Ground target identification method based on deep learning

Abstract

The invention belongs to the technical field of radar recognition, and particularly relates to a ground target recognition method based on deep learning. The method comprises the following steps: firstly, preprocessing HRRP data to eliminate amplitude sensitivity and translation sensitivity of the HRRP data; secondly, performing time-frequency analysis on HRRP data by using CWT, and extracting time-frequency characteristics of signals under different scales and enhancing robustness of a model to noise through multi-scale decomposition of the CWT; and inputting the processed time-frequency data into the convolutional neural network. And training the CNN by using a large number of marked HRRP data sets. In the training process, the network weight is continuously adjusted through a back propagation algorithm, and the network performance is optimized. And after the training is completed, the trained CNN model is used to classify and identify the test HRRP signal, and the performance of the test HRRP signal is evaluated. The method provided by the invention can describe the HRRP signal more comprehensively, improves the feature expression ability, and has a certain recognition performance advantage under a low signal-to-noise ratio.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar recognition, and in particular relates to a ground target recognition method based on deep learning. Background Art

[0002] With the development of radar technology, radar automatic target recognition (RATR) has become increasingly important in military and civilian fields. As an important way for radar systems to obtain target information, high-resolution range profile (HRRP) has attracted widespread attention because it can provide rich and detailed features of the target. HRRP generates a one-dimensional range profile of the target by analyzing the reflected signals of the target scattering center at different distance units. These profiles contain important information such as the size, shape and scattering characteristics of the target.

[0003] Traditional RATR methods rely on manual feature extraction, but these methods often require a lot of prior knowledge and parameter adjustment, and are difficult to adapt to practical problems such as target posture changes and noise interference. In recent years, deep learning technology has made revolutionary progress in image recognition, speech processing and other fields with its powerful data processing and feature extraction capabilities. The application of deep learning methods in RATR can be divided into three categories. The first category is based on autoencoders, which extract effective features by compressing input data for recognition. At the same time, autoencoders can be used to address the influence of noise in non-cooperative target recognition. However, their generalization ability is limited by the feature extraction of HRRP data. The second category is convolutional neural networks (CNNs), which are naturally introduced into HRRP target recognition due to their excellent performance in image processing, showing great potential for processing high-dimensional radar data, but they are difficult to extract global features due to limited receptive fields. The third category is time series models, such as recurrent neural networks (RNNs), long short-term memory networks (LSTMs), and transformers, which treat HRRP as sequence data. However, HRRP reflects the scattering distribution, so it is more appropriate to treat it as a one-dimensional profile rather than sequential data.

[0004] Most of the existing deep learning-based RATR methods directly process HRRP, ignoring the time-frequency characteristics of radar signals. In the latest research, people mainly apply signal time-frequency analysis to extract the micro-motion features of targets. The literature Target recognition algorithm based on HRRP time-spectrogram feature and multi-scale asymmetric convolutional neural network and Convolutional neural networks for radar HRRP target recognition and rejection use short-time Fourier transform to analyze the time-frequency information of HRRP. However, since the time resolution and frequency resolution obtained by short-time Fourier transform are restricted by the length of the window function, it cannot provide high time resolution and high frequency resolution at the same time, which limits its performance in analyzing non-stationary signals. The literature Towards Efficient Lesion Localization Based on Template Occlusion Strategy in Intelligent Diagnosis proposes an autoencoder with RNN, which uses discrete wavelet transform to automatically learn the hierarchical representation of signals. This method successfully improves the feature extraction ability of signals by utilizing the multi-scale characteristics of DWT. However, although DWT has certain advantages in processing signals, it has certain limitations in the continuity and adaptability of time-frequency analysis. Summary of the invention

[0005] In view of the above problems, the present invention proposes a method for classifying and identifying high-resolution one-dimensional range images based on deep learning, namely CWT-CNN. The method first preprocesses the HRRP data to eliminate its amplitude sensitivity and translation sensitivity; then uses CWT to perform time-frequency analysis on the HRRP data. Through the multi-scale decomposition of CWT, the time-frequency characteristics of the signal at different scales can be extracted and the robustness of the model to noise can be enhanced; then the time-frequency data obtained after processing is input into the convolutional neural network. CNN is trained using a large number of annotated HRRP data sets. During the training process, the network weights are continuously adjusted through the back propagation algorithm to optimize the network performance. After the training is completed, the trained CNN model is used to classify and identify the test HRRP signal, and its performance is evaluated. The proposed method can describe the HRRP signal more comprehensively, improve the expression ability of the features, and has certain recognition performance advantages under low signal-to-noise ratio.

[0006] The technical solution adopted by the present invention is:

[0007] A ground target recognition method based on deep learning includes the following steps:

[0008] S1. Obtain the radar’s high-resolution range image HRRP, expressed as r = [r1, r2, …, r L ] T , where r i represents the sub-echo of the ith range unit, and L represents the total number of range units; the amplitude of the complex HRRP data is taken to obtain the real HRRP data x = |r| = [x1, x2, ..., x l ] T ;

[0009] S2. Preprocessing the obtained real HRRP data, including:

[0010] Use L2 norm normalization:

[0011]

[0012] Calculate the centroid position of the HRRP signal:

[0013]

[0014] The center of the normalized data is shifted to the obtained centroid, and the shifted data is used as training data;

[0015] S3, constructing a target recognition model, including a wavelet transform module, a convolution module, a pooling and full connection module, and an output module;

[0016] The wavelet transform module performs wavelet transform on the training data to obtain time-frequency data X;

[0017] The convolution module extracts features from the time-frequency data X. The convolution module includes two convolution layers, each of which is composed of a convolution layer with a kernel size of 3×3. There is a batch normalization layer and an activation function after the convolution layer. The convolution module modeling is expressed as:

[0018]

[0019] Among them, w ijl is the weight of each position of the convolution kernel, x ijl is the value of each position in the corresponding convolution area, (k1, k2, d) is the convolution kernel size, b is the corresponding offset, the superscript k represents the kth dimension of the data, μ (k) and σ (k) are the mean and standard deviation of the current batch of input data, and the parameter γ (k) and β (k)It is used to scale and shift the normalized value, ∈ is a small number to ensure that the denominator is non-zero, and f(·) is the activation function;

[0020] The pooling and fully connected module includes a pooling layer and two fully connected layers. The pooling layer performs a downsampling operation on the input feature map, and the output of the fully connected layer is expressed as:

[0021]

[0022] in, and b i They are the weight matrix, input data, and offset matrix of the i-th fully connected layer respectively; the first fully connected layer is followed by a Dropout function, and the activation function of the second fully connected layer uses a Sigmoid function;

[0023] The output module adopts softmax classifier for classification;

[0024] S4, using the obtained training data to train the constructed target recognition model to obtain a trained target recognition model;

[0025] S5. After preprocessing, the acquired high-resolution range image of the target is input into the trained target recognition model to obtain a recognition result.

[0026] The beneficial effects of the present invention are as follows: the method of the present invention combines wavelet transform and convolutional neural network to identify ground targets. Continuous wavelet transform can extract the time-frequency information of HRRP data, and can effectively filter out the influence of noise, making the model more robust to noise; at the same time, the multi-scale characteristics of CWT are used to effectively extract local features in HRRP data. It has high recognition accuracy in small sample and high noise scenarios, and has certain advantages under different signal-to-noise ratios and different numbers of sample sets. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a network structure diagram of the present invention.

[0028] Figure 2 Schematic diagram of data preprocessing results.

[0029] Figure 3 Schematic diagram of model comparison under different signal-to-noise ratios. DETAILED DESCRIPTION

[0030] The present invention is described in detail below with reference to the accompanying drawings and embodiments:

[0031] like Figure 1 As shown, the method of the present invention comprises the following steps:

[0032] S1. Obtain the high-resolution range image of the radar r = [r1, r2, ..., r L ] T , where r i represents the sub-echo of the ith range unit, and L represents the total number of range units. The amplitude of the complex HRRP data is taken to obtain the real HRRP data x = |r| = [x1, x2, ..., x l ] T .

[0033] S2. Data preprocessing:

[0034] In view of amplitude sensitivity, this paper uses L2 norm normalization to process the real HRRP data, namely:

[0035]

[0036] The HRRP data is processed using the center of mass alignment to reduce the impact of target posture changes. The center of mass position of the HRRP signal can be calculated using the following formula, and the calculated center of mass is aligned to the center point of the signal.

[0037]

[0038] The HRRP data after L2 normalization and centroid alignment is as follows Figure 2 As shown, the amplitude range of the processed HRRP samples is [0,1], and the centroid is close to the center point.

[0039] S3. Input the recognition model and perform the following processing on the recognition model:

[0040] Wavelet Transform Module

[0041] Continuous wavelet transform provides a method to locate signal features in both time domain and frequency domain by analyzing the signal under different scales and translation parameters. This transform uses a wavelet function as a basis, and convolves with the original signal by changing its scale and position to obtain the time-frequency representation of the signal. The corresponding calculation formula is:

[0042]

[0043] Where x(t) is the original signal; ψ * is the complex conjugate of the wavelet function; a is the scale parameter, which is used to adjust the scale or frequency range of the wavelet function on the time axis; b is the translation parameter.

[0044] The complex Gaussian wavelet is a complex number consisting of the derivatives of the Gaussian function, which can be expressed as the P-order derivative of the Gaussian function, where P is an integer between 1 and 8. The mathematical expression of the complex Gaussian wavelet is as follows:

[0045]

[0046] Among them, C n is a normalization constant that ensures that the wavelet function has unit energy. Compared with other types of wavelets, complex Gaussian wavelets have localized properties in both time and frequency domains, which enables them to provide detailed information about the signal at specific time and frequency.

[0047] Convolutional module

[0048] In order to extract the feature information corresponding to the time-frequency data, a feature extraction module is designed. This module consists of two 2-D convolution blocks, and each convolution layer consists of a convolution layer with a kernel size of 3×3. Set the input data It is multi-channel, assuming that the weight of each position of the convolution kernel is w ijl , the corresponding convolution area has a value of x at each position ijl , the convolution kernel size is (k1, k2, d), then the output value of each convolution area after convolution is

[0049]

[0050] Among them, b is the corresponding offset.

[0051] Batch Normalization Processing Module

[0052] In order to improve the stability of the model, a batch normalization layer is added after the convolution layer. For d-dimensional input data The BN operation is obtained by Equation 6.

[0053]

[0054] Among them, BN is performed independently on each dimension of a batch of data, the superscript k represents the kth dimension of the data, μ (k) and σ (k) are the mean and standard deviation of the current batch of input data. After normalizing the data, use the parameter γ (k) and β (k) Scale and shift the normalized values.

[0055] Activation Function

[0056] Regarding the choice of nonlinear activation function, considering the problem of neuron "death" when using the ReLU function, the LeakyReLU function is selected to avoid the problem of gradient aliasing. At this point, the convolution block can be modeled as:

[0057]

[0058] Among them, f(·) is the activation function.

[0059] The Dropout function is added after the convolutional layer and the first fully connected layer. By adjusting the parameters of the Dropout function, some neurons in the network are randomly disabled.

[0060] Pooling layer and fully connected layer

[0061] The pooling layer extracts key features by downsampling the input feature map, improving the computational efficiency of the model and reducing overfitting, while also suppressing noise and reducing information redundancy. The formula is as follows:

[0062]

[0063] Among them, y ijl Represents the rectangular area related to the kth feature map The maximum pooling output value, x kpq Represents a rectangular area The element at (p,q) in .

[0064] Next, the feature map is passed into the fully connected layer, and the output can be obtained through formula 2-7:

[0065]

[0066] in, and b i are the weight matrix and bias matrix of the ith fully connected layer respectively. The activation function of the second fully connected layer uses the Sigmoid function, and the output layer generally uses the softmax classifier for classification. The loss function is:

[0067]

[0068] Where Y is the output of the network, and the latter is a regular term to prevent overfitting. If the value of the weight decay coefficient λ is too small, it may lead to overfitting, and vice versa.

[0069] S4, using the obtained training data to train the constructed target recognition model to obtain a trained target recognition model;

[0070] S5. After preprocessing, the acquired high-resolution range image of the target is input into the trained target recognition model to obtain a recognition result.

[0071] Example

[0072] The parameters used in this example are: the learning rate of the network model is 0.001, the batch size of the training data is 64, and the batch gradient descent method is used for network training. The technical solution of the present invention is used in sequence:

[0073] 1.According to step "S1", obtain the high-resolution range image r = [r1, r2, ..., r L ] T , where r i represents the sub-echo of the ith range unit, and L represents the total number of range units. The amplitude of the complex HRRP data is taken to obtain the real HRRP data x = |r| = [x1, x2, ..., x l ] T ;

[0074] 2. Perform L2 norm normalization on the real HRRP data obtained in step “S2” to obtain data S;

[0075] 3. According to step "S3", calculate the corresponding center of gravity W in the radar high-resolution image S:

[0076]

[0077] The center of the corresponding radar high-resolution range image S is shifted to the center of gravity W to obtain the moved data x′=[x′1,x′2,…,x′ l ] T

[0078] 4. According to step "S4", the HRRP data is input into the wavelet transform module, the wavelet basis function is set to cgau8, the length is 32, and the time-frequency data X corresponding to the HRRP is obtained, where the size of X is 31×L.

[0079] 5. According to steps “S5-S7”, each convolution layer is composed of a convolution layer with a kernel size of 3×3, and the corresponding output channel numbers are 32 and 64 respectively, the moving stride and the number of zero padding are 1, and the obtained time-frequency data X is input into the convolution module to extract feature information. The dimensions of the corresponding output data are 64×32×31×L and 64×64×31×L respectively;

[0080] 6. Following step “S8”, a Dropout function is added after the convolutional layer and the first fully connected layer, and the deactivation rate of its neurons is set to 0.5.

[0081] 7. According to step "S9", the pooling layer uses the maximum pooling method, the dimension of the pooling domain is 4×4, and the moving stride is 2. By downsampling the input feature map, the key features are extracted, and the corresponding output data Y pool The dimension of Y is 64×64×15×L. The number of neurons in the first fully connected layer is 1024, and the number of neurons in the second fully connected layer is the number of target classification categories. pool Input two fully connected layers and get the output data y=[y1,y2,y i ,…,y l ] T, corresponding to the probability of identifying the HRRP sample as the i-th class.

[0082] like Figure 3 As shown, the present invention has better effect than the existing method.

Claims

1. A ground target recognition method based on deep learning, characterized in that: The following steps are involved: S1. Obtain the radar’s high-resolution range image HRRP, expressed as r = [r1, r2, …, r L ] T , where r i represents the sub-echo of the ith range unit, and L represents the total number of range units; the amplitude of the complex HRRP data is taken to obtain the real HRRP data x = |r| = [x1, x2, ..., x l ] T ; S2. Preprocessing the obtained real HRRP data, including: Use L2 norm normalization: Calculate the centroid position of the HRRP signal: The center of the normalized data is shifted to the obtained centroid, and the shifted data is used as training data; S3, constructing a target recognition model, including a wavelet transform module, a convolution module, a pooling and full connection module, and an output module; The wavelet transform module performs wavelet transform on the training data to obtain time-frequency data X; The convolution module extracts features from the time-frequency data X. The convolution module includes two convolution layers, each of which is composed of a convolution layer with a kernel size of 3×3. There is a batch normalization layer and an activation function after the convolution layer. The convolution module modeling is expressed as: Among them, w ijl is the weight of each position of the convolution kernel, x ijl is the value of each position in the corresponding convolution area, (k1, k2, d) is the convolution kernel size, b is the corresponding offset, the superscript k represents the kth dimension of the data, μ (k) and σ (k) are the mean and standard deviation of the current batch of input data, and the parameter γ (k) and β (k) It is used to scale and shift the normalized value, ∈ is a small number to ensure that the denominator is non-zero, and f(·) is the activation function; The pooling and fully connected module includes a pooling layer and two fully connected layers. The pooling layer performs a downsampling operation on the input feature map, and the output of the fully connected layer is expressed as: in, and b i They are the weight matrix, input data, and offset matrix of the i-th fully connected layer respectively; the first fully connected layer is followed by a Dropout function, and the activation function of the second fully connected layer uses a Sigmoid function; The output module adopts softmax classifier for classification; S4, using the obtained training data to train the constructed target recognition model to obtain a trained target recognition model; S5. After preprocessing, the acquired high-resolution range image of the target is input into the trained target recognition model to obtain a recognition result.

Images