follows: The number of speed estimation matching channels is divided into N, based on the upper limit of the absolute value v of the sub-pulse speed compensation. cmax The speed calculation estimates the matching coefficient W = [w1; w2; ... w N The speed estimation matching coefficient w of the p-th matching channel. p The calculation formula is: ； w p The dimension of is 1×N, and the dimension of W is N×N, where the formula for calculating fd(p,k) is: ； Where f(k) is the transmission frequency of the k-th sub-pulse in the step-frequency signal, and c = 3 * 10 8 Let p be the speed of light, 1 ≤ p ≤ N, 1 ≤ k ≤ N-1, and the formula for calculating v(p) is: 。 4. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 3, characterized in that, Step 3 is as follows: Using the M range cell I / Q data obtained after pulse compression of N sub-pulses, the range cell t corresponding to the maximum amplitude is searched through modulus calculation, non-coherent accumulation, and constant false alarm rate detection.

5. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 4, characterized in that, Step 4 is as follows: Step 4.1: Using the M range cell I / Q data obtained after pulse compression processing of the N sub-pulses, form a matrix S according to the pulse-range cell, i.e., S = [s1; s2; ... s... N ], where s n Let S represent the I / Q data of the nth sub-pulse, where 1 ≤ n ≤ N, and the dimension of matrix S is N × M. Step 4.2: Multiply the speed estimation matching coefficient W with the matrix S, i.e., H = W * S, to obtain the speed matching filter matrix H, where the dimension of matrix H is N × M. Step 4.3: Calculate the velocity estimation matching channel Chn corresponding to the maximum amplitude of the t-th distance unit in the velocity matched filter matrix H. The calculation formula is as follows: ； Where [B, C] = max[A] means performing a maximum value search on the elements in vector A, storing the maximum value in B, and storing the index of the maximum value in C; Step 4.4: Based on the speed estimation matching channel Chn, calculate the speed estimation matching channel range ChnRng. The calculation formula is as follows: ； Step 4.5: Calculate the lower limit of the target compensation velocity estimation range based on ChnRng. and upper limit The calculation formula is: ； ； Where [B]=minval(A) means performing a minimum value search on the elements in set A, and storing the minimum value found in B; [B]=maxval(A) means performing a maximum value search on the elements in set A, and storing the maximum value found in B. The symbol represents the floor function; The symbol represents the floor function; Step 4.6: Calculate the target compensation speed estimation range The calculation formula is: ； Among them, v cRng The unit is meters per second; v cStep For speed compensation step, a value of 0.1 meters per second is used; v cRng The dimension is 1×J, that is, v cRng =[v1;v2; ... v J ].

6. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 5, characterized in that, Step 5 is as follows: Step 5.1: Traverse the target compensation speed range v cRng For each velocity value within the sub-pulse, velocity compensation is performed on the data S after pulse compression processing to obtain SC, i.e., SC = [sc1; sc2; ... sc J ], sc j This represents the result after applying speed compensation to S using the j-th speed, i.e., sc. j =[sc j1 ;sc j2 ; ··· sc jN ], sc jn This represents the velocity compensation result for the j-th velocity and the n-th sub-pulse. Since each sub-pulse has M distance units, sc jn The dimension is 1×M, sc jn The calculation formula is as follows: ； Where c = 3 * 10 8 Let B be the speed of light, exp(·) represent the natural exponential function, and B step This represents the step bandwidth; i represents the imaginary unit, i.e. ; Step 5.2: SC is processed by IFFT and modulo to obtain SF, i.e., SF = [sf1; sf2; ... sf... J ], sf j sf represents the result of taking the modulus after performing an N-point IFFT on the j-th compensation velocity. j The dimension is N×M, sf j The calculation formula is: ； Here, IFFT(A,B) represents the inverse Fourier transform function with B points applied to data A. Represents the modulus function for complex numbers; Step 5.3: Traverse M range cells, extract and stitch the SF to obtain J high-resolution one-dimensional range images HP corresponding to compensated velocities, i.e., HP = [hp1; hp2; ... hp... J ], hp j hp represents the high-resolution one-dimensional range image data corresponding to the j-th compensated velocity. j The dimension is 1×MC, where MC represents the distance dimension of the high-resolution one-dimensional range image HP. The formula for calculating MC is: ； Among them, B step For step bandwidth, B sub For sub-pulse bandwidth, The symbol represents the floor function, and the dimension of HP is J×MC.

7. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 6, characterized in that, The high-resolution one-dimensional range image data hp corresponding to the j-th compensated velocity mentioned in step 5.3 j The calculation process is as follows: Step 5.3.1: Calculate the center distance R of the imaging region. cen The formula for calculating the fine resolution distance element ΔR is as follows: ； Among them, R bgn This represents the position of the first distance unit in the imaging area, in meters, c = 3 * 10. 8 For the speed of light, f s B represents the distance sampling rate. step For step bandwidth; Step 5.3.2: Calculate the starting distance index IND of the high-resolution one-dimensional range image HP. bgn ,Right now IND bgn =[ind bgn (1);ind bgn (2); ··· ind bgn (M)],ind bgn The formula for calculating (m) is as follows: ； Step 5.3.3: Calculate the end distance index IND of the high-resolution one-dimensional range image HP. end ,Right now: IND end =[ind end (1);ind end (2); ··· ind end (M)],ind end The formula for calculating (m) is as follows: ； Among them, B sub Where is the sub-pulse bandwidth, and Coe is the index calculation constant, with a value of 0.

25. The symbol represents the floor function; Step 5.3.4, according to IND bgn Calculate the starting distance index P of the data SF, i.e., P = [p(1); p(2); ... p(M)], where p(i) is calculated using the following formula: ； Step 5.3.5, according to IND end Calculate the end distance index Q of the data SF, i.e., Q = [q(1), q(2), ... q(M)], where q(i) is calculated using the following formula: ； Where mod(A,B) represents the modulo function of A with respect to B; Step 5.3.6, according to sf j P, Q, IND bgn and IND end Calculate HP j The calculation formula is: 。 8. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 7, characterized in that, Step 6 is as follows: Step 6.1: Calculate the high-resolution one-dimensional range image hp corresponding to each compensated velocity. j The kurtosis K(j) and envelope entropy E(j), j=1,2,…,J; The formula for calculating kurtosis K(j) is: ； in ; The formula for calculating the envelope entropy E(j) is: ； Among them, hp j (i) represents the high-resolution one-dimensional range image hp corresponding to the j-th compensated velocity. j The data in the i-th distance cell; Step 6.2: Denote the ratio of kurtosis K to envelope entropy E as the peak-entropy ratio, and use the peak-entropy ratio as the objective function. Within the compensation speed range v cRng The maximum value of the objective function f(K,E) is found in the search function, and the formula is as follows: ； Let the index corresponding to the maximum value be denoted as IMAX, and let HP be denoted as MAX. imax HPO is a high-resolution one-dimensional range image after target velocity compensation.

9. The step-frequency imaging method based on velocity-matched filtering and false scattering point removal according to claim 8, characterized in that, Step 7 is as follows: Step 7.1: Write the distance cell indices corresponding to values ​​greater than the threshold Thr in the HPO data into set V, where the threshold Thr = 0.125 * HPO. max HPO max To find the maximum value of HPO, using the set V and the high-resolution one-dimensional range image HPO, sample data SP for spurious scattering point detection is selected from HPO, using the formula: ； The length of the sample data SP is M sp ; Step 7.2: Using the spurious scattering point sample data SP and the one-dimensional range image data HPO, calculate the spurious scattering point correlation coefficient CR. The dimension of CR is 1×M. cr , that is, CR=[cr(1);cr(2);···cr(M cr )], where M cr =MC-M sp +1, the formula for calculating cr(u) is as follows: ； in, Represents the covariance function. Represents the standard deviation function; Step 7.3: Place the u values ​​corresponding to cr(u) greater than 0.7 in the correlation coefficient CR of spurious scattering points into set R, M. R The number of elements in set R; Step 7.4: Add the elements in set R that satisfy the relative amplitude ratio of the spurious scattering points to set D, with the following condition: ； Step 7.5: Range HPO data Points within the range are identified as false scattering points and are removed, i.e.: ； Where R ii Let i be the i-th element in set R. D jj Let jj be the j-th element in set D. M D The number of elements in set D; Step 7.6: The high-resolution one-dimensional range image HPO obtained after removing false scattering points is the final output high-resolution one-dimensional range image HPO.

10. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the step-frequency imaging method based on velocity-matched filtering and false scattering point removal as described in any one of claims 1 to 9.

Images