Reverberation suppression method and device suitable for underwater linear target detection

Abstract

The invention relates to the technical field of reverberation suppression suitable for underwater linear target detection, and discloses a reverberation suppression method and device suitable for underwater linear target detection, and the method comprises the steps: transmitting a detection signal through a preset high-frequency limited band model, and receiving an echo signal; obtaining a first peak point according to the echo signal value, taking the first peak point as a starting point, and selecting a plurality of sampling points to form a plurality of signal windows; performing fractional Fourier transform on each window signal to obtain a u-domain signal; performing narrowband filtering on the u-domain signal, and extracting a target signal; and performing inverse transformation on the target u-domain signal to obtain a suppressed target signal. The method has the beneficial effects that the method does not depend on array aperture and system bandwidth, is suitable for data processing of a signal domain, does not need training data in an algorithm implementation process, is small in operand, is easy to implement, can effectively focus on a spectral range of a target signal, and remarkably improves the extraction precision of the target signal.

Description

Technical Field

[0001] The present invention relates to the technical field of reverberation suppression applicable to underwater linear target detection, and in particular to a reverberation suppression method and device applicable to underwater linear target detection. Background Art

[0002] With the continuous advancement of technology, autonomous underwater vehicles (AUVs) have been widely used in underwater exploration and surveying. Due to their maneuverability and ease of operation, AUVs are capable of performing tasks in complex underwater environments, such as underwater target search, underwater terrain survey, and marine environmental monitoring. Side-scan sonar, as a key underwater imaging technology, can acquire high-resolution images of underwater targets in real time, thus playing a vital role in underwater target detection.

[0003] Side-scan sonar detects underwater objects by emitting sound waves and receiving their reflected echoes. Its imaging quality is affected by a variety of factors, particularly when searching for submerged targets. Seabed reverberation is a major factor interfering with target detection. Seabed reverberation is caused by the interaction of sound waves with the seabed, sediment, and other materials, resulting in reflected sound waves and spurious noise in the echo signal. In complex ocean environments, changes in seabed topography and water conditions can cause variations in the reverberation signal, impacting the quality of the sonar image. Summary of the Invention

[0004] Based on this, it is necessary to address the existing reverberation suppression problem applicable to underwater linear target detection and propose a reverberation suppression method and device applicable to underwater linear target detection.

[0005] A reverberation suppression method suitable for underwater linear target detection, the method comprising:

[0006] Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0007] Traversing the numerical values of the echo signals in chronological order to obtain the first peak point;

[0008] Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0009] Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain;

[0010] performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0011] An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0012] Furthermore, before the step of traversing the numerical values of the echo signals in chronological order to obtain the first peak point, the method further includes:

[0013] Select an averaging window Ω of length K = 2M + 1 k =(kM,k-M+1,…,k+M-1,k+M; where K is the length and Ωk is the averaging window;

[0014] According to the formula Find the mean of the signal; where is the signal mean, X i represents the signal value at the i-th time point within the averaging window;

[0015] According to the formula Reassign the signal value in the averaging window; where Y i represents the value assigned to the signal at the i-th time point within the averaging window, and r is the preset threshold coefficient;

[0016] According to the formula and formula Calculate the target signal value as the subsequent echo signal calculation; where μ i Indicates the median value corresponding to the signal value at the i-th time point in the averaging window, N i Indicates the target signal value corresponding to the signal value at the i-th time point in the averaging window.

[0017] Furthermore, the step of performing fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain includes:

[0018] By formula Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain; where p is the fractional Fourier transform order, F p is the operator symbol of fractional Fourier transform, x a (t) represents the signal value at the tth time point in the ath window signal, u represents the u domain, X ap (u) represents the u-domain signal corresponding to the a-th window signal, K p (t,u) is the transformation kernel function, and j is the representation of the complex signal, u is the frequency point in the u-domain, and t is the time point.

[0019] Furthermore, in performing the fractional Fourier transform, the fractional Fourier transform order is preferably Where k is the frequency modulation of the detection signal, and s is the dimension normalization coefficient.

[0020] Furthermore, the step of performing narrowband filtering on each of the U-domain signals to obtain a corresponding target U-domain signal includes:

[0021] Get the template signal s(t);

[0022] By formula Get the narrowband filter window; where S p (u) is the narrowband filter window;

[0023] According to the formula Z ap (u)=S p (u)×X ap (u) narrow-band filtering is performed on the u-domain signal to obtain the corresponding target u-domain signal, where Z ap (u) represents the target u-domain signal corresponding to the a-th window signal.

[0024] Furthermore, after the steps of taking the first peak point as a starting point, sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining multiple windows, and obtaining a window signal corresponding to each window, the method further includes:

[0025] Other signals in the echo signal that are not within the window are set to zero.

[0026] Furthermore, the preset high-frequency band-limited model is a sonar bright spot model.

[0027] A reverberation suppression device suitable for underwater linear target detection, the device comprising:

[0028] A transmitting module is used to transmit a detection signal through a preset high-frequency band-limited model and receive an echo signal of a preset duration after the detection signal is transmitted;

[0029] An acquisition module, configured to traverse the numerical values of the echo signal in chronological order to obtain a first peak point;

[0030] a selection module, configured to use the first peak point as a starting point, and sequentially select a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0031] A transformation module is used to perform a fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain;

[0032] a filtering module, configured to perform narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0033] The inverse transform module is used to perform an inverse transform of the opposite order of the fractional Fourier transform on each target u-domain signal to obtain the target signal after each window signal is suppressed.

[0034] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0035] Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0036] Traversing the numerical values of the echo signals in chronological order to obtain the first peak point;

[0037] Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0038] Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain;

[0039] performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0040] An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0041] A computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform the following steps:

[0042] Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0043] Traversing the numerical values of the echo signals in chronological order to obtain the first peak point;

[0044] Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0045] Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain;

[0046] performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0047] An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0048] The beneficial effects of the present invention are as follows: by adopting a preset high-frequency band-limited model to transmit a detection signal, selecting a suitable window in the echo signal for processing, applying fractional Fourier transform, converting the signal into the u domain, and further extracting the spectral characteristics of the target signal through narrowband filtering, it is independent of the array aperture and system bandwidth, suitable for data processing in the signal domain, the algorithm implementation process does not require training data, has a small amount of computation, is easy to implement, can effectively focus on the spectral range of the target signal, and significantly improves the extraction accuracy of the target signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0050] in:

[0051] Figure 1 FIG2 is a diagram showing an application environment of a reverberation suppression method suitable for underwater linear target detection in one embodiment;

[0052] Figure 2 is a flow chart of a reverberation suppression method applicable to underwater linear target detection in one embodiment;

[0053] Figure 3 1 is a structural block diagram of a reverberation suppression device suitable for underwater linear target detection in one embodiment;

[0054] Figure 4 FIG. 1 is a structural block diagram of a computer device in one embodiment. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] Figure 1FIG. 1 is a diagram of an application environment for reverberation suppression suitable for underwater linear target detection in one embodiment. Figure 1 This reverberation suppression method for underwater linear target detection is applied to a reverberation suppression system for underwater linear target detection. This reverberation suppression system for underwater linear target detection includes a terminal 110 and a server 120. Terminal 110 and server 120 are connected via a network. Terminal 110 can be a desktop terminal or a mobile terminal. The mobile terminal can be at least one of a mobile phone, tablet computer, and laptop computer. Server 120 can be implemented as a standalone server or a server cluster consisting of multiple servers. Terminal 110 is used to acquire echo signals, and server 120 is used to analyze the echo signals.

[0057] For bottom targets, when using side-scan sonar for detection, the relative speed between the sonar and the target is small, and the influence of Doppler can be ignored. According to the bright spot model, in the case of high frequencies, the echo is composed of the superposition of several sub-echoes, and each sub-echo can be regarded as a wave emitted from a certain scattering point. Since the influence of Doppler is ignored, when the transmitted signal is an LFM signal (Linear Frequency Modulation, linear frequency modulation signal), the echo signal is also an LFM signal with the same frequency modulation, but the amplitude of the signal has changed. The scattering units are evenly distributed within the directivity angle coverage range of the receiving array. Each unit contains a large number of scatterers. Each scatterer will generate a backscattered wave for the incident sound signal. The received signal is actually the superposition of the target's reflected signal and the reflected signals of other scatterers, so the detection of the target is easily masked by the bottom reverberation. The linear target can specifically be a submarine cable.

[0058] like Figure 2 As shown, in one embodiment, a reverberation suppression method suitable for underwater linear target detection is provided. This method can be applied to both a terminal and a server. This embodiment uses the application to a terminal as an example. The reverberation suppression method suitable for underwater linear target detection specifically includes the following steps:

[0059] S1: transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0060] S2: traverse the numerical values of the echo signal in chronological order to obtain the first peak point;

[0061] S3: taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining multiple windows, and acquiring a window signal corresponding to each window;

[0062] S4: Perform fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain;

[0063] S5: performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0064] S6: Perform an inverse transform of an order opposite to the fractional Fourier transform on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0065] As described in step S1 above, a detection signal is transmitted through a preset high-frequency band-limited model, and an echo signal of a preset duration is received after the detection signal is transmitted. The detection signal is transmitted through a preset high-frequency band-limited model, and the frequency range of the detection signal is within a specific high-frequency band so that it can penetrate the underwater medium and effectively detect the target. After transmitting the signal, the underwater unmanned vehicle will receive an echo signal of a preset duration. The echo signal is the sound wave signal reflected back by the target and the seabed terrain, which contains the target's location information and the noise information of the seabed environment.

[0066] As described in step S2 above, the numerical values of the echo signal are traversed in time sequence to obtain the first peak point. After receiving the echo signal, the numerical values of the echo signal are traversed in time sequence to identify the first peak point. This peak point usually represents a strong reflection in the echo signal, generally corresponding to the initial reflection receiving point of the target. Since the target echo position is unknown, a sliding window signal interception method can be adopted. When the intercepted signal does not contain or only contains a part of the target echo, the peak signal corresponding to the u0 position of the u domain is the scattered signal of a scattering unit on the seabed that reaches the receiving point at this time. At this time, the signal filtered and recovered at the u0 position is the scattered signal of a scattering unit on the seabed (in practice, it may be a signal of the in-phase superposition of the scattered signals of some scattering units); when the intercepted signal just completely contains the target echo signal, the recovered signal is the target echo signal (it may also contain the seabed scattered signal superimposed in phase with the target echo), thereby removing the interference of the seabed reverberation. In order to intercept a complete cycle of the LFM signal (Linear Frequency Modulation) reflected by the target, the starting point of the peak signal is used as a marker. When the starting point of a signal processing window coincides with the starting point of the signal, the signal containing the target echo is just intercepted. Whether the starting point of the target echo signal can be selected exactly is crucial for the processing of the algorithm. Among them, the peak point is the local maximum in the signal, that is, the adjacent signal value is smaller than the value of the point. In the echo signal, the first peak point usually represents the first significant change in the signal, corresponding to the echo reflection of the target object. In the time-sequential traversal, when the value of the current point is greater than the value of the previous point and greater than the value of the next point, the point is the peak point. If the echo signal is monotonically rising or falling, the program will continue to traverse until the first local maximum is found.

[0067] As described in step S3 above, the first peak point is used as the starting point, and a preset number of sampling points are sequentially selected based on the starting point to form a window, thereby obtaining multiple windows and acquiring the window signal corresponding to each window. The first maximum point is used as the starting point, and a preset number of sampling points are sequentially selected based on the starting point to form a window. In this way, the entire echo signal can be divided into multiple windows, each corresponding to a relatively small signal segment. The signal in each window represents the sonar echo information within that time period, allowing analysis to be performed using the time domain signal as the smallest unit.

[0068] As described in step S4 above, each of the u-domain signals is narrowband filtered to obtain the corresponding target u-domain signal. The fractional Fourier transform is an extended fractional Fourier transform method that can perform more refined time-frequency analysis of the signal and better extract the features of different frequency bands. Through this step, the u-domain signal of each window signal in the u-domain can be obtained, where the u-domain signal can reflect the characteristics of the echo signal in the frequency domain and lay the foundation for subsequent signal processing.

[0069] As described in step S5 above, each of the u-domain signals is narrowband filtered to obtain a corresponding target u-domain signal. Narrowband filters can effectively remove spectral components of non-target signals, particularly interfering signals such as seafloor reverberation. By precisely adjusting the filter bandwidth, components within the target signal frequency range can be retained while suppressing other irrelevant frequency components. The filtered signal is the target u-domain signal, which retains the target's spectral characteristics while removing most of the noise and reverberation components.

[0070] As described in step S6 above, an inverse transform of the opposite order of the fractional Fourier transform is performed on each target u-domain signal to obtain the target signal after each window signal is suppressed. The inverse Fourier transform restores the signal from the frequency domain to the time domain and reconstructs a clearer and more accurate target signal. This helps further reduce the impact of seabed reverberation and other interference on the signal, effectively removing seabed reverberation and noise, improving the quality of the target signal, and thus enhancing the detection accuracy and reliability of underwater linear targets.

[0071] In one embodiment, before the step S2 of traversing the numerical values of the echo signals in chronological order to obtain the first peak point, the method further includes:

[0072] S101: Select an averaging window Ω with a length of K=2M+1 k =(kM,k-M+1,…,k+M-1,k+M; where K is the length and Ωk is the averaging window;

[0073] S102: According to the formula Find the mean of the signal; where is the signal mean, X i represents the signal value at the i-th time point within the averaging window;

[0074] S103: According to the formula Reassign the signal value in the averaging window; where Y i represents the value assigned to the signal at the i-th time point within the averaging window, and r is the preset threshold coefficient;

[0075] S104: According to the formula and formula Calculate the target signal value as the subsequent echo signal calculation; where μ i Indicates the median value corresponding to the signal value at the i-th time point in the averaging window, N i Indicates the target signal value corresponding to the signal value at the i-th time point in the averaging window.

[0076] As described in steps S101-S104 above, background equalization is achieved. Due to the presence of strong bottom reverberation, the amplitude of the seabed reverberation is often greater than the amplitude of the target echo, affecting the selection of the starting point. In this solution, background equalization is used to process the complete echo. The specific processing steps are as follows:

[0077] In step S101, an averaging window Ω of length K=2M+1 is selected. k =(kM,k-M+1,…,k+M-1,k+M); where K is the length and Ωk is the averaging window. That is, an averaging window with a length of K=2M+1 is pre-defined, and equalization is performed in units of the averaging window.

[0078] In steps S102-S104, the signal mean is calculated, and the value is reassigned based on the signal mean. Finally, the weight coefficient is calculated, and the original signal is divided by the weight coefficient as the final echo signal, thereby achieving the effect of background equalization. Among them, the value of M can be set by yourself. It should not be set too large, otherwise it will cause the local background to be consistent. It should not be too small, otherwise the background equalization effect cannot be achieved. The threshold coefficient is preferably set to Experiments have shown that selecting the above value as the threshold coefficient can achieve better background equalization effect.

[0079] In one embodiment, the step S4 of performing fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain includes:

[0080] S401: By formula Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain; where p is the fractional Fourier transform order, F p is the operator symbol of fractional Fourier transform, x a (t) represents the signal value at the tth time point in the ath window signal, u represents the u domain, X ap (u) represents the u-domain signal corresponding to the a-th window signal, K p (t,u) is the transformation kernel function, and j is the representation of the complex signal, u is the frequency point in the u-domain, and t is the time point.

[0081] In one embodiment, when performing fractional Fourier transform, the fractional Fourier transform order is preferably Where k is the frequency modulation of the detection signal, s is the dimensional normalization factor, which refers to a constant introduced during dimensional analysis in order to make the dimensions of different physical quantities comparable or unified in calculations, arctan is the inverse tangent function, and π is the pi.

[0082] As described in step S401 above, the time domain signal is converted into a signal in the u domain to facilitate subsequent filtering processing. Specifically, the u domain refers to the frequency component of the signal after transformation, which is similar to the frequency domain in the traditional fractional Fourier transform, but it is more flexible and can be used to describe different time-frequency characteristics of the signal. The key idea of the fractional Fourier transform is to introduce an "order" (p) to control the degree of transformation between the time domain and the frequency domain. This "order" determines the characteristics of the signal representation after the transformation. Specifically, when p = 0, the signal is completely in the time domain, when p = 1, the signal is completely in the frequency domain, and the value between the two indicates that the signal is in a domain between time and frequency, namely the u domain. In this way, flexible conversion between time and frequency can be performed according to actual needs, making signal processing more efficient and accurate.

[0083] In one embodiment, the step S5 of performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal includes:

[0084] S501: Acquire a template signal s(t);

[0085] S502: By formula Get the narrowband filter window; where S p (u) is the narrowband filter window;

[0086] S503: According to formula Z ap (u)=S p (u)×X ap (u) narrow-band filtering is performed on the u-domain signal to obtain the corresponding target u-domain signal, where Z ap (u) represents the target u-domain signal corresponding to the a-th window signal.

[0087] As described in steps S501-S503 above, narrowband filtering of the u-domain signal is achieved. The template signal refers to a known signal template under a specific environment or under an expected target. It is usually used to help identify the target signal, especially to extract target features from spurious signals (such as noise, reverberation, etc.). The template signal can be extracted from existing data or obtained through experiments or simulations. In underwater detection, the template signal can be a known reflection echo from the target or an idealized signal (for example, an echo of a specific frequency, duration or specific shape). The signal has specific spectral characteristics in the u-domain and can be used to help filter out the target signal in subsequent processing steps. The generation of the narrowband filter window is designed based on the spectral characteristics of the template signal. Specifically, the spectrum of the template signal contains the frequency components of the target signal, and the narrowband filter window limits its spectral range to the frequency band of the target signal. By filtering the u-domain signal through the narrowband filter window, the frequency components containing the target signal can be obtained. The main goal of this step is to remove frequency components that are not related to the target signal (such as seabed reverberation, noise, etc.) and retain the frequency characteristics of the target signal. After narrowband filtering, the resulting signal is the target u-domain signal. This signal represents the frequency characteristics of the target and removes interference and noise, making it easier to process and analyze later.

[0088] In one embodiment, after the step S3 of taking the first peak point as a starting point and sequentially selecting a preset number of sampling points based on the starting point as a window to obtain multiple windows and obtaining a window signal corresponding to each window, the method further includes:

[0089] S411: Setting other signals in the echo signal that are not within the window to zero.

[0090] The portion of the echo signal that does not fall within the current window range is set to 0. Specifically, for the signal range corresponding to each window, all other signals not within the window (whether in the time domain or frequency domain) are set to 0. By eliminating signals outside the window, background noise and reverberation components unrelated to the target can be effectively eliminated. The signal within the window is retained and further processed, while the signal outside the window is ignored, allowing the focus to be on accurately analyzing and processing the target signal.

[0091] In one embodiment, the preset high-frequency band-limited model is a sonar bright spot model.

[0092] The sonar brightspot model describes the echo characteristics of sonar targets by equating them to a set of spatially distributed brightspots. This model, based on the fact that most sonar systems use high-frequency, band-limited signals, describes the target's echo characteristics through a transfer function. At high frequencies, the various components of the echo, such as specular reflections, angular waves, and various elastically scattered waves, can be equated to the echo from a single scattering center, known as a brightspot.

[0093] Reference Figure 3 The present invention provides a reverberation suppression device suitable for underwater linear target detection, comprising:

[0094] The transmitting module 10 is used to transmit a detection signal through a preset high-frequency band-limited model and receive an echo signal of a preset duration after the detection signal is transmitted;

[0095] An acquisition module 20 is configured to traverse the numerical values of the echo signal in chronological order to obtain a first peak point;

[0096] A selection module 30 is configured to use the first peak point as a starting point, and sequentially select a preset number of sampling points based on the starting point as a window, thereby obtaining multiple windows, and obtaining a window signal corresponding to each window;

[0097] a transform module 40 for performing a fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain;

[0098] A filtering module 50 is configured to perform narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0099] The inverse transform module 60 is configured to perform an inverse transform of an order opposite to that of the fractional Fourier transform on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0100] In one embodiment, the reverberation suppression device for underwater linear target detection further includes:

[0101] Average window selection module, used to select the average window Ω with a length of K = 2M + 1 k =(kM,k-M+1,…,k+M-1,k+M; where K is the length and Ωk is the averaging window;

[0102] According to the formula Find the mean of the signal; where is the signal mean, X i represents the signal value at the i-th time point within the averaging window;

[0103] Re-assignment module, used to Reassign the signal value in the averaging window; where Y i represents the value assigned to the signal at the i-th time point within the averaging window, and r is the preset threshold coefficient;

[0104] The target signal value calculation module is used to calculate the target signal value according to the formula and formula Calculate the target signal value as the subsequent echo signal calculation; where μ i Indicates the median value corresponding to the signal value at the i-th time point in the averaging window, N i Indicates the target signal value corresponding to the signal value at the i-th time point in the averaging window.

[0105] In one embodiment, the transformation module 40 includes:

[0106] Transformation submodule, used to transform the Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain; where p is the fractional Fourier transform order, F p is the operator symbol of fractional Fourier transform, x a (t) represents the signal value at the tth time point in the ath window signal, u represents the u domain, X ap (u) represents the u-domain signal corresponding to the a-th window signal, K p (t,u) is the transformation kernel function, and j is the representation of the complex signal, u is the frequency point in the u-domain, and t is the time point.

[0107] In one embodiment, when performing fractional Fourier transform, the fractional Fourier transform order is preferably Where k is the frequency modulation of the detection signal, and s is the dimension normalization coefficient.

[0108] In one embodiment, the filtering module 50 includes:

[0109] The template signal acquisition submodule is used to obtain the template signal s(t);

[0110] The narrowband filter window acquisition submodule is used to obtain the formula Get the narrowband filter window; where S p (u) is the narrowband filter window;

[0111] The narrowband filter submodule is used to calculate the Z ap (u)=S p (u)×X ap (u) narrow-band filtering is performed on the u-domain signal to obtain the corresponding target u-domain signal, where Z ap (u) represents the target u-domain signal corresponding to the a-th window signal.

[0112] In one embodiment, the reverberation suppression device for underwater linear target detection further includes:

[0113] The zeroing module is used to set other signals in the echo signal that are not within the window to zero.

[0114] In one embodiment, the preset high-frequency band-limited model is a sonar bright spot model.

[0115] Figure 4 FIG1 shows an internal structure diagram of a computer device in an embodiment. The computer device can be a terminal or a server. Figure 4 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement the reverberation suppression method for underwater linear target detection. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement the reverberation suppression method for underwater linear target detection. It will be understood by those skilled in the art that Figure 4 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 shown in the figure, or combine certain components, or have a different component arrangement.

[0116] In one embodiment, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0117] Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0118] Traversing the numerical values of the echo signals in chronological order to obtain the first peak point;

[0119] Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0120] Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain;

[0121] performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0122] An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0123] By adopting a preset high-frequency band-limited model to transmit the detection signal, selecting a suitable window in the echo signal for processing, applying fractional Fourier transform, converting the signal to the u-domain, and further extracting the spectral characteristics of the target signal through narrowband filtering, the algorithm does not rely on the array aperture and system bandwidth and is suitable for data processing in the signal domain. The algorithm implementation process does not require training data, has a small amount of computation, is easy to implement, can effectively focus on the spectral range of the target signal, and significantly improves the extraction accuracy of the target signal.

[0124] In one embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the processor performs the following steps:

[0125] Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted;

[0126] Traversing the numerical values of the echo signals in chronological order to obtain the first peak point;

[0127] Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window;

[0128] Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain;

[0129] performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal;

[0130] An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

[0131] By adopting a preset high-frequency band-limited model to transmit the detection signal, selecting a suitable window in the echo signal for processing, applying fractional Fourier transform, converting the signal to the u-domain, and further extracting the spectral characteristics of the target signal through narrowband filtering, the algorithm does not rely on the array aperture and system bandwidth and is suitable for data processing in the signal domain. The algorithm implementation process does not require training data, has a small amount of computation, is easy to implement, can effectively focus on the spectral range of the target signal, and significantly improves the extraction accuracy of the target signal.

[0132] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0133] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A reverberation suppression method suitable for underwater linear target detection, characterized in that: The method comprises: Transmitting a detection signal through a preset high-frequency band-limited model, and receiving an echo signal of a preset duration after the detection signal is transmitted; Traversing the numerical values of the echo signals in chronological order to obtain the first peak point; Taking the first peak point as a starting point, and sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window; Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain; performing narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal; An inverse transform of an order opposite to the fractional Fourier transform is performed on each target u-domain signal to obtain a target signal after each window signal is suppressed.

2. The reverberation suppression method for underwater linear target detection according to claim 1, characterized in that: Before the step of traversing the numerical values of the echo signals in time order to obtain the first peak point, the method further includes: Select an averaging window Ω of length K = 2M + 1 k =(kM,k-M+1,…,k+M-1,k+M; where K is the length and Ωk is the averaging window; According to the formula Find the mean of the signal; where is the signal mean, X i represents the signal value at the i-th time point within the averaging window; According to the formula Reassign the signal value in the averaging window; where Y i represents the value assigned to the signal at the i-th time point within the averaging window, and r is the preset threshold coefficient; According to the formula and formula Calculate the target signal value as the subsequent echo signal calculation; where μ i Indicates the median value corresponding to the signal value at the i-th time point in the averaging window, N i Indicates the target signal value corresponding to the signal value at the i-th time point in the averaging window.

3. The reverberation suppression method for underwater linear target detection according to claim 1, characterized in that: The step of performing fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain includes: By formula Perform fractional Fourier transform on each window signal to obtain the u-domain signal of each window signal in the u-domain; where p is the fractional Fourier transform order, F p is the operator symbol of fractional Fourier transform, x a (t) represents the signal value at the tth time point in the ath window signal, u represents the u domain, X ap (u) represents the u-domain signal corresponding to the a-th window signal, K p (t,u) is the transformation kernel function, and j is the representation of the complex signal, u is the frequency point in the u-domain, and t is the time point.

4. The reverberation suppression method for underwater linear target detection according to claim 3, characterized in that: In performing fractional Fourier transform, the fractional Fourier transform order is preferably Where k is the frequency modulation of the detection signal, and s is the dimension normalization coefficient.

5. The reverberation suppression method for underwater linear target detection according to claim 3 or 4, characterized in that: The step of performing narrowband filtering on each of the U-domain signals to obtain a corresponding target U-domain signal includes: Get the template signal s(t); By formula Get the narrowband filter window; where S p (u) is the narrowband filter window; According to the formula Z ap (u)=S p (u)×X ap (u) narrow-band filtering is performed on the u-domain signal to obtain the corresponding target u-domain signal, where Z ap (u) represents the target u-domain signal corresponding to the a-th window signal.

6. The reverberation suppression method for underwater linear target detection according to claim 1, characterized in that: After the steps of taking the first peak point as a starting point, sequentially selecting a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window, the method further includes: Other signals in the echo signal that are not within the window are set to zero.

7. The reverberation suppression method for underwater linear target detection according to claim 1, characterized in that: The preset high-frequency band-limited model is a sonar bright spot model.

8. A reverberation suppression device suitable for underwater linear target detection, characterized in that: The device comprises: A transmitting module is used to transmit a detection signal through a preset high-frequency band-limited model and receive an echo signal of a preset duration after the detection signal is transmitted; An acquisition module, configured to traverse the numerical values of the echo signal in chronological order to obtain a first peak point; a selection module, configured to use the first peak point as a starting point, and sequentially select a preset number of sampling points based on the starting point as a window, thereby obtaining a plurality of windows, and obtaining a window signal corresponding to each window; A transformation module is used to perform a fractional Fourier transform on each window signal to obtain a u-domain signal of each window signal in the u-domain; a filtering module, configured to perform narrowband filtering on each of the u-domain signals to obtain a corresponding target u-domain signal; The inverse transform module is used to perform an inverse transform of the opposite order of the fractional Fourier transform on each target u-domain signal to obtain the target signal after each window signal is suppressed.

9. A computer-readable storage medium, characterized in that A computer program is stored, and when the computer program is executed by a processor, the processor is caused to perform the steps of the reverberation suppression method applicable to underwater linear target detection as claimed in any one of claims 1 to 7.

10. A computer device, characterized in that: The device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the reverberation suppression method applicable to underwater linear target detection as described in any one of claims 1 to 7.

Images