Target echo signal detection method and device, computer equipment, computer readable storage medium and computer program product

Abstract

The invention relates to a target echo signal detection method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring an original echo signal, wherein the original echo signal comprises a noise signal and a target echo signal; performing frequency domain matched filtering processing and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result; mapping the signal filtering result to a high-dimensional feature space, and performing dichotomy according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, the target hyperplane is determined according to the distance from sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane; and generating a detection result of the target echo signal according to the signal category label. By adopting the method, the target echo signal can be accurately identified and detected in a complex noise environment.

Description

Technical Field

[0001] The present application relates to the field of signal processing technology, and in particular to a method, device, computer equipment, computer-readable storage medium and computer program product for detecting a target echo signal. Background Art

[0002] In complex marine environments, due to the influence of underwater environmental noise, active sonar often receives target echo signals mixed with a lot of noise and reverberation. Therefore, how to detect target echo signals in signals mixed with noise and reverberation has become a current research hotspot.

[0003] In the prior art, a frequency domain adaptive filter can be used to match filter the input signal received by the active sonar based on the frequency domain characteristics of the target echo signal, thereby obtaining the detection result of the target echo signal. However, due to the large fluctuation of the background noise of the signal processed by the frequency domain adaptive filter, it is easy to cause inaccurate detection results of the target echo signal. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for detecting target echo signals with high detection accuracy in complex background noise in order to solve the above technical problems.

[0005] In a first aspect, the present application provides a method for detecting a target echo signal, comprising:

[0006] Acquire an original echo signal, wherein the original echo signal includes a noise signal and a target echo signal;

[0007] Performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result;

[0008] Mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, wherein the target hyperplane is determined according to the distances from sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane;

[0009] A detection result of the target echo signal is generated according to the signal category label.

[0010] In one embodiment, performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result includes:

[0011] Inputting the original echo signal into a matched filter, and using the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal;

[0012] Using the matched filter to perform matched filtering on the frequency domain signal to obtain a preliminary filtering result, wherein the preliminary filtering result includes a single frequency component and a noise component;

[0013] Adopting the matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristic to obtain a target line spectrum signal, wherein the target frequency domain characteristic is obtained by performing Fourier transform on the frequency domain samples of the transmission signal corresponding to the target echo signal;

[0014] The matched filter is used to perform inverse Fourier transform processing on the target line spectrum signal to obtain the signal filtering result in the time domain.

[0015] In one embodiment, the step of using the matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristics to obtain the target line spectrum signal includes:

[0016] Using the matched filter to multiply the initial weight coefficient with the preliminary filtering result to obtain a processed enhanced line spectrum signal;

[0017] Determining a minimum mean square error corresponding to the initial weight coefficient using the matched filter according to the target frequency domain characteristics and the enhanced line spectrum signal;

[0018] The matched filter is used based on the minimum mean square error criterion, and the initial weight coefficient is dynamically adjusted according to the minimum mean square error to obtain the target weight coefficient;

[0019] The target line spectrum signal is obtained by multiplying the target weight coefficient with the preliminary filtering result using the matched filter.

[0020] In one embodiment, the using the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal includes:

[0021] Using the signal weight corresponding to the preset direction, weighted processing is performed on the original echo signal received by each array element in the multi-element array to obtain a processed weighted echo signal;

[0022] Superimposing the weighted echo signals corresponding to each of the array elements to obtain the target beam signal corresponding to the multi-element array;

[0023] The target beam signal is subjected to Fourier transform processing by using the matched filter to obtain the frequency domain signal.

[0024] In one embodiment, the method further comprises:

[0025] constructing a range-azimuth spectrum corresponding to the target beam signal;

[0026] According to the signal category label of the signal filtering result, a target detection position corresponding to the target echo signal is located in the range-azimuth spectrum.

[0027] In one embodiment, mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result includes:

[0028] Normalizing the signal filtering result to obtain a normalized signal filtering result;

[0029] Mapping the normalized signal filtering result to the high-dimensional feature space using a linear kernel function to obtain a corresponding linearly separable data set;

[0030] According to the positional relationship between the linearly separable data set and the target hyperplane, the linearly separable data set is binary-classified to obtain the signal category label.

[0031] In a second aspect, the present application also provides a detection device for a target echo signal, comprising:

[0032] A signal acquisition module, used to acquire an original echo signal, wherein the original echo signal includes a noise signal and a target echo signal;

[0033] A matched filtering module, used for performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result;

[0034] A signal classification module, used for mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, wherein the target hyperplane is determined according to the distances from sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane;

[0035] The result generating module is used to generate the detection result of the target echo signal according to the signal category label.

[0036] In a third aspect, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method for detecting a target echo signal as described in any one of the embodiments of the first aspect is implemented.

[0037] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for detecting a target echo signal described in any one of the embodiments of the first aspect above is implemented.

[0038] In a fifth aspect, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the method for detecting a target echo signal described in any one of the embodiments of the first aspect is implemented.

[0039] The above-mentioned target echo signal detection method, device, computer equipment, computer-readable storage medium and computer program product obtain the original echo signal, perform frequency domain matching filtering and adaptive line spectrum enhancement processing on the original echo signal, map the signal filtering result obtained after processing to a high-dimensional feature space, and perform binary classification on the signal filtering result according to the target hyperplane in the high-dimensional feature space, generate the detection result of the target echo signal based on the signal category label obtained by the binary classification, improve the signal-to-noise ratio of the original echo signal through frequency domain matching filtering and adaptive line spectrum enhancement processing, and improve the accuracy of the detection result through a binary classification algorithm based on a support vector machine, thereby realizing accurate recognition and detection of target echo signals in complex noise environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0041] FIG1 is a diagram showing an application environment of a method for detecting a target echo signal in one embodiment;

[0042] FIG2 is a schematic flow chart of a method for detecting a target echo signal in one embodiment;

[0043] Figure 3 A schematic flow chart of a signal filtering result generating step in one embodiment;

[0044] Figure 4 It is a schematic flow chart of a target line spectrum signal generation step in one embodiment;

[0045] Figure 5 is a schematic flow chart of a beamforming step in an embodiment;

[0046] Figure 6 A schematic diagram of a flow chart of a signal category label generation step in one embodiment;

[0047] Figure 7 is a schematic flow chart of a method for detecting a target echo signal in another embodiment;

[0048] Figure 8 is an algorithm flow chart of a matched filter in one embodiment;

[0049] Fig. 9A A range-azimuth spectrum diagram based on conventional signal processing results in one embodiment;

[0050] Fig. 9B A range-azimuth spectrum diagram of a result of an adaptive line spectrum enhancement technique based on matched filtering in one embodiment;

[0051] Fig. 10A is a range-azimuth spectrum diagram based on a constant false alarm algorithm in one embodiment;

[0052] Fig. 10B A distance-azimuth spectrum graph based on a support vector machine in one embodiment;

[0053] Fig.11 is a structural block diagram of a target echo signal detection device 1100 in one embodiment;

[0054] Fig.12 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0055] 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.

[0056] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0057] The target echo signal detection method provided in the embodiment of the present application can be applied in the application environment shown in FIG. 1 .

[0058] The sonar detection device 102 communicates with the server 104 through a network. During operation, the sonar detection device 102 can transmit a signal to an object to be detected, receive an echo signal corresponding to the transmitted signal returned from the object, and generate a corresponding detection result according to the received echo signal. However, since the echo signal is easily affected by underwater environmental noise during underwater propagation, the original echo signal received by the sonar detection device 102 may be mixed with noise signals and target echo signals.

[0059] The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers. The server 104 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0060] Specifically, the server 104 can obtain the original echo signal sent by the sonar detection device 102. Since the target echo signal is generated based on the transmission signal, the server 104 can construct a matched filter to obtain the target echo signal based on the adaptive line spectrum enhancement technology of the frequency domain matched filter. The server 104 can use the matched filter corresponding to the target echo signal to perform frequency domain matched filtering on the original echo signal, and then perform adaptive line spectrum enhancement on the signal after the frequency domain matched filtering process to obtain a signal filtering result corresponding to the original echo signal. The server 104 can store a target hyperplane determined according to the distance from the sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane in the high-dimensional feature space. The signal filtering result is mapped to the high-dimensional feature space, and the signal filtering result is classified into two categories according to the target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result. The server 104 can generate a detection result of the target echo signal according to the signal category label.

[0061] In this embodiment, by using a server to perform matched filtering processing on the original echo signal received by the sonar detection equipment, the signal-to-noise ratio of the echo signal can be improved. The signal filtering result obtained by the server based on the matched filtering processing is binary classified in a high-dimensional feature space to separate the noise signal and the target echo signal, which can improve the detection accuracy of the target echo signal in a complex noise environment.

[0062] In an exemplary embodiment, as shown in FIG2 , a method for detecting a target echo signal is provided, which is described by taking the method applied to the server 104 in FIG1 as an example, and includes the following steps S202 to S208 . Among them:

[0063] Step S202, obtaining the original echo signal.

[0064] Among them, the original echo signal can be used to characterize the signal directly received by the sonar detection device. The original echo signal includes a noise signal and a target echo signal. The noise signal may include underwater environmental noise, the noise of the sonar detection device itself, etc. The target echo signal can be used to characterize the echo signal that matches the transmission signal of the sonar detection device.

[0065] Exemplarily, an echo signal processing program may be run in the server. The server may receive a data signal sent by the sonar exploration device through a network. The echo signal processing program analyzes and processes the data signal to obtain an original echo signal sent by the sonar detection device in real time.

[0066] Step S204: Perform frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result.

[0067] The matched filter can be used to characterize a linear filter with the largest ratio between the instantaneous power of a signal at the output and the average power of noise.

[0068] Exemplarily, the server may store an adaptive line spectrum enhancement algorithm based on frequency domain matched filtering and a transmission signal corresponding to the sonar detection device and the target echo signal. Optionally, in some embodiments, the adaptive line spectrum enhancement algorithm based on frequency domain matched filtering may be used to first perform matched filtering processing on the original echo signal, and then based on the frequency domain characteristics of the transmission signal after matched filtering (which can be obtained by Fourier transforming the frequency domain samples of the transmission signal), the line spectrum enhancement processing is performed on the original echo signal after matched filtering in the frequency domain, thereby obtaining a signal filtering result corresponding to the original echo signal.

[0069] Step S206, mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to the target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result.

[0070] The target hyperplane may be determined based on the distances from the sample data points of the target echo signal samples and the noise signal samples in the high-dimensional feature space to the same hyperplane. For example, the target hyperplane may be an optimal hyperplane that maximizes the distances from the sample data points of the target echo signal samples and the noise signal samples in the high-dimensional feature space to the hyperplane.

[0071] The signal category label may include a target echo signal category and a noise signal category.

[0072] Exemplarily, the server may pre-process sample data points mapped with target echo signal samples and noise signal samples in a high-dimensional feature space to construct a target hyperplane of sample data points for distinguishing target echo signal categories and noise signal categories.

[0073] The server can map the signal filtering result corresponding to the original echo signal from the current low-dimensional data space to the high-dimensional feature space to obtain the mapping position of the signal filtering result in the high-dimensional feature space. The signal filtering result is classified into two categories according to the spatial relationship between the mapping position of the signal filtering result and the target hyperplane, and the result obtained by the two-category classification is used as the signal category label corresponding to the signal filtering result.

[0074] Step S208: generating a detection result of the target echo signal according to the signal category label.

[0075] Optionally, in some embodiments, when the signal category tag includes the target echo signal category, a detection result including the target echo signal in the signal filtering result may be generated. Alternatively, in other embodiments, when the signal category tag does not include the target echo signal category, a detection result not including the target echo signal in the signal filtering result may be generated.

[0076] In the above-mentioned target echo signal detection method, by acquiring the original echo signal, the original echo signal is subjected to frequency domain matched filtering processing and adaptive line spectrum enhancement processing, the signal filtering result obtained after processing is mapped to a high-dimensional feature space, and the signal filtering result is binary classified according to the target hyperplane in the high-dimensional feature space, and the detection result of the target echo signal is generated based on the signal category label obtained by the binary classification. The signal-to-noise ratio of the original echo signal can be improved through frequency domain matched filtering and adaptive line spectrum enhancement processing, and the accuracy of the detection result can be improved through a binary classification algorithm based on a support vector machine, thereby realizing accurate recognition and detection of target echo signals in complex noise environments.

[0077] In an exemplary embodiment, Figure 3 As shown, the above step S204 may also include the following steps S302 to S308. Among them:

[0078] Step S302: input the original echo signal into a matched filter, and use the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal.

[0079] The matched filter can be used to represent an optimal linear filter, which is designed to maximize the output signal-to-noise ratio when receiving a signal, thereby improving the detection performance of the system. Exemplarily, in this embodiment, a linear filter with a filter coefficient of a default initial value can be used as the matched filter.

[0080] Since the time complexity of the Fourier transform of the original echo signal in the time domain is relatively high, the server may transform the original echo signal into the frequency domain before performing the subsequent processing operations of steps S304 to S308. Exemplarily, an initialized matched filter may be stored in the server. The original echo signal is input into the matched filter, and the matched filter is used to perform Fourier transform processing on the original echo signal to transform the original echo signal in the time domain into signal data in the frequency domain, that is, to obtain a frequency domain signal corresponding to the original echo signal.

[0081] Step S304: Use a matched filter to perform matched filtering on the frequency domain signal to obtain a preliminary filtering result.

[0082] The preliminary filtering result may include a single frequency component and a noise component. The single frequency component may be used to characterize the signal component corresponding to the target echo signal. The noise component may be used to characterize the signal component corresponding to the noise signal.

[0083] For example, the server may use a matched filter to multiply the frequency domain signal using a transfer function corresponding to its own filter coefficient, and use the multiplication result obtained after processing as the frequency domain signal after matched filtering. The transfer function may be a mathematical expression used to describe the relationship between the filter output signal and the input signal, which reflects the response characteristics of the filter to signals of different frequencies.

[0084] Step S306: Adopt a matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristics to obtain a target line spectrum signal.

[0085] The target frequency domain characteristic may be obtained by performing Fourier transform on the frequency domain samples of the transmission signal corresponding to the target echo signal.

[0086] Exemplarily, since the single-frequency component in the preliminary filtering result obtained in step S304 is easily submerged in the noise component, the server can use a matched filter to process the preliminary filtering result using an adaptive line spectrum enhancement algorithm (such as Adaptive Line Enhancer, referred to as ALE, an adaptive line spectrum enhancer algorithm, which is used to suppress irrelevant components of its input signal and attenuate narrowband components with very small amplitude) to enhance the single-frequency component corresponding to the target echo signal. The server can perform Fourier transform processing on the frequency domain samples of the transmitted signal in advance, use the signal data results obtained after the frequency domain sample processing as the target frequency domain characteristics of the transmitted signal after matched filtering, and store the target frequency domain characteristics in the matched filter. The matched filter is used to perform adaptive line spectrum enhancement processing on the preliminary filtering result obtained by the matched filter in the frequency domain based on the target frequency domain characteristics, and obtain the target line spectrum signal with the single frequency component in the preliminary filtering result.

[0087] Step S308: Use a matched filter to perform inverse Fourier transform processing on the target line spectrum signal to obtain a signal filtering result in the time domain.

[0088] Exemplarily, the server uses a matched filter to perform inverse Fourier transform processing on the target line spectrum signal in the frequency domain. For example, a fast inverse Fourier transform algorithm can be used to perform calculations on the target line spectrum signal to transform the target line spectrum signal in the frequency domain into signal data in the time domain, that is, to obtain the signal filtering result in the time domain output by the matched filter.

[0089] In this embodiment, the original echo signal is processed by using an adaptive line spectrum enhancement algorithm based on frequency domain matched filtering to obtain the corresponding signal filtering result, which can not only greatly improve the signal-to-noise ratio of the signal output after matched filter processing, but also simplify the computational time complexity in the matched filtering process and improve the generation efficiency of the signal filtering result. At the same time, since the single-frequency component corresponding to the target echo signal in the preliminary filtering result is easily submerged in the noise component corresponding to the noise signal, by using the adaptive line spectrum enhancement method in this embodiment, the signal-to-noise ratio of the matched filter output signal can be improved, and noise and other interference can be removed or reduced.

[0090] In an exemplary embodiment, Figure 4 As shown, the above step S306 may also include the following steps S402 to S408. Among them:

[0091] Step S402: using a matched filter to multiply the initial weight coefficient with the preliminary filtering result to obtain a processed enhanced line spectrum signal.

[0092] Step S404: using a matched filter to determine the minimum mean square error corresponding to the initial weight coefficients according to the target frequency domain characteristics and the enhanced line spectrum signal.

[0093] The weight coefficient of the matched filter, that is, the conjugate of the signal spectrum, can be realized by the transfer function of the matched filter. The initial weight coefficient can be used to represent the default value of the initialization configuration of the matched filter.

[0094] Exemplarily, the server can use a matched filter to multiply the initial weight coefficient stored in the server with the preliminary filtering result, and use the signal data result obtained by the multiplication as the enhanced line spectrum signal. The signal output expectation of the matched filter is determined according to the target frequency domain characteristics. The enhanced line spectrum signal obtained by processing the signal output expectation and the initial weight coefficient is processed to generate the minimum mean square error corresponding to the initial weight coefficient.

[0095] Step S406: using a matched filter based on a minimum mean square error criterion, dynamically adjusting the initial weight coefficient according to the minimum mean square error to obtain a target weight coefficient.

[0096] Step S408: using a matched filter to multiply the target weight coefficient with the preliminary filtering result to obtain a target line spectrum signal.

[0097] Among them, the minimum mean square error criterion, also known as the Minimum Mean-Square Error, referred to as the MMSE criterion, is a data rule that uses the minimum value of the minimum mean square error output by the filter or estimator to measure the index of the performance of the filter or estimator. In this embodiment, when the difference between the output obtained after the matched filter performs adaptive line spectrum enhancement processing and the signal output expectation is smaller, the minimum mean square error value obtained is also smaller, indicating that the performance of the weight coefficient currently used by the matched filter is better. Conversely, when the minimum mean square error is larger, it indicates that the performance of the weight coefficient currently used by the matched filter is worse.

[0098] Exemplarily, the server can dynamically adjust the initial weight coefficient used by the matched filter according to the minimum mean square error calculated in step S404 using the matched filter, for example, by increasing or decreasing the initial weight coefficient, and calculating the minimum mean square error corresponding to the adjusted weight coefficient in accordance with the method of steps S402 to S404. Based on the minimum mean square error criterion, the weight coefficient that can minimize the minimum mean square error numerically is determined as the target weight coefficient. The target weight coefficient is multiplied by the preliminary filtering result using the matched filter to obtain the target line spectrum signal.

[0099] In this embodiment, the weight coefficient is calculated by an adaptive method, and the weight coefficient is dynamically adjusted based on the minimum mean square error criterion to enhance the single-frequency signal component, thereby increasing the signal-to-noise ratio of the target line spectrum signal.

[0100] In an exemplary embodiment, Figure 5 As shown, the above step S302 may also include the following steps S502 to S506. Among them:

[0101] Step S502: using the signal weight corresponding to the preset direction, weighted processing is performed on the original echo signal received by each array element in the multi-element array to obtain a processed weighted echo signal.

[0102] The preset direction may be used to indicate the incident direction of the original echo signal.

[0103] A multi-element array can be used to represent a structure composed of several array elements (also called transducers) arranged in a certain pattern. The purpose of a multi-element array is to obtain a desired target beam pattern in space.

[0104] An array element can be used to represent a single receiver or sensor that makes up a sonar array in a sonar detection device. An array element can be an antenna, a microphone, a hydroacoustic sensor, etc. An array element can be used to receive a signal from a signal source.

[0105] Exemplarily, the server may store signal weights corresponding to preset directions, for example, the signal weight in the preset direction is 1, and the signal weight in the non-preset direction is 0.5, or the corresponding signal weight is determined according to the angle between the non-preset direction and the preset direction. The signal weight corresponding to the preset direction is used to perform weighted operation processing on the original echo signal received by each array element in the multi-element array of the sonar detection device to obtain a processed weighted echo signal corresponding to each array element.

[0106] Step S504: superimpose the weighted echo signals corresponding to each array element to obtain the target beam signal corresponding to the multi-element array.

[0107] Step S506: Use a matched filter to perform Fourier transform processing on the target beam signal to obtain a frequency domain signal.

[0108] Exemplarily, the server may superimpose the weighted echo signals corresponding to each array element to form a target beam signal corresponding to the multi-element array, and perform Fourier transform processing on the target beam signal using a matched filter corresponding to the target echo signal to obtain a frequency domain signal corresponding to the original beam signal in the frequency domain.

[0109] In this embodiment, a beamforming algorithm based on a multi-element array is used to adjust the amplitude and phase corresponding to each array element to enhance the signal in a specific direction in the original echo signal and eliminate or suppress interference and noise in other directions. This can improve the signal-to-noise ratio of the target beam signal and estimate the incident azimuth of the original echo signal.

[0110] In an exemplary embodiment, Figure 5 Based on the embodiment provided in, the target echo signal detection method in the present application may also include:

[0111] A range-azimuth spectrum corresponding to the target beam signal is constructed; and a target detection position corresponding to the target echo signal is located in the range-azimuth spectrum according to the signal category label of the signal filtering result.

[0112] The range-azimuth spectrum can be used to describe the location of the object to be detected. The distance is the distance between the object to be detected and the sonar detection device. The azimuth is the azimuth of the object to be detected relative to the sonar detection device.

[0113] Exemplarily, the server can calculate the distance between the object to be detected and the sonar detection device based on the delay time of the original echo signal relative to the transmitted signal. According to the signal strength of the array element of the target beam signal in each direction, the azimuth of the object to be detected and the sonar detection device is determined. Using the azimuth and distance, a distance-azimuth spectrum corresponding to the target beam signal is constructed. According to the signal category label obtained by binary classification of the signal filtering result in the high-dimensional feature space, it is determined whether there is a signal under the target echo signal category in the signal filtering result. Or, the spatial position of the signal under the target echo signal category in the signal filtering result relative to the sonar detection device. Thereby achieving the technical effect of locating the target detection position corresponding to the target echo signal in the distance-azimuth spectrum.

[0114] In this embodiment, by providing a visualized range-azimuth spectrum, the target echo signal can be located, and the spatial position of the target echo signal can be intuitively displayed, thereby facilitating subsequent further research and processing of the target echo signal.

[0115] In an exemplary embodiment, Figure 6 As shown, step S206 may include the following steps S602 to S606. Among them:

[0116] Step S602, normalizing the signal filtering result to obtain a normalized signal filtering result.

[0117] Step S604: Use a linear kernel function to map the normalized signal filtering result to a high-dimensional feature space to obtain a corresponding linearly separable data set.

[0118] Exemplarily, in the data preprocessing stage of the signal filtering result, the server can normalize the signal filtering result to obtain a normalized signal filtering result. The normalized signal filtering result is processed by using a pre-stored linear kernel function to obtain a linearly separable data set obtained by mapping the normalized signal filtering result to a high-dimensional feature space.

[0119] Step S606, performing binary classification on the linearly separable data set according to the positional relationship between the linearly separable data set and the target hyperplane to obtain a signal category label.

[0120] Exemplarily, the server may perform binary classification on the linearly separable data set according to the positional relationship between the linearly separable data set and the target hyperplane. For example, when the data points in the linearly separable data set are distributed above the target hyperplane, the binary classification result corresponding to the linearly separable data set may be determined to be the target echo signal category. Otherwise, the binary classification result corresponding to the linearly separable data set may be determined to be the noise signal category.

[0121] In this embodiment, by normalizing the signal filtering result, it can help improve the signal classification efficiency. By using a linear kernel function to map the low-dimensional signal filtering result to a high-dimensional feature space, the problem that the characteristic vectors of the target echo signal and the noise signal cannot be linearly separated can be solved. By using a binary classification algorithm based on a support vector machine to determine the signal category to which the signal filtering result belongs, the target detection problem can be converted into a classification problem, reducing the computational complexity. At the same time, in this embodiment, the linear kernel function is used for mapping, which can also improve the mapping efficiency of linearly separable data sets.

[0122] In an exemplary embodiment, Figure 7 As shown, a method for detecting a target echo signal is also provided, comprising the following steps S702 to S710. Among them:

[0123] Step S702: acquiring the original echo signal received by each array element in the multi-element array.

[0124] Step S704: beamforming is performed on the original echo signal of each array element to obtain a target beam signal corresponding to the multi-element array.

[0125] Exemplarily, the server may use the signal weight corresponding to the preset direction to perform weighted processing on the original echo signal received by each array element, so as to enhance the signal in the specific preset direction, and eliminate or suppress the interference and noise in other directions, thereby obtaining the processed weighted echo signal. The weighted echo signal corresponding to each array element is superimposed to obtain the target beam signal corresponding to the multi-element array, thereby improving the output signal-to-noise ratio of the target beam signal, and at the same time, the incident azimuth of the target beam signal can be estimated.

[0126] Step S706, processing the target beam signal based on an adaptive line spectrum enhancement algorithm of frequency domain matched filtering to obtain a signal filtering result output by the matched filter.

[0127] For example, Figure 8 As shown in FIG. 1 , an algorithm flow chart of an adaptive line spectrum enhancement technology based on frequency domain matched filtering is provided. The server can Figure 8 The matched filter shown performs fast Fourier transform processing on the target beam signal in the time domain to obtain the corresponding frequency domain signal in the frequency domain. The frequency domain signal is subjected to matched filtering processing by the matched filter to obtain a preliminary filtering result, and then the preliminary filtering result is subjected to adaptive line spectrum enhancement processing to obtain the corresponding target line spectrum signal. The target line spectrum signal in the frequency domain is subjected to inverse fast Fourier transform processing to obtain the corresponding signal filtering result in the time domain.

[0128] Optionally, in some implementations, the target beam signal input to the matched filter is as follows:

[0129]

[0130] in, Represents the output after beamforming (spatial filtering), that is, the target beam signal described in this application. represents the beamforming weight vector, represents the steering vector, Represents the data received by the mth array element, represents the Gaussian white noise received by the array.

[0131] when When: The above target beam signal can be simplified to:

[0132]

[0133] Where M represents the number of array elements, and the meanings of the other parameters are as defined above.

[0134] Here we assume that the noise observed by each array element is random, spatially and temporally uncorrelated, with a mean of 0 and a variance of Gaussian white noise. Since the amplitude of the echo signal and the noise signal only affects the signal-to-noise ratio of the output signal. Therefore, after ignoring the influence of the amplitude, the above first frequency domain signal can be further simplified as:

[0135]

[0136] in, is the first frequency domain signal input into the matched filter. is the target echo signal corresponding to the transmitted signal, and its amplitude and phase are unknown. is a band-limited white noise with a mean of 0.

[0137] In the frequency domain, the signal can be matched filtered by using a matched filter according to the following formula:

[0138]

[0139] in, is the second frequency domain signal. is the transfer function of the matched filter. is the first frequency domain signal. is the spectrum function of the target echo signal. is the spectrum function of the noise signal.

[0140] Set at the preset time ( ), target echo signal The instantaneous power signal-to-noise ratio is the largest, then:

[0141]

[0142] in, is the transfer function of the matched filter. k is an arbitrary constant, which can generally be 1. are Fourier transform coefficients. express The complex conjugate of .

[0143] So further:

[0144]

[0145] in, is the second frequency domain signal. k is an arbitrary constant. is the spectrum function of the target echo signal. are Fourier transform coefficients. is the spectrum function of the noise signal. The meanings of the other parameters are as given above.

[0146] From the above formula, we can see that the result of the original echo signal after being processed by the matched filter in the frequency domain is: the second frequency domain signal It includes two parts. One of these two parts is , which can be regarded as the target echo signal component containing a single frequency component. The other part is , which can be regarded as the sum of band-limited white noise. The single-frequency component of the first part is submerged in the noise of the second part, and the detection performance of the matched filter can be further improved by means of line spectrum enhancement.

[0147] Alternatively, if Fig. 9AAs shown in FIG. 1 , a distance-azimuth spectrum diagram of a conventional signal processing result is provided, where the vertical axis represents its normalized amplitude, the horizontal axis represents its distance, and the vertical axis represents the azimuth angle. Fig. 9B The figure shows a distance-azimuth spectrum diagram of the adaptive line spectrum enhancement technology based on matched filtering. The vertical axis represents its normalized amplitude, the horizontal axis represents its distance, and the vertical axis represents its azimuth. After comparison, it can be clearly seen that Fig. 9B The results obtained by using the adaptive line spectrum enhancement technology based on matched filtering provided in this embodiment are compared with Fig. 9A As a result, the signal-to-noise ratio of the echo signal is significantly improved.

[0148] Step S708: Use a linear kernel function to map the normalized filtering results to a high-dimensional feature space, and perform binary classification according to the target hyperplane of the high-dimensional feature space to obtain a corresponding signal category label.

[0149] Alternatively, in some embodiments, the linear kernel function may be as follows:

[0150]

[0151] in, It can be expressed using vector and vector The kernel function is composed of the inner product of .

[0152] Optionally, in some implementations, the target hyperplane may be determined by:

[0153]

[0154]

[0155] in, is the target hyperplane. According to the "maximum interval" principle (that is, the principle that the distance between data points of different categories and the hyperplane is the largest), the parameters are determined. and . is the slack variable, is a penalty parameter used to balance the maximization of the margin and the minimization of the error. By solving and To determine the optimal target hyperplane, and then realize the separation of noise signal and target echo signal.

[0156] Step S710, generating a detection result of the target echo signal according to the signal category label, and locating the target detection position corresponding to the target echo signal in the range-azimuth spectrum corresponding to the target beam signal.

[0157] Alternatively, if Fig. 10AFIG. 1 shows a schematic diagram of detecting a target echo signal from a target beam signal on a range-azimuth spectrum using a constant false alarm algorithm. Fig. 10B FIG. 2 shows a result diagram of target echo signal detection on the range-azimuth spectrum using the support vector machine provided in this embodiment. By comparison, it can be seen that Fig. 10B The positioning of the target echo signal achieved by the technical means of this embodiment has significantly less interference and noise than Fig. 10A The detection result of the target echo signal obtained by using the constant false alarm algorithm.

[0158] In this embodiment, by adopting the beamforming technology of the multi-element array, the original echo signal is converted into the corresponding target beam signal, the signal in a specific preset direction can be enhanced, the interference and noise in other directions can be eliminated or suppressed, and the output signal-to-noise ratio of the signal can be improved. By adopting the frequency domain adaptive matched filtering technology for matched filtering and adaptive line spectrum enhancement, the signal-to-noise ratio of the signal filtering result can be improved. By adopting the binary classification method based on the support vector machine to determine the signal category label, and then obtaining the detection result of the target echo signal, the low-dimensional data can be mapped to the high-dimensional feature space, and the non-linearly separable target echo signal and noise signal can be converted into a linearly separable data set, thereby realizing the accurate detection and positioning of the target echo signal, and at the same time, the time complexity of the target echo signal detection can be reduced.

[0159] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0160] Based on the same inventive concept, the embodiment of the present application also provides a target echo signal detection device for implementing the target echo signal detection method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more target echo signal detection devices provided below can refer to the limitations of the target echo signal detection method above, and will not be repeated here.

[0161] In an exemplary embodiment, Fig.11As shown, a target echo signal detection device 1100 is provided, comprising: a signal acquisition module 1102, a matched filtering module 1104, a signal classification module 1106 and a result generation module 1108, wherein:

[0162] The signal acquisition module 1102 is used to acquire an original echo signal, where the original echo signal includes a noise signal and a target echo signal.

[0163] The matched filtering module 1104 is used to perform frequency domain matched filtering processing and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result.

[0164] The signal classification module 1106 is used to map the signal filtering results to a high-dimensional feature space, and perform binary classification according to the target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering results. The target hyperplane is determined based on the distance from the sample data points of the target echo signal samples and the noise signal samples in the high-dimensional feature space to the same hyperplane.

[0165] The result generating module 1108 is used to generate the detection result of the target echo signal according to the signal category label.

[0166] In an exemplary embodiment, the matched filtering module 1104 includes a frequency domain transformation unit, which is used to input the original echo signal into a matched filter, and use the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal; a frequency domain filtering unit, which is used to use the matched filter to perform matched filtering processing on the frequency domain signal to obtain a preliminary filtering result, and the preliminary filtering result includes a single frequency component and a noise component; a line spectrum enhancement unit, which is used to use the matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristics to obtain a target line spectrum signal, and the target frequency domain characteristics are obtained by Fourier transforming the frequency domain samples of the transmitted signal corresponding to the target echo signal; a time domain transformation unit, which is used to use the matched filter to perform inverse Fourier transform processing on the target line spectrum signal to obtain a signal filtering result in the time domain.

[0167] In an exemplary embodiment, the line spectrum enhancement unit is also used to use a matched filter to multiply the initial weight coefficient with the preliminary filtering result to obtain a processed enhanced line spectrum signal; use a matched filter to determine the minimum mean square error corresponding to the initial weight coefficient based on the target frequency domain characteristics and the enhanced line spectrum signal; use a matched filter based on the minimum mean square error criterion to dynamically adjust the initial weight coefficient according to the minimum mean square error to obtain the target weight coefficient; use a matched filter to multiply the preliminary filtering result with the target weight coefficient to obtain the target line spectrum signal.

[0168] In an exemplary embodiment, the frequency domain transformation unit is further used to perform weighted processing on the original echo signal received by each array element in the multi-element array using a signal weight corresponding to a preset direction to obtain a processed weighted echo signal; superimpose the weighted echo signal corresponding to each array element to obtain a target beam signal corresponding to the multi-element array; and perform Fourier transform processing on the target beam signal using a matched filter to obtain a frequency domain signal.

[0169] In an exemplary embodiment, the target echo signal detection device 1100 also includes a target positioning module for constructing a range-azimuth spectrum corresponding to the target beam signal; and locating the target detection position corresponding to the target echo signal in the range-azimuth spectrum according to the signal category label of the signal filtering result.

[0170] In an exemplary embodiment, the signal classification module 1106 is also used to normalize the signal filtering results to obtain normalized signal filtering results; use a linear kernel function to map the normalized signal filtering results to a high-dimensional feature space to obtain a corresponding linearly separable data set; and classify the linearly separable data set into two categories based on the positional relationship between the linearly separable data set and the target hyperplane to obtain a signal category label.

[0171] Each module in the above-mentioned target echo signal detection device 1100 can be implemented in whole or in part by software, hardware and a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0172] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.12As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as original echo signals, matching filters, signal filtering results, target hyperplanes, signal category labels, and detection results. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for detecting a target echo signal is implemented.

[0173] Those skilled in the art will understand that Fig.12 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0174] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0175] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0176] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0177] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method 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 the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0178] 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 application.

[0179] The above-described 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 present application. 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 these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for detecting a target echo signal, characterized in that: The method comprises: Acquire an original echo signal, wherein the original echo signal includes a noise signal and a target echo signal; Performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result; Mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, wherein the target hyperplane is determined according to the distances from sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane; A detection result of the target echo signal is generated according to the signal category label.

2. The method according to claim 1, characterized in that The performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result includes: Inputting the original echo signal into a matched filter, and using the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal; Using the matched filter to perform matched filtering on the frequency domain signal to obtain a preliminary filtering result, wherein the preliminary filtering result includes a single frequency component and a noise component; Adopting the matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristic to obtain a target line spectrum signal, wherein the target frequency domain characteristic is obtained by performing Fourier transform on the frequency domain samples of the transmission signal corresponding to the target echo signal; The matched filter is used to perform inverse Fourier transform processing on the target line spectrum signal to obtain the signal filtering result in the time domain.

3. The method according to claim 2, characterized in that The step of using the matched filter to perform adaptive line spectrum enhancement processing on the single frequency component in the preliminary filtering result based on the target frequency domain characteristics to obtain the target line spectrum signal includes: Using the matched filter to multiply the initial weight coefficient with the preliminary filtering result to obtain a processed enhanced line spectrum signal; Determining a minimum mean square error corresponding to the initial weight coefficient using the matched filter according to the target frequency domain characteristics and the enhanced line spectrum signal; The matched filter is used based on the minimum mean square error criterion, and the initial weight coefficient is dynamically adjusted according to the minimum mean square error to obtain the target weight coefficient; The target line spectrum signal is obtained by multiplying the target weight coefficient with the preliminary filtering result using the matched filter.

4. The method according to claim 2, characterized in that: The using the matched filter to perform Fourier transform processing on the original echo signal to obtain a frequency domain signal corresponding to the original echo signal includes: Using the signal weight corresponding to the preset direction, weighted processing is performed on the original echo signal received by each array element in the multi-element array to obtain a processed weighted echo signal; Superimposing the weighted echo signals corresponding to each of the array elements to obtain the target beam signal corresponding to the multi-element array; The target beam signal is subjected to Fourier transform processing by using the matched filter to obtain the frequency domain signal.

5. The method according to claim 4, characterized in that The method further comprises: constructing a range-azimuth spectrum corresponding to the target beam signal; According to the signal category label of the signal filtering result, a target detection position corresponding to the target echo signal is located in the range-azimuth spectrum.

6. The method according to claim 1, characterized in that Mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, includes: Normalizing the signal filtering result to obtain a normalized signal filtering result; Mapping the normalized signal filtering result to the high-dimensional feature space using a linear kernel function to obtain a corresponding linearly separable data set; According to the positional relationship between the linearly separable data set and the target hyperplane, the linearly separable data set is binary-classified to obtain the signal category label.

7. A device for detecting a target echo signal, characterized in that: The device comprises: A signal acquisition module, used to acquire an original echo signal, wherein the original echo signal includes a noise signal and a target echo signal; A matched filtering module, used for performing frequency domain matched filtering and adaptive line spectrum enhancement processing on the original echo signal to obtain a corresponding signal filtering result; A signal classification module, used for mapping the signal filtering result to a high-dimensional feature space, and performing binary classification according to a target hyperplane in the high-dimensional feature space to obtain a signal category label corresponding to the signal filtering result, wherein the target hyperplane is determined according to the distances from sample data points of the target echo signal sample and the noise signal sample in the high-dimensional feature space to the same hyperplane; The result generating module is used to generate the detection result of the target echo signal according to the signal category label.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. 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 method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images