A space-based radar sea surface target identification method

Abstract

The application provides a space-based radar sea surface target recognition method, which comprises the following steps: acquiring a sea surface high-resolution range profile (HRRP) matrix of a ship, performing amplitude normalization processing on the HRRP matrix, and superimposing and accumulating the normalized HRRP matrix along an azimuth direction; extracting structural features of the accumulated HRRP; extracting micro-motion features from the accumulated HRRP azimuth direction by using short-time Fourier transform and Fourier coefficient fitting respectively; inputting the structural features and the micro-motion features into a meta-classifier to obtain a sea surface target recognition result, wherein the meta-classifier is composed of a first base classifier and a second base classifier, the first base classifier is obtained by training according to structural feature samples, and the second base classifier is obtained by training according to micro-motion feature samples. The structural features and the micro-motion features of the HRRP echo matrix are extracted respectively, and the stacked classifier is used for feature fusion and feature classification, so that the problem that high-similarity false targets are difficult to distinguish in one-dimensional HRRP recognition is solved, and the recognition accuracy and reliability are improved.

Description

Technical Field

[0001] This application relates to the field of radar target recognition technology, and in particular to a method for identifying sea surface targets using space-based radar. Background Technology

[0002] Automatic Target Recognition (RATR) radar holds immense value in maritime transport management and navigation safety, with profound implications for both military and civilian applications. Space-based radar boasts a wide field of view, completely unrestricted by airspace or territorial waters, enabling continuous and prolonged maritime observation. It allows for all-weather, 24 / 7 surveillance of targets of interest, possessing significant potential for broad military and civilian applications. Furthermore, it holds crucial strategic importance for safeguarding national security and promoting economic and social development in my country.

[0003] Extensive research has been conducted on the identification of surface ships. However, existing identification methods based on high-resolution range profiles (HRRPs) primarily focus on the structural characteristics of the ships. The temporal echoes of radar ship target HRRPs also contain rich micro-motion characteristics. From the perspective of space-based radar, the rolling and undulating motion of a ship alters its attitude angle within the radar, causing the scattering point to move systematically across range cells. Summary of the Invention

[0004] This application aims to at least partially address one of the technical problems in the related art.

[0005] Therefore, the first objective of this application is to propose a method for identifying sea surface targets using space-based radar.

[0006] The second objective of this application is to propose a space-based radar sea surface target identification device.

[0007] The third objective of this application is to propose an electronic device.

[0008] The fourth objective of this application is to provide a computer-readable storage medium.

[0009] The fifth objective of this application is to provide a computer program product.

[0010] To achieve the above objectives, a first aspect of this application provides a space-based radar method for identifying sea surface targets, including:

[0011] The ship's high-resolution range image (HRRP) matrix is ​​obtained, the HRRP matrix is ​​normalized, and the normalized HRRP matrix is ​​superimposed and accumulated along the azimuth direction.

[0012] The structural features of the accumulated HRRP are extracted, including the equivalent scattering center dimension, equivalent size, entropy, standard deviation, bias, irregularity, frequency domain incoherent range profile, and central moment features.

[0013] Micro-motion features were extracted from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively.

[0014] The structural features and the micro-motion features are input into the meta-classifier to obtain the target recognition result on the sea surface. The meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on the structural feature samples, and the second base classifier is trained based on the micro-motion feature samples.

[0015] Optionally, the step of acquiring the high-resolution sea surface range image (HRRP) matrix of the ship, performing amplitude normalization on the HRRP matrix, and then stacking and accumulating the normalized HRRP matrix along the azimuth direction includes:

[0016] Obtain the HRRP matrix of the ship, denoted as:

[0017]

[0018] Where X is the HRRP matrix, M is the number of rows in the HRRP matrix, representing the number of time frames; N is the number of columns in the HRRP matrix, representing the number of distance units; x ij This represents the signal strength of a single element in the i-th row and j-th column of the HRRP matrix, i.e., the j-th distance unit in the i-th frame.

[0019] The HRRP matrix is ​​normalized row by row, where the HRRP of the i-th row is denoted as X. i =[x i1 …x iN The elements in the HRRP matrix are normalized according to the following formula, which is expressed as follows:

[0020]

[0021] The normalized HRRP matrices are superimposed along the azimuth direction, and the expression is:

[0022]

[0023] Among them, X i 'Indicates the accumulated HRRP.

[0024] Optionally, the extraction of the accumulated structural features of HRRP includes:

[0025] For the accumulated HRRP image, i.e. Xi ′=[x i1 …x iN ], denoted as X i =[x(1)…x(N)], the structural features are calculated using the following formula, including:

[0026] Calculate the dimension N of the equivalent scattering center scatter The mathematical expression is:

[0027]

[0028] Where ε(·) is the unit jump function, defined as:

[0029]

[0030] Calculate the equivalent size E size The mathematical expression is:

[0031] E size =S(N) scatter )-S(1)

[0032] Where S is the position vector of the range cell in the HRRP signal whose echo intensity is greater than half of the maximum value, and its calculation expression is:

[0033] S = {i|x(i)≥max(X) / 2}

[0034] The mathematical expression for calculating entropy is:

[0035]

[0036] The standard deviation std(dB) is calculated using the following mathematical expression:

[0037]

[0038] The mathematical expression for calculating deviation (dB) is:

[0039]

[0040] The mathematical expression for calculating the irregularity (dB) is:

[0041]

[0042] Calculate the frequency domain incoherent range image The mathematical expression is:

[0043]

[0044] To calculate the central moment characteristic, let the first-order origin moment be: Then the p-th order central moment C p The mathematical expression is:

[0045]

[0046] Optionally, the step of extracting micro-motion features from the accumulated HRRP orientation using Fourier coefficient fitting includes:

[0047] The range cell containing the strongest scattering point is extracted and its change over time. A Fourier series is used to fit the position of the strong scattering point within the range cell as a function of the ship's attitude angle. The coefficients of the Fourier series are then used as the ship's micro-motion characteristics. The formula for calculating the Fourier series is as follows:

[0048]

[0049] Where f(t) represents a function of the HRRP signal, with time t as the independent variable, a0 as the constant term in the Fourier series, and a n Let b be the coefficient of the cosine term in the Fourier series. n ω represents the coefficients of the sine term in the Fourier series, ω represents the angular frequency of the fundamental frequency, and n represents the harmonic order.

[0050] Optionally, the step of extracting micro-motion features from the accumulated HRRP orientation using short-time Fourier transform includes:

[0051] Multiple range cells near a strong scattering point are selected, and the changes in multiple range cells are monitored simultaneously using short-time Fourier transform. Let the accumulated HRRP of the i-th range cell be X. i , Let the amplitude of the i-th HRRP signal be the value of the n-th distance cell. Perform a short-time Fourier transform on the signal data in each distance cell, and the calculation formula is as follows:

[0052]

[0053] Where w(m) is the selected window function;

[0054] The multiple short-time Fourier transform matrices calculated from selected distance units are summed, and the summed matrix is ​​denoted as STFT. Singular value decomposition is then performed on the STFT matrix to reduce its feature dimension, i.e.:

[0055]

[0056] Among them, STFT m×n U represents the matrix after the short-time Fourier transform; m×k Let Σ be an m×k matrix containing the left singular vectors of the STFT matrix; k×k It is a k×k diagonal matrix containing singular values; It is the transpose of matrix V, with a size of k×n. Matrix V contains the right singular vector of the STFT matrix.

[0057] Based on the singular values ​​obtained from the decomposition, the micro-motion characteristics of the ship are extracted.

[0058] To achieve the above objectives, a second aspect of this application provides a space-based radar sea surface target identification device, comprising:

[0059] The preprocessing module is used to obtain the high-resolution range image (HRRP) matrix of the ship at sea surface, perform amplitude normalization on the HRRP matrix, and then stack and accumulate the normalized HRRP matrix along the azimuth direction.

[0060] The structural feature extraction module is used to extract the structural features of the accumulated HRRP, wherein the structural features include the equivalent scattering center dimension, equivalent size, entropy, standard deviation, bias, irregularity, frequency domain incoherent range profile, and central moment features;

[0061] The micro-motion feature extraction module is used to extract micro-motion features from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively.

[0062] The target recognition module is used to input the structural features and the micro-motion features into a meta-classifier to obtain the target recognition result on the sea surface. The meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on the structural feature samples, and the second base classifier is trained based on the micro-motion feature samples.

[0063] To achieve the above objectives, a third aspect of this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0064] The memory stores computer-executed instructions;

[0065] The processor executes computer execution instructions stored in the memory to implement the method as described in any one of the first aspects.

[0066] To achieve the above objectives, a fourth aspect of this application provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, are used to implement the method as described in any one of the first aspects.

[0067] To achieve the above objectives, a fifth aspect of this application provides a computer program product that, when executed by a processor, implements the method described in any one of the first aspects.

[0068] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0069] By extracting the structural and micro-motion features of the HRRP echo matrix separately, and using a stacked classifier for feature fusion and classification, the problem of difficulty in distinguishing highly similar false targets in one-dimensional HRRP recognition is overcome, thereby improving the recognition accuracy and reliability.

[0070] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0071] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0072] Figure 1 A schematic flowchart illustrating a space-based radar method for identifying sea surface targets provided in an embodiment of this application;

[0073] Figure 2 A flowchart illustrating the training and stacking process of the meta-classifier provided in the embodiments of this application;

[0074] Figure 3 This is a schematic diagram of the structure of a space-based radar sea surface target identification device provided in an embodiment of this application. Detailed Implementation

[0075] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0076] Automatic Target Recognition (RATR) radar holds immense value in maritime transport management and navigation safety, with profound implications for both military and civilian applications. Space-based radar boasts a wide field of view, completely unrestricted by airspace or territorial waters, enabling continuous and prolonged maritime observation. It allows for all-weather, 24 / 7 surveillance of targets of interest, possessing significant potential for broad military and civilian applications. Furthermore, it holds crucial strategic importance for safeguarding national security and promoting economic and social development in my country.

[0077] Extensive research has been conducted on the identification of surface ships. However, existing identification methods based on high-resolution range profiles (HRRPs) primarily focus on the structural characteristics of the ships. The temporal echoes of radar ship target HRRPs also contain rich micro-motion characteristics. From the perspective of space-based radar, the rolling and undulating motion of a ship alters its attitude angle within the radar, causing the scattering point to move systematically across range cells.

[0078] Furthermore, existing target recognition technologies face two main challenges: firstly, the attitude angles observed by space-based radar differ significantly from those of other radars, and current research has not fully demonstrated the applicability of recognition technologies to space-based radars; secondly, current research has failed to combine high-range resolution image structural features and micro-motion features.

[0079] However, as is well known, in the feature extraction stage, the more comprehensive and representative the extracted information is, the better the accuracy and reliability of the identification. In order to overcome the shortcomings of the current low accuracy of sea surface ship target identification, this application proposes a space-based radar sea surface target identification method, which combines micro-motion features with structural features, and performs feature fusion and identification through a stacked classifier.

[0080] Figure 1 This is a flowchart illustrating a space-based radar method for identifying sea surface targets, provided in an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0081] Step 101: Obtain the high-resolution range image (HRRP) matrix of the ship at sea surface, perform amplitude normalization on the HRRP matrix, and then stack and accumulate the normalized HRRP matrix along the azimuth direction.

[0082] In this embodiment of the application, the HRRP (High Resolution Range Image) matrix of the ship is first obtained to represent the signal reflection intensity at different time frames and different range units, denoted as:

[0083]

[0084] Where X is the HRRP matrix, M is the number of rows in the HRRP matrix, representing the number of time frames; N is the number of columns in the HRRP matrix, representing the number of range cells; x ij This represents the signal strength of a single element in the i-th row and j-th column of the HRRP matrix, i.e., the j-th distance cell in the i-th frame.

[0085] To enhance data comparability and signal-to-noise ratio, and to extract effective features, the HRRP matrix is ​​first normalized row by row, where the HRRP of the i-th row is denoted as X. i =[xi1 …x iN The elements in the HRRP matrix are normalized according to the following formula, which is expressed as follows:

[0086]

[0087] It should be noted that in radar signal processing, different targets and acquisition conditions may lead to amplitude variations. The main purpose of normalization is to eliminate differences in the amplitude of different HRRP signals, making the scale of different samples consistent. This helps to enhance the contrast of features, making subsequent processing and classification more stable.

[0088] Finally, the normalized HRRP matrices are superimposed along the azimuth direction, and the expression is:

[0089]

[0090] Among them, X i 'Indicates the accumulated HRRP.

[0091] In one possible embodiment, N = 10 is set.

[0092] It is understandable that in practical applications, the azimuth of the target may vary slightly, and analyzing each HRRP individually may result in a slight bias. The overlay operation can average across multiple azimuths, thereby better suppressing noise and preserving the main features of the target.

[0093] Step 102: Extract the structural features of the accumulated HRRP.

[0094] For the accumulated HRRP image, i.e. X i ′=[x i1 …x iN ], denoted as X i =[x(1)…x(N)], structural features can be extracted using the following formula, including:

[0095] 1. Calculate the dimension N of the equivalent scattering center. scatter This feature quantifies the number of distance cells exceeding a certain threshold in HRRP. It can remove some clutter effects without eliminating most signal characteristics. Its mathematical expression is:

[0096]

[0097] Where ε(·) is the unit jump function, defined as:

[0098]

[0099] In this feature extraction scenario, it can be used to determine whether the value of each distance cell is greater than zero and set the result to 1 or 0 so that distance cells that meet the conditions can be counted in the scattering center count.

[0100] 2. Calculate the equivalent size E size This quantification metric describes the radial dimension of the target. If S is defined as the position vector of the range cell in the HRRP signal where the echo intensity is greater than half of its maximum value, then:

[0101] S = {i|x(i)≥max(X) / 2}

[0102] The mathematical expression for the equivalent size is:

[0103] E size =S(N) scatter )-S(1)

[0104] 3. Calculate the entropy of the HRRP signal, an index used to quantify the randomness of variables. Its mathematical expression is:

[0105]

[0106] 4. Calculate the standard deviation of HRRP, std(dB). This indicator is used to quantify the dispersion of the data. Its mathematical expression is:

[0107]

[0108] 5. Calculate the deviation (dB) of HRRP, an indicator used to quantify the distribution characteristics of the data. Its mathematical expression is:

[0109]

[0110] 6. Calculate the HRRP irregularity (dB). This index quantifies the locality deviation of the HRRP signal, reflecting the relative magnitude relationship between the current distance cell and the preceding and following distance cells. Its mathematical expression is:

[0111]

[0112] 7. Calculate the frequency domain incoherent range profile of HRRP. This indicator can suppress the orientation sensitivity of HRRP. Its mathematical expression is:

[0113]

[0114] 8. Calculate the central moment feature of HRRP. The central moment is a basic translation-invariant feature that can effectively reveal the structural information of a ship.

[0115] Let the first-order moment at the origin be: Then the p-th order central moment C p The mathematical expression is:

[0116]

[0117] In the embodiments of this application, central moment features of orders 2 to 6 are selected.

[0118] Step 103: Extract micro-motion features from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively.

[0119] Existing HRRP-based identification of surface ships primarily focuses on the structural characteristics of the ship. However, the time-domain echo of radar ship target HRRP also contains rich micro-motion characteristics. From the perspective of space-based radar, the rolling and undulating motion of the ship changes its attitude angle in the radar, causing the scattering point to move regularly across range cells.

[0120] For ship motion models, this application mainly considers the ship's roll and the linear motion between the ship and the radar, which can cause HRRP flickering and range drift. To address these micro-motions, this application uses Fourier series to fit the position of strong scattering points in the range cell as a function of the ship's attitude angle; on the other hand, it employs short-time Fourier transform to extract the time-frequency features of specific cells.

[0121] Specifically, the steps for extracting micro-motion features from the accumulated HRRP orientation using Fourier coefficient fitting include:

[0122] The range cell containing the strongest scattering point is extracted and its change over time. A Fourier series is used to fit the position of the strong scattering point within the range cell as a function of the ship's attitude angle. The coefficients of the Fourier series are then used as the ship's micro-motion characteristics. The formula for calculating the Fourier series is as follows:

[0123]

[0124] Where f(t) represents a function of the HRRP signal, with time t as the independent variable, a0 as the constant term in the Fourier series, and a n Let b be the coefficient of the cosine term in the Fourier series. n ω represents the coefficients of the sine term in the Fourier series, ω represents the angular frequency of the fundamental frequency, and n represents the harmonic order.

[0125] It is understandable that different types of ships have different masses, centers of gravity, and shapes, and their micro-motion characteristics such as sway amplitude and sway period will also vary.

[0126] In one possible embodiment, the order of the Fourier series is set to 3, and its coefficients are used as the micromotion characteristics of the ship.

[0127] Specifically, the steps for extracting micro-motion features from the accumulated HRRP orientation using short-time Fourier transform include:

[0128] The purpose of selecting multiple range cells near the strong scattering point is to monitor the changes in the reflected signal on different range cells over time, and to simultaneously monitor the changes in multiple range cells using short-time Fourier transform.

[0129] Let the accumulated HRRP of the i-th image be X. i , Let the amplitude of the i-th HRRP signal be the value of the n-th distance cell. Perform a short-time Fourier transform on the signal data in each distance cell, and the calculation formula is as follows:

[0130]

[0131] Here, w(m) is the selected window function used for weighting to smooth the data. By performing Fourier transform on the data within different time windows, frequency domain features are extracted, thereby capturing micro-motion information.

[0132] Furthermore, the multiple short-time Fourier transform matrices calculated from the selected range cells are summed, and the summed matrix is ​​denoted as STFT. Singular value decomposition is then performed on the STFT matrix to reduce its feature dimension, i.e.:

[0133]

[0134] Among them, STFT m×n The matrix represented by U is the result of signal transformation after multiple frames of data and multiple range units, used to describe the micro-motion characteristics of the target; m×k Σ is an m×k matrix containing the left singular vectors of the STFT matrix, used to preserve the main temporal or distance-related features of the matrix; k×k Let be a k×k diagonal matrix containing singular values, arranged in descending order of magnitude, representing the energy distribution and principal component strengths in the data. The largest singular value corresponds to the most dominant feature, and the gradually decreasing singular values ​​correspond to secondary features. It is the transpose of the V matrix, with a size of k×n. The V matrix contains the right singular vectors of the STFT matrix, which represent the contributions of different features of the signal in the frequency domain.

[0135] Finally, through singular value decomposition, this application can choose to retain larger singular values ​​and their corresponding vectors, thereby reducing the data dimensionality while retaining the main features of the signal to describe the micro-motion features of the target.

[0136] Step 104: Input the structural features and micro-motion features into the meta-classifier to obtain the sea surface target recognition result.

[0137] In this embodiment, the meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on structural feature samples, and the second base classifier is trained based on micro-motion feature samples.

[0138] It is understandable that a meta-classifier is the same as a stacked classifier. A stacked classifier is an ensemble learning technique that uses multiple base classifiers to learn and predict different features, and then uses the predicted probabilities as new features to train a new meta-classifier, thereby achieving feature fusion and classification.

[0139] In one possible embodiment, a flowchart illustrating the training and stacking process of the meta-classifier is shown below. Figure 2 As shown.

[0140] Reference Figure 2 First, training samples are collected and preprocessed.

[0141] It is understood that the training samples are high-resolution range images of the ship's other sea surface HRRP matrix samples, and the preprocessing steps can be synchronously referred to the normalization step and accumulation step in step 101. This application will not repeat the description of this step.

[0142] Then, based on the preprocessed training samples, structural features and micro-motion features are extracted, and different base classifiers are trained separately to learn and recognize the structural features of sea surface targets and the micro-motion features of targets. Each base classifier is trained independently to optimally identify and classify different target types within its specific feature domain.

[0143] After training, the two base classifiers are stacked to form a meta-classifier. The meta-classifier performs further comprehensive analysis and decision-making based on the information obtained from the first and second base classifiers. By combining structural features and micro-motion features, the meta-classifier can more comprehensively identify and classify sea surface targets, improving classification accuracy and robustness.

[0144] Understandably, after each classifier is trained, it needs to be tested using test samples. Evaluation using test samples allows us to obtain performance metrics such as accuracy, precision, and recall for each base classifier, and to identify potential overfitting or underfitting issues.

[0145] Finally, the meta-classifier, based on the outputs of the two base classifiers and the test results, derives the final sea surface target identification result. Furthermore, feedback from test samples allows for further optimization of the meta-classifier's weights or stacking method, improving its overall performance.

[0146] In this application embodiment, support vector machine, random forest, and logistic regression classifier can be used as the base classifier, and logistic regression classifier can be used as the meta classifier. This application does not make specific limitations on this.

[0147] To further demonstrate the beneficial effects of this application, a set of simulation experiments were also conducted, and the specific results are shown in the following embodiments.

[0148] Example 1

[0149] (1) Simulation experimental conditions:

[0150] This invention uses the electromagnetic simulation software CST to simulate simulation data of a certain X-band broadband space-based radar on surface ships. Four models are used: a small civilian ship (18 meters), a Fujian warship (178 meters), a Type 055 guided-missile destroyer (158 meters), and a diluent corner reflector array (25 meters). All models were adjusted and scaled using Hypermesh software.

[0151] (2) Experimental data setup:

[0152] The training dataset contains 280 HRRP matrix data sets per class, with each HRRP matrix containing 1280 frames of one-dimensional HRRP; the test dataset contains 120 HRRP matrix data sets per class, with each HRRP matrix containing 1280 frames of one-dimensional HRRP.

[0153] This application uses three different classifiers as base classifiers for comparison: Support Vector Machine, Random Forest, and Logistic Regression. Specific parameter settings are as follows:

[0154] 1) Support Vector Machine: SVM uses grid search to optimize. The kernel function of the base classifier trained by structural features is set to radial basis function, and the kernel function coefficient gamma is set to 0.1; the kernel function of the base classifier trained by micro-motion features is set to radial basis function, and the kernel function coefficient gamma is set to 1.

[0155] 2) Random Forest: The decision tree of the base classifier trained with structural features is set to 100; the decision tree of the base classifier trained with micro-motion features is set to 1000.

[0156] 3) Logistic Regression: The maximum number of iterations for the logistic regression classifier is 10,000.

[0157] 4) Meta-classifier: The meta-classifiers all use logistic regression classifiers, with a maximum number of iterations of 1000.

[0158] (3) Experimental Results

[0159] Tables 1, 2, and 3 show the recognition accuracy rates of the base classifier trained using only structural features, the base classifier trained using only micro-motion features, and the meta-classifier trained using both structural and micro-motion features, respectively. The data in Tables 1, 2, and 3 show that the recognition accuracy of the classifier is significantly improved after fusing structural and micro-motion features, proving the effectiveness of this fusion method.

[0160] Table 1

[0161] Classifier methods accuracy SVM 80.3％ Random Forest 85.4％ Logistic Regression 64.0％

[0162] As shown in Table 1, when using only structural features, Random Forest performed best, achieving an accuracy of 85.4%, followed by SVM, while Logistic Regression had the lowest accuracy.

[0163] Table 2

[0164] Classifier methods accuracy SVM 72.0％ Random Forest 70.6％ Logistic Regression 66.9％

[0165] As shown in Table 2, the recognition accuracy is generally low when using only micro-motion features, especially for Random Forest and SVM, where the accuracy is below 72%. Micro-motion features may be more unstable in recognition than structural features, especially when the structural features of the target are not obvious, and the classification effect of relying solely on micro-motion features is not ideal.

[0166] Table 3

[0167] Classifier methods accuracy SVM 87.1％ Random Forest 88.5％ Logistic Regression 75.2％

[0168] As shown in Table 3, the accuracy of all classifiers significantly improved after fusing structural features and micro-motion features. Random Forest achieved an accuracy of 88.5%, followed by SVM at 87.1%, and Logistic Regression also improved to 75.2%. Compared with the data in Tables 1 and 2, these figures show a significant improvement, especially for Random Forest and SVM.

[0169] The comparison of the above data leads to the conclusion that fusing structural features and micro-motion features can significantly improve the recognition accuracy of the classifier. This fusion method utilizes the stability of structural features and the dynamic information of micro-motion features, thereby providing a more comprehensive description of the target and improving classification performance. This improvement has practical significance for applications such as sea surface target recognition.

[0170] To achieve the above embodiments, this application also proposes a space-based radar sea surface target identification device. Figure 3 This is a schematic diagram of the structure of a space-based radar sea surface target identification device 10 provided in an embodiment of this application.

[0171] like Figure 3 As shown, the device includes:

[0172] The preprocessing module 100 is used to acquire the high-resolution range image (HRRP) matrix of the ship at sea surface, perform amplitude normalization on the HRRP matrix, and then stack and accumulate the normalized HRRP matrix along the azimuth direction.

[0173] The structural feature extraction module 200 is used to extract the structural features of the accumulated HRRP, including the equivalent scattering center dimension, equivalent size, entropy, standard deviation, bias, irregularity, frequency domain incoherent range profile, and central moment features.

[0174] The micro-motion feature extraction module 300 is used to extract micro-motion features from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively.

[0175] The target recognition module 400 is used to input structural features and micro-motion features into the meta-classifier to obtain the target recognition result on the sea surface. The meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on structural feature samples, and the second base classifier is trained based on micro-motion feature samples.

[0176] To implement the above embodiments, this application also proposes an electronic device, including: a processor and a memory communicatively connected to the processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory to implement the method provided in the foregoing embodiments.

[0177] To implement the above embodiments, this application also proposes a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided in the foregoing embodiments.

[0178] To implement the above embodiments, this application also proposes a computer program product, including a computer program that, when executed by a processor, implements the methods provided in the foregoing embodiments.

[0179] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in this application all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0180] It should be noted that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. Furthermore, such collection / sharing should only be conducted after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes authorization of relevant user information before the user uses the function. In addition, any necessary steps must be taken to protect and safeguard access to such personal information data and ensure that others with access to personal information data comply with their privacy policies and procedures.

[0181] This application is intended to provide an implementation scheme for users to selectively prevent the use or access to their personal information data. Specifically, this disclosure is intended to provide hardware and / or software to prevent or block access to such personal information data. Once personal information data is no longer needed, risks can be minimized by restricting data collection and deleting data. Furthermore, where applicable, such personal information is de-identified to protect user privacy.

[0182] In the foregoing descriptions of the embodiments, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0183] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0184] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0185] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0186] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0187] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0188] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0189] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

[0190] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.

[0191] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for identifying sea surface targets using space-based radar, characterized in that, Includes the following steps: The ship's high-resolution range image (HRRP) matrix is ​​obtained, the HRRP matrix is ​​normalized, and the normalized HRRP matrix is ​​superimposed and accumulated along the azimuth direction. The structural features of the accumulated HRRP are extracted, including the equivalent scattering center dimension, equivalent size, entropy, standard deviation, bias, irregularity, frequency domain incoherent range profile, and central moment features. Micro-motion features were extracted from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively. The structural features and the micro-motion features are input into the meta-classifier to obtain the sea surface target recognition result. The meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on the structural feature samples, and the second base classifier is trained based on the micro-motion feature samples. The extracted structural features of the accumulated HRRP include: For the extracted and accumulated HRRP, i.e. , recorded as Structural features are calculated using the following formulas, including: Calculate the dimension of the equivalent scattering center The mathematical expression is: in, The unit jump function is defined as: Calculate the equivalent size The mathematical expression is: in, Let be the position vector of the range cell in the HRRP signal where the echo intensity is greater than half of the maximum value. Its calculation expression is: Calculate entropy The mathematical expression is: Calculate the standard deviation The mathematical expression is: Calculate the deviation The mathematical expression is: Calculate irregularity The mathematical expression is: Calculate the frequency domain incoherent range image The mathematical expression is: To calculate the central moment characteristic, let the first-order origin moment be: ,but order central moments The mathematical expression is: ； The step of extracting micro-motion features from the accumulated HRRP orientation using Fourier coefficient fitting includes: The range cell containing the strongest scattering point is extracted and its change over time. A Fourier series is used to fit the position of the strong scattering point within the range cell as a function of the ship's attitude angle. The coefficients of the Fourier series are then used as the ship's micro-motion characteristics. The formula for calculating the Fourier series is as follows: in, A function representing the HRRP signal, expressed in terms of time. As the independent variable, For the constant term in the Fourier series, Let be the coefficients of the cosine terms in the Fourier series. Let be the coefficients of the sine term in the Fourier series. The angular frequency of the fundamental frequency. For harmonic order; The extraction of micro-motion features from the accumulated HRRP orientation using short-time Fourier transform includes: Multiple range cells near a strong scattering point are selected, and the changes in multiple range cells are monitored simultaneously using short-time Fourier transform. Let the accumulated value be... HRRP is , For the first The first HRRP The amplitude of each distance cell is calculated by performing a short-time Fourier transform on the signal data in each distance cell, using the following formula: in, It is the selected window function; The multiple short-time Fourier transform matrices calculated from the selected range cells are summed, and the summed matrix is ​​denoted as . For the matrix Singular value decomposition is performed to reduce its feature dimensionality, i.e.: in, Represents the matrix after the short-time Fourier transform; Given an m×k matrix, containing The left singular vector of the matrix; It is a k×k diagonal matrix containing singular values; yes The transpose of a matrix, of size k×n. The matrix contains The right singular vector of the matrix; Based on the singular values ​​obtained from the decomposition, the micro-motion characteristics of the ship are extracted.

2. The method according to claim 1, characterized in that, The process of acquiring the high-resolution sea surface range image (HRRP) matrix of the ship, performing amplitude normalization on the HRRP matrix, and then stacking and accumulating the normalized HRRP matrix along the azimuth direction includes: Obtain the HRRP matrix of the ship, denoted as: in, The HRRP matrix is ​​defined as follows: M is the number of rows in the HRRP matrix, representing the number of time frames; N is the number of columns in the HRRP matrix, representing the number of distance units. In the HRRP matrix, the first... Okay, number A single element of the column, i.e., the first Frame number, the first Signal strength of each distance unit; The HRRP matrix is ​​normalized row-wise, where the first row is... HRRP record The elements in the HRRP matrix are normalized according to the following formula, which is expressed as follows: The normalized HRRP matrices are superimposed along the azimuth direction, and the expression is: in, This represents the accumulated HRRP.

3. A space-based radar sea surface target identification device based on the method of any one of claims 1-2, characterized in that, include: The preprocessing module is used to obtain the high-resolution range image (HRRP) matrix of the ship at sea surface, perform amplitude normalization on the HRRP matrix, and then stack and accumulate the normalized HRRP matrix along the azimuth direction. The structural feature extraction module is used to extract the structural features of the accumulated HRRP, wherein the structural features include the equivalent scattering center dimension, equivalent size, entropy, standard deviation, bias, irregularity, frequency domain incoherent range profile, and central moment features; The micro-motion feature extraction module is used to extract micro-motion features from the accumulated HRRP orientation using short-time Fourier transform and Fourier coefficient fitting, respectively. The target recognition module is used to input the structural features and the micro-motion features into a meta-classifier to obtain the target recognition result on the sea surface. The meta-classifier is composed of a first base classifier and a second base classifier stacked together. The first base classifier is trained based on the structural feature samples, and the second base classifier is trained based on the micro-motion feature samples.

4. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-2.

5. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-2.

6. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1-2.

Images