Manifold filtering clutter suppression method, device and equipment based on geodetic coordinates

Abstract

The present application relates to a method, device and equipment for suppressing clutter by manifold filtering based on geodesic coordinates. The method comprises: obtaining a radar echo signal to be processed and performing matched filtering processing to obtain an echo signal matrix and divide it into a plurality of range pulse units; using a Hamming window to respectively construct a covariance matrix corresponding to each range pulse unit; respectively calculating the geodesic coordinates of each covariance matrix, vectorizing each geodesic coordinate, and then reducing the dimension into a corresponding three-dimensional vector using a PCA algorithm, visualizing it in a three-dimensional space, and determining a clutter suppression area in the three-dimensional space according to the visualization result; using a manifold filter to perform matrix transformation processing on all covariance matrices to obtain a target covariance matrix after clutter suppression; using the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result. The purpose of greatly improving the target detection performance under complex backgrounds is achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of radar target detection, and relates to a manifold filtering clutter suppression method, device and equipment based on geodetic coordinates. Background Art

[0002] Radar target detection is the process of processing radar signals to determine whether a target echo exists. Radar target detection technology is widely used and plays a huge role in public and national defense security fields such as air and sea target early warning detection. However, in the problem of complex background target detection, the target echo is usually submerged in strong clutter backgrounds such as the sea surface and the ground, which brings great challenges to target detection. Therefore, a robust and efficient clutter suppression method is a key technology in achieving target detection in complex backgrounds.

[0003] Traditional clutter suppression methods usually use the energy difference between the target echo and the clutter to distinguish them, or use the Doppler frequency difference between the target echo and the clutter to distinguish them, so as to achieve clutter suppression through corresponding processing. However, in the process of implementing the present invention, the inventor found that the clutter suppression method based on energy difference uses the method of setting an empirical threshold to screen out some clutter, but this method cannot suppress strong clutter in the signal. Therefore, the clutter suppression method based on energy difference cannot effectively improve the target detection performance under complex backgrounds. The method based on Doppler frequency difference removes the clutter Doppler frequency corresponding component by fast Fourier transform processing on the radar signal, thereby achieving clutter suppression. When the target moves slowly, the target Doppler frequency is similar to the clutter Doppler frequency, and the fast Fourier transform processing cannot effectively separate the target echo and the clutter, so that the clutter cannot be suppressed to improve the detection performance. In summary, the traditional clutter suppression method still has the technical problem of low target detection performance under complex backgrounds, and a robust and effective clutter suppression method is urgently needed to improve the target detection performance under complex backgrounds. Summary of the invention

[0004] In view of the problems existing in the above-mentioned traditional methods, the present invention proposes a manifold filtering clutter suppression method based on geodetic coordinates, which can greatly improve the target detection performance in complex backgrounds, a manifold filtering clutter suppression device based on geodetic coordinates, a radar signal processing device and a computer-readable storage medium.

[0005] In order to achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0006] On the one hand, a method for suppressing clutter by manifold filtering based on geodetic coordinates is provided, comprising the steps of:

[0007] Acquire the radar echo signal to be processed;

[0008] Perform matched filtering on the radar echo signal to obtain an echo signal matrix and divide the echo signal matrix into multiple range pulse units;

[0009] The covariance matrix corresponding to each distance pulse unit is constructed using the Hamming window;

[0010] Calculate the geodesic coordinates of each covariance matrix separately;

[0011] After vectorizing each geodetic coordinate, the PCA algorithm is used to reduce the dimension into the corresponding three-dimensional vector, and each three-dimensional vector is visualized in three-dimensional space;

[0012] Determine the clutter suppression area in the three-dimensional space according to the visualization results of each three-dimensional vector;

[0013] Determine the manifold filter according to the three-dimensional space, use the manifold filter to perform matrix transformation processing on all covariance matrices, and obtain the target covariance matrix after clutter suppression;

[0014] The trace of the target covariance matrix is ​​used as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0015] In one embodiment, the step of constructing the covariance matrix corresponding to each range pulse unit using a Hamming window includes:

[0016] For each distance pulse unit, respectively, a set number of vectors before and after adjacent pulses of the same distance are taken to form a complex vector corresponding to each distance pulse unit;

[0017] After multiplying each complex vector by the Hamming window, the covariance matrix corresponding to each range pulse unit is constructed.

[0018] In one embodiment, the step of calculating the geodesic coordinates of each covariance matrix includes:

[0019] Use all covariance matrices to solve and obtain the coordinate origin of the geodetic coordinates;

[0020] The geodesic coordinates of each covariance matrix are calculated based on the coordinate origin.

[0021] In one embodiment, the step of determining the clutter suppression area in the three-dimensional space according to the visualization results of each three-dimensional vector includes:

[0022] Determine the parameters of the preset plane equation according to the visualization results of each three-dimensional vector;

[0023] According to the preset plane equation after the parameters are determined, the three-dimensional space is divided into two parts; one part is the clutter suppression area, and the other part is the non-clutter suppression area.

[0024] In one embodiment, the method for determining the parameters of the preset plane equation includes:

[0025] The determination is made through manual interpretation based on the visualization results of each three-dimensional vector, or through the assistance of prior information based on the visualization results of each three-dimensional vector.

[0026] In one embodiment, the process of dividing the echo signal matrix into a plurality of range pulse units includes:

[0027] According to the pulse length and the number of pulses in the echo signal matrix, the echo signal matrix is ​​divided into a plurality of range pulse units.

[0028] In one embodiment, the method further comprises the steps of:

[0029] The radar is controlled to transmit a coherent pulse train and receive a radar echo signal of the coherent pulse train.

[0030] On the other hand, a manifold filtering clutter suppression device based on geodetic coordinates is also provided, comprising:

[0031] An echo acquisition module is used to acquire the radar echo signal to be processed;

[0032] A filtering and dividing module is used to perform matched filtering on the radar echo signal to obtain an echo signal matrix and divide the echo signal matrix into a plurality of range pulse units;

[0033] A matrix construction module, used to construct the covariance matrix corresponding to each distance pulse unit by using a Hamming window;

[0034] A coordinate calculation module, used to calculate the geodesic coordinates of each covariance matrix respectively;

[0035] A three-dimensional processing module, which is used to vectorize each geodetic coordinate and then reduce the dimension into a corresponding three-dimensional vector using the PCA algorithm, and visualize each three-dimensional vector in three-dimensional space;

[0036] A region determination module, used for determining a clutter suppression region in a three-dimensional space according to the visualization results of each three-dimensional vector;

[0037] A manifold filter module is used to determine a manifold filter according to the three-dimensional space, and use the manifold filter to perform matrix transformation processing on all covariance matrices to obtain a target covariance matrix after clutter suppression;

[0038] The result output module is used to use the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0039] On the other hand, a radar signal processing device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned manifold filtering clutter suppression method based on geodetic coordinates when executing the computer program.

[0040] On the other hand, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned manifold filtering clutter suppression method based on geodetic coordinates are implemented.

[0041] One of the above technical solutions has the following advantages and beneficial effects:

[0042] The above-mentioned manifold filtering clutter suppression method, device and equipment based on geodesic coordinates, after performing matched filtering processing on the obtained radar echo signal and dividing it into range pulse units, constructing the covariance matrix of each unit, calculating the geodesic coordinates of the covariance matrix and then performing vectorization, dimensionality reduction and visualization processing, and then determining the clutter suppression area based on the three-dimensional visualization result, and then determining the manifold filter and using it for matrix transformation processing of the covariance matrix to obtain the target covariance matrix after clutter suppression, and finally using the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0043] Compared with the traditional method, the above scheme uses the radar signal covariance matrix to represent its energy, time correlation, Doppler and other information, which can comprehensively consider the differences between clutter and target echo in these three aspects, avoiding the problem of being unable to effectively distinguish the target when relying on only one feature. At the same time, the geodesic coordinates of the signal covariance matrix are calculated to characterize the distribution area of ​​the matrix on the Hermitian positive definite matrix manifold. Since the coordinates reflect the differences between the target echo and clutter in energy, time correlation, Doppler and other characteristics, the target echo and clutter can be effectively distinguished. Therefore, the manifold mapping is designed using the geodesic coordinates and the radar signal is processed by manifold mapping. When the traditional method cannot distinguish the target echo from the clutter, the clutter suppression in complex scenes can be realized to accurately detect the target echo, thereby greatly improving the target detection performance under complex backgrounds. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the conventional technology, the drawings required for use in the embodiments or the conventional technology descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 1 is a flow chart of a method for suppressing clutter using a manifold filter based on geodetic coordinates in one embodiment;

[0046] Figure 2 is a flow chart of a manifold filtering clutter suppression method based on geodetic coordinates in another embodiment;

[0047] Figure 3 is a schematic diagram of the amplitude of an echo signal matrix in one embodiment;

[0048] Figure 4 is a schematic diagram of three-dimensional vector visualization in one embodiment;

[0049] Figure 5 A schematic diagram of the amplitude of an echo signal matrix of data used in an experiment in one embodiment;

[0050] Figure 6 A schematic diagram of three-dimensional vector visualization of data used in an experiment in an embodiment;

[0051] Figure 7 is a schematic diagram of clutter suppression experimental results obtained using the first set of parameters in one embodiment;

[0052] Figure 8 is a schematic diagram of clutter suppression experimental results obtained using the second set of parameters in one embodiment;

[0053] Fig. 9 Schematic diagram of the module structure of a manifold filtering clutter suppression device based on geodetic coordinates in one embodiment. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0055] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0056] It should be noted that the reference to "embodiment" herein means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The presentation of the phrase at various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments.

[0057] Those skilled in the art will appreciate that the embodiments described herein may be combined with other embodiments. The term "and / or" used in the present specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0058] In the research work, the inventor found that the covariance matrix is ​​a high-dimensional feature that characterizes the statistical characteristics of the signal, which contains the energy, time correlation and Doppler information of the radar signal. Matrix information geometry is an emerging discipline that studies the geometric structure of the matrix from the perspective of information. In the theory of matrix information geometry, the covariance matrix of radar echo data constitutes a Hermitian positive definite matrix manifold. On this matrix manifold, the geometric distance between the two covariance matrices represents the comprehensive difference in energy, time correlation and Doppler frequency between the two corresponding radar signals. Due to the characteristic differences between the target echo data and the clutter data, the covariance matrices corresponding to the two will be distributed in different areas on the Hermitian positive definite matrix manifold. Geodesic coordinates are a special coordinate system on the manifold, which can represent the geometric distribution of the covariance matrix on the manifold, such as the distance and direction of the matrix from the origin of the coordinates. The present application uses geodesic coordinates to represent the position of the radar signal on the manifold, and uses manifold mapping to reduce the energy of the clutter area, thereby achieving clutter suppression under complex backgrounds.

[0059] The following will describe the implementation of the present invention in detail with reference to the accompanying drawings in the embodiment diagram of the present invention.

[0060] See also Figure 1 In one embodiment, the present application provides a manifold filtering clutter suppression method based on geodetic coordinates, comprising the following processing steps S12 to S26:

[0061] S12, obtaining the radar echo signal to be processed.

[0062] It can be understood that the radar acquires the corresponding radar echo signal by transmitting coherent pulses to the detection area, and the radar echo signal is usually composed of multiple echoes.

[0063] S14, performing matched filtering processing on the radar echo signal to obtain an echo signal matrix and dividing the echo signal matrix into a plurality of range pulse units.

[0064] It can be understood that in this field, the transmitted signal waveform can be used to design a matched filter so as to perform matched filtering processing on each pulse in the radar echo signal. The processed pulse vectors can be arranged to form an echo signal matrix. The dimension of the matrix is ​​determined by the number and length of the pulse vectors, and the elements in the matrix can be divided into range-pulse units (referred to as range-pulse units).

[0065] S16, using a Hamming window to construct the covariance matrix corresponding to each range pulse unit.

[0066] It can be understood that for each distance pulse unit, several data of adjacent pulses at the same distance can be taken, and after applying a Hamming window, a covariance matrix corresponding to each unit can be constructed.

[0067] S18, calculating the geodesic coordinates of each covariance matrix respectively.

[0068] S20, vectorizing each geodetic coordinate and reducing the dimension into a corresponding three-dimensional vector using a PCA algorithm, and visualizing each three-dimensional vector in a three-dimensional space.

[0069] It can be understood that the PCA algorithm is also known as Principle Compoent Analysis. After calculating the geodesic coordinates of each covariance matrix, the geodesic coordinates can be vectorized and then the PCA algorithm can be used to reduce the dimension of all vectorized coordinate vectors to three-dimensional vectors, and all three-dimensional vectors can be visualized in three-dimensional space.

[0070] S22, determining a clutter suppression area in the three-dimensional space according to the visualization results of each three-dimensional vector;

[0071] S24, determining a manifold filter according to the three-dimensional space, and using the manifold filter to perform matrix transformation processing on all covariance matrices to obtain a target covariance matrix after clutter suppression.

[0072] It can be understood that the filter suppression area in the three-dimensional space can be set by visualizing the results, and then the three-dimensional space is used to determine the applicable manifold filter, and then the manifold filter is used to perform matrix transformation processing on all covariance matrices to filter the three-dimensional vectors in the clutter suppression area, thereby suppressing the covariance matrix corresponding to the clutter. The target covariance matrix is ​​the covariance matrix corresponding to the three-dimensional vectors originally determined as the non-clutter suppression area after clutter suppression.

[0073] S26, using the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0074] It can be understood that for each range-pulse unit, the trace of the covariance matrix obtained by the corresponding processing is taken as the energy value, and the final clutter suppression result can be obtained, that is, the echo signal matrix after clutter suppression by manifold filtering.

[0075] The above-mentioned manifold filtering clutter suppression method based on geodesic coordinates, after performing matched filtering processing on the obtained radar echo signal and dividing it into range pulse units, constructs the covariance matrix of each unit, calculates the geodesic coordinates of the covariance matrix and then performs vectorization, dimensionality reduction and visualization processing, and then determines the clutter suppression area based on the three-dimensional visualization result, and then determines the manifold filter and uses it for matrix transformation processing of the covariance matrix to obtain the target covariance matrix after clutter suppression, and finally uses the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0076] Compared with the traditional method, the above scheme uses the radar signal covariance matrix to represent its energy, time correlation, Doppler and other information, which can comprehensively consider the differences between clutter and target echo in these three aspects, avoiding the problem of being unable to effectively distinguish the target when relying on only one feature. At the same time, the geodesic coordinates of the signal covariance matrix are calculated to characterize the distribution area of ​​the matrix on the Hermitian positive definite matrix manifold. Since the coordinates reflect the differences between the target echo and clutter in energy, time correlation, Doppler and other characteristics, the target echo and clutter can be effectively distinguished. Therefore, the manifold mapping is designed using the geodesic coordinates and the radar signal is processed by manifold mapping. When the traditional method cannot distinguish the target echo from the clutter, the clutter suppression in complex scenes can be realized to accurately detect the target echo, thereby greatly improving the target detection performance under complex backgrounds.

[0077] In one embodiment, Figure 2 As shown, the above method further includes step S10:

[0078] S10, controlling the radar to transmit a coherent pulse train and receiving a radar echo signal of the coherent pulse train.

[0079] It can be understood that in practical applications, a radar can be used to transmit a coherent pulse train, such as M coherent pulses, and receive a radar echo signal of the pulse train. For ease of description, the radar echo signal can be recorded as Represents the jth echo pulse, j=1, 2,…,M.

[0080] In one embodiment, the process of dividing the echo signal matrix into a plurality of range pulse units in the above step S14 may specifically include the following processing:

[0081] According to the pulse length and the number of pulses in the echo signal matrix, the echo signal matrix is ​​divided into a plurality of range pulse units.

[0082] It can be understood that after matching the received radar echo signal, the pulse vector sequence x1, x2, ..., x M, assuming that there is no vector with a length of N, then the echo signal matrix composed of vector arrangements is It is a complex N×M dimensional matrix, where the i-th row and j-th column in the matrix is ​​the (i, j)-th range-pulse unit, 1≤i≤N, 1≤j≤M. Figure 3 As shown, it is a schematic diagram of the result obtained by taking the amplitude of each element in the echo signal matrix X.

[0083] In one embodiment, the above step S16 may further include the following processing steps:

[0084] For each distance pulse unit, respectively, a set number of vectors before and after adjacent pulses of the same distance are taken to form a complex vector corresponding to each distance pulse unit;

[0085] After multiplying each complex vector by the Hamming window, the covariance matrix corresponding to each range pulse unit is constructed.

[0086] It can be understood that the number of adjacent units can be set to l (that is, the number is set, and its specific value can be selected according to the actual application needs). For the distance-pulse unit (i, j), take (i, jl), ..., (i, j), ..., (i, j+l), a total of 2l+1 data to form a complex vector, and then multiply the complex vector by the Hamming window to construct its corresponding covariance matrix.

[0087] Specifically, for each distance-pulse unit (i, j), the data of the l units before and after the adjacent pulses at the same distance are taken to form a complex vector with a length of 2l+1, which is expressed as:

[0088]

[0089] Multiply the complex vector by the Hamming window w = [w1w2…w 2l+1 ] and then get the vector:

[0090] y i,j =[w1x i,j-l wxya i,j-l+1 … w l+1 x i,j … w 2l x i,j+l-1 w 2l+1 x i,j+l ]

[0091] Using vector y i,j Construct the covariance matrix C i,j , which is constructed as follows:

[0092]

[0093] in,

[0094]

[0095] Here Represents x k The conjugation of.

[0096] In one embodiment, the above step S18 may further include the following processing steps:

[0097] Use all covariance matrices to solve and obtain the coordinate origin of the geodetic coordinates;

[0098] The geodesic coordinates of each covariance matrix are calculated based on the coordinate origin.

[0099] It can be understood that for calculating all NM covariance matrices C i,j (1≤i≤N,1≤j≤M) geodesic coordinates: First, use the covariance matrix of all NM to find the origin C0 of the geodesic coordinates. The process is as follows:

[0100]

[0101] Among them, exp(·) and log(·) are the exponential function and logarithmic function of the matrix respectively. Then, calculate each matrix C i,j Geodetic coordinates of Its expression is:

[0102]

[0103] In some embodiments, with respect to the above step S20, further, the geodesic coordinates corresponding to each range-pulse unit are Vectorize to get α i,j , the process is:

[0104]

[0105] in, and Respectively represent the geodetic coordinates The matrix composed of the real and imaginary parts of each element in. Using the PCA algorithm, all vectors α i,j Dimensionality reduction to three dimensions gives β i,j , and all NM three-dimensional vectors β i,j Visualize in three-dimensional space. Figure 4 As shown, Figure 3 All vectors of data used β i,j An example of 3D visualization.

[0106] In one embodiment, the above step S22 may further include the following processing steps:

[0107] Determine the parameters of the preset plane equation according to the visualization results of each three-dimensional vector;

[0108] According to the preset plane equation after the parameters are determined, the three-dimensional space is divided into two parts; one part is the clutter suppression area, and the other part is the non-clutter suppression area.

[0109] It can be understood that there are two preset plane equations, namely A1x+B1y+C1z=D1 and A2x+B2y+C2z=D2. The parameters of these two plane equations (A1, B1, C1, D1) and (A2, B2, C2, D2) need to be determined through visualization results to divide the three-dimensional space into two parts:

[0110] Ω={(x,y,z)|A1x+B1y+C1z>D1 and A2x+B2y+C2z>D2}

[0111]

[0112] Among them, Ω is set as the clutter suppression area (that is, where the clutter is located), and the other part is the non-clutter suppression area (that is, where the target echo is located).

[0113] In one embodiment, in the above step S22, the method for determining the parameters of the preset plane equation includes:

[0114] The determination is made through manual interpretation based on the visualization results of each three-dimensional vector, or through the assistance of prior information based on the visualization results of each three-dimensional vector.

[0115] Optionally, based on the above visualization results, the parameters of the above two plane equations can be directly determined by manual interpretation, or they can be directly determined by the assistance of commonly used prior information. Both parameter determination methods can quickly achieve the desired purpose, thereby quickly realizing three-dimensional partitioning.

[0116] In some implementations, regarding the above step S24, specifically, after setting non-negative real numbers r and s according to actual application needs, the manifold filter is determined based on the above three-dimensional partition as follows:

[0117]

[0118] Among them, ||·|| F represents the Frobenius norm of the matrix. i,j Obtaining a matrix by matrix transformation That is, the target covariance matrix.

[0119] Finally, for each range-pulse unit (i, j), let each element in the filtered signal matrix Y be That is, the final clutter suppression result (matrix) Y is obtained. Representation Matrix Trace of.

[0120] In some embodiments, in order to more intuitively and comprehensively illustrate the above-mentioned manifold filtering clutter suppression method based on geodetic coordinates, the following is an example of experimental verification of radar sea detection data published in the Journal of Radars to experimentally illustrate the manifold filtering clutter suppression method based on geodetic coordinates proposed in the present invention.

[0121] It should be noted that the implementation cases given in this specification are only illustrative and are not the sole limitation for the specific implementation cases of the present invention. Those skilled in the art can adopt the above-mentioned geodetic coordinate-based manifold filtering clutter suppression method to achieve clutter suppression for different application scenarios under the guidance of the implementation cases provided by the present invention.

[0122] The data file in the second group of data in the first phase of 2019 was selected, and the 404th to 1193th pulse units and the 1st to 100th distance units were selected as the experimental data. Two simulation targets were added to the data to obtain the signal matrix X. Figure 5 This is the result graph obtained by taking the amplitude of each element in the signal matrix X.

[0123] Set parameter l = 5, execute steps S16 to S20 in the above method of this application, and obtain all vectors β i,j The 3D visualization results are as follows Figure 6 As shown. By manual interpretation, the two planes A1x+B1y+C1z=D1 and A2x+B2y+C2z=D2 (as shown Figure 6 Plane 1 and Plane 2 are indicated in the figure. The relevant parameters are shown in Table 1 below:

[0124] Table 1

[0125] A B C D Plane 1 -0.2853 0.5053 1.7791 -0.8 Plane 2 -0.3941 -1.6037 1.4762 -1.2

[0126] Setting the manifold filter parameters to (r = 0.1, s = 2) and (r = 0.0001, s = 4) respectively, we get the following Figure 7 and Figure 8 The filtering result diagram is shown. The results show that the method proposed in this application can more effectively filter out clutter in a complex background.

[0127] It should be understood that although Figure 1 and Figure 2The steps in the flowchart are shown in sequence as indicated by the arrows, but the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of the steps, and the steps can be executed in other orders. Figure 1 and Figure 2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequentially, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0128] See also Fig. 9 In one embodiment, a manifold filtering clutter suppression device 100 based on geodetic coordinates is provided, comprising an echo acquisition module 11, a filter division module 13, a matrix construction module 15, a coordinate calculation module 17, a three-dimensional processing module 19, a region determination module 21, a manifold filtering module 23 and a result output module 25. Among them, the echo acquisition module 11 is used to obtain the radar echo signal to be processed. The filter division module 13 is used to perform matched filtering on the radar echo signal to obtain an echo signal matrix and divide the echo signal matrix into a plurality of range pulse units. The matrix construction module 15 is used to use a Hamming window to respectively construct the covariance matrix corresponding to each range pulse unit. The coordinate calculation module 17 is used to respectively calculate the geodetic coordinates of each covariance matrix. The three-dimensional processing module 19 is used to respectively vectorize each geodetic coordinate and then use the PCA algorithm to reduce the dimension into a corresponding three-dimensional vector, and visualize each three-dimensional vector in three-dimensional space. The region determination module 21 is used to determine the clutter suppression region in three-dimensional space according to the visualization results of each three-dimensional vector. The manifold filter module 23 is used to determine the manifold filter according to the three-dimensional space, and use the manifold filter to perform matrix transformation processing on all covariance matrices to obtain the target covariance matrix after clutter suppression. The result output module 25 is used to use the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0129] The above-mentioned manifold filtering clutter suppression device 100 based on geodetic coordinates, through the cooperation of various modules, performs matched filtering processing on the obtained radar echo signal and divides it into range pulse units, constructs the covariance matrix of each unit, calculates the geodetic coordinates of the covariance matrix and then performs vectorization, dimensionality reduction and visualization processing, and then determines the clutter suppression area based on the three-dimensional visualization result, and then determines the manifold filter and uses it for matrix transformation processing of the covariance matrix to obtain the target covariance matrix after clutter suppression, and finally uses the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

[0130] Compared with the traditional method, the above scheme uses the radar signal covariance matrix to represent its energy, time correlation, Doppler and other information, which can comprehensively consider the differences between clutter and target echo in these three aspects, avoiding the problem of being unable to effectively distinguish the target when relying on only one feature. At the same time, the geodesic coordinates of the signal covariance matrix are calculated to characterize the distribution area of ​​the matrix on the Hermitian positive definite matrix manifold. Since the coordinates reflect the differences between the target echo and clutter in energy, time correlation, Doppler and other characteristics, the target echo and clutter can be effectively distinguished. Therefore, the manifold mapping is designed using the geodesic coordinates and the radar signal is processed by manifold mapping. When the traditional method cannot distinguish the target echo from the clutter, the clutter suppression in complex scenes can be realized to accurately detect the target echo, thereby greatly improving the target detection performance under complex backgrounds.

[0131] In one embodiment, the matrix construction module 15 includes a complex vector quantum module and a Hamming matrix submodule. The complex vector quantum module is used to take a set number of vectors before and after the adjacent pulses of the same distance for each distance pulse unit to form a complex vector corresponding to each distance pulse unit. The Hamming matrix submodule is used to construct a covariance matrix corresponding to each distance pulse unit after multiplying each complex vector by a Hamming window.

[0132] In one embodiment, the coordinate calculation module 17 includes an origin calculation submodule and a geodesic calculation submodule. The origin calculation submodule is used to solve the coordinate origin of the geodesic coordinates using all covariance matrices. The geodesic calculation submodule is used to calculate the geodesic coordinates of each covariance matrix according to the coordinate origin.

[0133] In one embodiment, the region determination module 21 includes a parameter submodule and a space division submodule. The parameter submodule is used to determine the parameters of the preset plane equation according to the visualization results of each three-dimensional vector. The space division submodule is used to divide the three-dimensional space into two parts according to the preset plane equation after the parameters are determined; wherein one part is the clutter suppression area and the other part is the non-clutter suppression area.

[0134] In one embodiment, the parameter submodule determines the parameters of the preset plane equation by manually interpreting and determining the parameters according to the visualization results of each three-dimensional vector, or by assisting in determining the parameters according to the visualization results of each three-dimensional vector with prior information.

[0135] In one embodiment, the filtering division module 13 can be used to divide the echo signal matrix into multiple range pulse units according to the pulse length and the number of pulses in the echo signal matrix.

[0136] In one embodiment, the above-mentioned manifold filtering clutter suppression device 100 based on geodetic coordinates may further include a signal transceiver module for controlling the radar to transmit a coherent pulse train and receive radar echo signals of the coherent pulse train.

[0137] Regarding the specific definition of the manifold filter clutter suppression device 100 based on geodetic coordinates, please refer to the corresponding definition of the manifold filter clutter suppression method based on geodetic coordinates in the above text, which will not be repeated here. Each module in the above-mentioned manifold filter clutter suppression device 100 based on geodetic coordinates can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of a device with a specific data processing function in the form of hardware, or can be stored in the memory of the aforementioned device in the form of software, so that the processor can call and execute the operations corresponding to the above modules. The aforementioned device can be, but is not limited to, various types of signal detection devices already available in the art.

[0138] In one embodiment, a radar signal processing device is also provided, including a memory and a processor, the memory stores a computer program, and the processor implements the following processing steps when executing the computer program: obtaining a radar echo signal to be processed; performing matched filtering processing on the radar echo signal to obtain an echo signal matrix and dividing the echo signal matrix into a plurality of range pulse units; using a Hamming window to respectively construct a covariance matrix corresponding to each range pulse unit; respectively calculating the geodesic coordinates of each covariance matrix; respectively vectorizing each geodesic coordinate and reducing the dimension into a corresponding three-dimensional vector using a PCA algorithm, and visualizing each three-dimensional vector in a three-dimensional space; determining a clutter suppression region in a three-dimensional space according to the visualization results of each three-dimensional vector; determining a manifold filter according to the three-dimensional space, and using the manifold filter to perform matrix transformation processing on all covariance matrices to obtain a target covariance matrix after clutter suppression; using the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain a final clutter suppression result.

[0139] It can be understood that the above-mentioned radar signal processing device, in addition to the memory and processor mentioned above, also includes other software and hardware components not listed in this specification. The specific components can be determined according to the model of the specific signal processing device in different application scenarios, and this specification will not list them one by one in detail.

[0140] In one embodiment, when the processor executes the computer program, the processor may also implement the additional steps or sub-steps in each embodiment of the above-mentioned manifold filtering clutter suppression method based on geodetic coordinates.

[0141] In one embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, the following processing steps are implemented: obtaining a radar echo signal to be processed; performing matched filtering processing on the radar echo signal to obtain an echo signal matrix and dividing the echo signal matrix into a plurality of range pulse units; using a Hamming window to respectively construct a covariance matrix corresponding to each range pulse unit; respectively calculating the geodesic coordinates of each covariance matrix; respectively vectorizing each geodesic coordinate and reducing the dimension into a corresponding three-dimensional vector using a PCA algorithm, and visualizing each three-dimensional vector in a three-dimensional space; determining a clutter suppression region in a three-dimensional space according to the visualization results of each three-dimensional vector; determining a manifold filter according to the three-dimensional space, and using the manifold filter to perform matrix transformation processing on all covariance matrices to obtain a target covariance matrix after clutter suppression; using the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain a final clutter suppression result.

[0142] In one embodiment, when the computer program is executed by a processor, it can also implement the steps or sub-steps added in the above embodiments of the manifold filtering clutter suppression method based on geodetic coordinates.

[0143] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus dynamic random access memory (Rambus DRAM, referred to as RDRAM) and interface dynamic random access memory (DRDRAM).

[0144] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0145] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and all of them belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A manifold filtering clutter suppression method based on geodetic coordinates, characterized in that: Includes steps: Acquire the radar echo signal to be processed; Performing matched filtering on the radar echo signal to obtain an echo signal matrix and dividing the echo signal matrix into a plurality of range pulse units; Using a Hamming window, respectively construct a covariance matrix corresponding to each of the distance pulse units; Calculating the geodesic coordinates of each of the covariance matrices respectively; After vectorizing each of the geodetic coordinates, the PCA algorithm is used to reduce the dimension into a corresponding three-dimensional vector, and each of the three-dimensional vectors is visualized in a three-dimensional space; Determining a clutter suppression area in the three-dimensional space according to the visualization results of each of the three-dimensional vectors; Determine a manifold filter according to the three-dimensional space, and use the manifold filter to perform matrix transformation processing on all the covariance matrices to obtain a target covariance matrix after clutter suppression; The trace of the target covariance matrix is ​​used as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

2. The method for suppressing clutter by manifold filtering based on geodetic coordinates according to claim 1, characterized in that: The steps of constructing the covariance matrices corresponding to the range pulse units using the Hamming window include: For each of the distance pulse units, respectively, a set number of vectors before and after adjacent pulses of the same distance are taken to form a complex vector corresponding to each of the distance pulse units; After multiplying each of the complex vectors by a Hamming window, the covariance matrix corresponding to each of the range pulse units is constructed.

3. The method for clutter suppression based on manifold filtering of geodetic coordinates according to claim 1 or 2, characterized in that: The step of respectively calculating the geodesic coordinates of each of the covariance matrices comprises: Using all the covariance matrices to solve and obtain the coordinate origin of the geodetic coordinates; The geodesic coordinates of each of the covariance matrices are calculated according to the coordinate origin.

4. The method for clutter suppression based on manifold filtering of geodetic coordinates according to claim 3 is characterized in that: The step of determining a clutter suppression area in the three-dimensional space according to the visualization results of each of the three-dimensional vectors comprises: Determining parameters of a preset plane equation according to the visualization results of each of the three-dimensional vectors; According to the preset plane equation after parameter determination, the three-dimensional space is divided into two parts; one part is a clutter suppression area, and the other part is a non-clutter suppression area.

5. The method for suppressing clutter by manifold filtering based on geodetic coordinates according to claim 4, characterized in that: The method for determining the parameters of the preset plane equation includes: The determination is performed manually based on the visualization results of each of the three-dimensional vectors, or is determined with the assistance of prior information based on the visualization results of each of the three-dimensional vectors.

6. The method for clutter suppression based on manifold filtering of geodetic coordinates according to claim 1, characterized in that: The process of dividing the echo signal matrix into a plurality of range pulse units includes: The echo signal matrix is ​​divided into a plurality of range pulse units according to the pulse length and the number of pulses in the echo signal matrix.

7. The method for clutter suppression based on geodetic coordinates according to claim 1, characterized in that: Also includes the steps: The radar is controlled to transmit a coherent pulse train and receive the radar echo signal of the coherent pulse train.

8. A manifold filter clutter suppression device based on geodetic coordinates, characterized in that: include: An echo acquisition module is used to acquire the radar echo signal to be processed; A filtering and dividing module, used for performing matched filtering on the radar echo signal to obtain an echo signal matrix and dividing the echo signal matrix into a plurality of range pulse units; A matrix construction module, used to construct the covariance matrix corresponding to each of the range pulse units using a Hamming window; A coordinate calculation module, used for respectively calculating the geodesic coordinates of each of the covariance matrices; A three-dimensional processing module, used to vectorize each of the geodesic coordinates and then reduce the dimension into a corresponding three-dimensional vector using a PCA algorithm, and visualize each of the three-dimensional vectors in a three-dimensional space; A region determination module, used for determining a clutter suppression region in the three-dimensional space according to the visualization results of each of the three-dimensional vectors; A manifold filtering module, used for determining a manifold filter according to the three-dimensional space, and performing matrix transformation processing on all the covariance matrices using the manifold filter to obtain a target covariance matrix after clutter suppression; The result output module is used to use the trace of the target covariance matrix as the element value of the corresponding echo signal matrix to obtain the final clutter suppression result.

9. A radar signal processing device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the manifold filtering clutter suppression method based on geodetic coordinates according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the manifold filtering clutter suppression method based on geodetic coordinates according to any one of claims 1 to 7 are implemented.

Images