Target line spectrum extraction method and system for improving orientation estimation capability of vector hydrophone

Abstract

The invention provides a target line spectrum extraction method and system for improving the orientation estimation capability of a vector hydrophone. The method comprises the following steps: S1, obtaining vector hydrophone data and carrying out normalization processing; the vector hydrophone data comprises scalar channel data; s2, based on vector hydrophone data, obtaining high-order time interval characteristics, and extracting a target line spectrum; and S3, calculating the spectrum intensity of the target line spectrum to obtain a target orientation. According to the method, the effective spatial texture information of the time-frequency image of the signal is utilized to obtain the high-order time interval characteristic, after the target line spectrum is extracted, the interference line spectrum is eliminated or the spectrum intensity of the interference line spectrum is reduced, then the weighted histogram statistical method is utilized to highlight the effect of the target strong line spectrum, the influence of the interference line spectrum is reduced, and the orientation estimation capability is improved.

Description

Technical Field

[0001] The invention relates to the field of sonar, and in particular to a target line spectrum extraction method and system for improving the azimuth estimation capability of a vector hydrophone. Background Art

[0002] Single vector hydrophone detection has the characteristics of strong concealment, low power, small scale, and flexible application, so it is widely used in small detection platforms. The azimuth estimation of single vector hydrophone refers to the process of directly estimating the spatial azimuth information of the target after a series of processing using the received sound pressure and velocity information. The acoustic energy flow method based on complex sound intensity analysis is a typical example of this type. The implementation of the acoustic energy flow method in the frequency domain is applied to histogram statistics to obtain more comprehensive target azimuth information. The weighted histogram statistics method is suitable for detecting broadband signals containing line spectra, which can highlight the role of strong line spectra. The estimated azimuths of all frequency points are counted in the corresponding azimuth confidence intervals in a weighted manner according to the spectral intensity, and the azimuth estimation curve at a certain moment is obtained. The peak of the curve corresponds to the azimuth estimation value of a target. This method is also called the weighted histogram statistics method. If the spectral intensity of the interference line spectrum is strong, it is easy to generate non-target curve peaks when using the weighted histogram statistics method.

[0003] The usual method to deal with the above problem is to use filtering methods to extract the signal within the target bandwidth and reduce the line spectrum interference outside the bandwidth, or use the integration method to eliminate the non-stationary and irrelevant line spectrum signals in the data. These two methods are used for special processing to achieve the purpose of eliminating the interference line spectrum. These two methods are ineffective for the unknown target signal frequency band and the stable line spectrum in the environment. At present, the single vector hydrophone is affected by the interference line spectrum, which will cause azimuth estimation errors or false alarms. Summary of the invention

[0004] In view of the defects in the prior art, an object of the present invention is to provide a target line spectrum extraction method and system for improving the azimuth estimation capability of a vector hydrophone.

[0005] A target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone provided by the present invention comprises:

[0006] Step S1: acquiring vector hydrophone data and performing normalization processing;

[0007] The vector hydrophone data includes scalar channel data;

[0008] Step S2: based on the vector hydrophone data, obtain high-order time interval characteristics and extract the target line spectrum;

[0009] Step S3: Calculate the spectrum intensity of the target line spectrum to obtain the target direction.

[0010] Preferably, step S1 includes the following sub-steps:

[0011] Step S1.1: Obtain a frame of data from the vector hydrophone, and form a one-dimensional array after Fourier transformation, where each number in the array represents the amplitude of a frequency point; accumulate multiple frames of data arranged in time sequence after Fourier transformation to form a two-dimensional matrix representing time frequency;

[0012] Step S1.2: normalize the two-dimensional matrix.

[0013] Preferably, step S2 includes the following sub-steps:

[0014] Step S2.1: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows, and solve the time gap property for it:

[0015]

[0016] Among them, P t represents the pixel value, E{·} is the data expectation;

[0017] Step S2.2: Use the high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain the high-order time gap feature:

[0018]

[0019] Preferably, if expert experience can be used to extract a segment of data without target line spectrum and containing only background in the current environment sonar image database, a high-order temporal intermittent feature extraction method based on background learning is selected;

[0020] If it is not possible to use expert experience to extract a segment of the current environment sonar image database without threatening targets and only containing data with both interference and targets, then the integral accumulation high-order temporal intermittent feature extraction without prior knowledge is chosen.

[0021] Preferably, the high-order temporal intermittent feature extraction process based on background learning includes:

[0022] Select the mean value U of the background data without a target within d seconds in the database, and m groups of data within d seconds:

[0023]

[0024] Select n groups of data within the time step of t seconds before the current moment, and calculate the difference between them and U for post-processing:

[0025]

[0026] Here, k represents the order.

[0027] Preferably, the integral accumulation high-order temporal intermittent feature extraction process without prior knowledge includes:

[0028] Select the historical mean value U of the data group nm within the time step before the current time s seconds:

[0029]

[0030] Where n is all the data currently stored in the device;

[0031] Select t seconds as the time length, and then calculate the difference between v sets of data within t seconds before the current moment and U for post-processing:

[0032]

[0033] Preferably, step S3 includes calculating the spectral intensity of the signal line spectrum after eliminating the interference line spectrum by high-order time intervals, and using the weighted histogram method to weight the acoustic energy flow according to the spectral intensity. The peak of the curve after fitting is the target direction.

[0034] According to the present invention, a target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone is provided, comprising:

[0035] Module M1: Acquire vector hydrophone data and perform normalization;

[0036] The vector hydrophone data includes scalar channel data;

[0037] Module M2: Based on the vector hydrophone data, obtain the high-order time interval characteristics and extract the target line spectrum;

[0038] Module M3: Calculate the spectrum intensity of the target line spectrum to obtain the target direction.

[0039] Preferably, the module M1 includes the following submodules:

[0040] Module M1.1: Obtain a frame of data from the vector hydrophone, and form a one-dimensional array after Fourier transformation. Each number in the array represents the amplitude of a frequency point; accumulate multiple frames of data arranged in time sequence after Fourier transformation to form a two-dimensional matrix representing time frequency;

[0041] Module M1.2: Normalize the two-dimensional matrix.

[0042] Preferably, the module M2 includes the following submodules:

[0043] Module M2.1: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows, and solve the time gap property for it:

[0044]

[0045] Among them, P t represents the pixel value, E{·} is the data expectation;

[0046] Module M2.2: Use high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain high-order time gap characteristics:

[0047]

[0048] Preferably, if expert experience can be used to extract a segment of data without target line spectrum and containing only background in the current environment sonar image database, a high-order temporal intermittent feature extraction method based on background learning is selected;

[0049] If it is not possible to use expert experience to extract a segment of the current environment sonar image database without threatening targets and only containing data with both interference and targets, then the integral accumulation high-order temporal intermittent feature extraction without prior knowledge is chosen.

[0050] Preferably, the high-order temporal intermittent feature extraction process based on background learning includes:

[0051] Select the mean value U of the background data without a target within d seconds in the database, and m groups of data within d seconds:

[0052]

[0053] Select n groups of data within the time step of t seconds before the current moment, and calculate the difference between them and U for post-processing:

[0054]

[0055] Here, k represents the order.

[0056] Preferably, the integral accumulation high-order temporal intermittent feature extraction process without prior knowledge includes:

[0057] Select the historical mean value U of the data group nm within the time step before the current time s seconds:

[0058]

[0059] Where n is all the data currently stored in the device;

[0060] Select t seconds as the time length, and then calculate the difference between v sets of data within t seconds before the current moment and U for post-processing:

[0061]

[0062] Preferably, the module M3 includes calculating the spectrum intensity of the signal line spectrum after eliminating the interference line spectrum by high-order time interval, and using the weighted histogram method to weight the acoustic energy flow according to the spectrum intensity. The peak of the curve after fitting is the target direction.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1. The present invention utilizes the effective spatial texture information of the time-frequency image of the signal to obtain high-order time interval characteristics. After extracting the target line spectrum, the interference line spectrum is eliminated or the spectrum intensity of the interference line spectrum is reduced. Then, the weighted histogram statistical method is used to highlight the role of the target strong line spectrum, reduce the influence of the interference line spectrum, and improve the azimuth estimation capability.

[0065] 2. The algorithm used in the present invention is simple in structure, does not require any parameters to be set in advance, is flexible and fast, and can eliminate interference line spectra without prior knowledge of the interference signal frequency and the target signal frequency. It can still be extracted when the target line spectrum produces a Doppler frequency shift.

[0066] 3. The present invention can eliminate the stable interference signal line spectrum in the environment that cannot be eliminated in the integral processing, and at the same time can effectively suppress the interference line spectrum, improve the azimuth estimation capability, and reduce the false alarm rate.

[0067] Other beneficial effects of the present invention will be explained in the specific implementation manner through the introduction of specific technical features and technical solutions. Through the introduction of these technical features and technical solutions, those skilled in the art should be able to understand the beneficial technical effects brought about by the technical features and technical solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0069] Figure 1 The present invention is a flow chart of the method.

[0070] Figure 2 It is a processing flow chart of the present invention.

[0071] Figure 3 This is a flow chart of the HOT-Lac feature extraction method based on background learning of the present invention.

[0072] Figure 4 This is a time-frequency diagram containing a target line spectrum and an interference line spectrum without performing HOT-Lac characteristic line spectrum extraction in the present invention.

[0073] Figure 5 This is a schematic diagram of the results of the HOT-Lac characteristic line spectrum extraction method based on background learning in the present invention.

[0074] Figure 6 This is a flow chart of the integral accumulation HOT-Lac feature extraction method without prior knowledge of the present invention.

[0075] Figure 7 This is the orientation estimation diagram of the weighted histogram statistics method before and after the target line spectrum extraction of the present invention. DETAILED DESCRIPTION

[0076] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0077] The problem solved by the present invention is that a single vector hydrophone is affected by interference line spectrum, resulting in azimuth estimation error or false alarm. The basic idea is to use the effective spatial texture information of the time-frequency image of the signal to obtain the high-order time interval (HOT-Lac) feature, extract the target line spectrum, remove the interference line spectrum or reduce the spectrum intensity of the interference line spectrum, and then use the weighted histogram statistical method to highlight the role of the target strong line spectrum and reduce the influence of the interference line spectrum, thereby improving the azimuth estimation capability.

[0078] Interstitiality is a technical term in fractal geometry that provides a measure of how a pattern fills space, reflecting the degree of deviation between the actual pixel value and the expected value, and is defined as follows:

[0079]

[0080] P t : pixel value.

[0081] It can be seen from the formula that the more uneven the pixel distribution of the image is, the larger the value of L is.

[0082] The time-frequency image uses the joint features of time and frequency to characterize the signals collected by these vector hydrophones. There are differences in time or frequency characteristics between the target line spectrum and the interference line spectrum, that is, the target line spectrum and the interference line spectrum have different intensity changes over time, or they have different frequency changes over time. Due to the non-stationary behavior of the target, the change degree of the target line spectrum is often greater. Therefore, the ability to distinguish the target line spectrum and the interference line spectrum can be improved by using the method of high-order central moments. The method of using high-order central moments can better characterize the heavy-tail characteristics of interference noise, increase the inter-class distance between the target line spectrum and the interference line spectrum, and effectively distinguish the target line spectrum from the interference line spectrum. Different orders can extract different inter-class distances between the target line spectrum and the interference line spectrum, and can be applied in different scenarios. Therefore, the use of high-order time gap (HOT-Lac) features can better remove the relatively stable interference line spectrum.

[0083] Reference Figure 1 As shown, a target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone comprises:

[0084] Step 1: Obtain a frame of data from the vector hydrophone and form a one-dimensional array through Fourier transformation. Each number in the array represents the amplitude of a frequency point. Multiple frames of data arranged in time sequence can be accumulated to form a two-dimensional matrix representing time frequency after Fourier transformation. This matrix is ​​the time-frequency diagram of the signal, with the row number representing the frequency on the X-axis and the column number representing the time on the Y-axis. The elements in the array are the amplitudes of each frequency.

[0085] Step 2: Normalize the accumulated matrix representing the time position.

[0086] Step 3: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows. T seconds represents the step length of the time scale sliding on the image. t cannot be too large to prevent the change characteristics of the target line spectrum and the interference line spectrum from not being extracted, while ensuring that t is large enough to ensure that rich target line spectrum information is obtained. The time interval is solved as follows:

[0087]

[0088] Among them, P t represents the pixel value, E{·} is the data expectation;

[0089] Step 4: Use the high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain the high-order time gap (HOT-Lac) feature. As shown in the following formula:

[0090]

[0091] L kThe larger the value, the greater the possibility of having the target line spectrum in the corresponding direction. k The lower the value, the more likely it is that the corresponding direction is an interference line spectrum.

[0092] Step 5: Calculate the spectrum intensity of the signal line spectrum after removing the interference line spectrum by high-order time intervals, and use the weighted histogram method to weight the acoustic energy flow according to the spectrum intensity. The peak of the curve after fitting is the target direction.

[0093] The present invention utilizes the effective spatial texture information of the signal's time-frequency image to obtain high-order temporal intermittent characteristics, and after extracting the target line spectrum, removes the interference line spectrum or reduces the spectral intensity of the interference line spectrum, and then uses the weighted histogram statistics method to highlight the role of the target strong line spectrum, reduce the influence of the interference line spectrum, and improve the azimuth estimation capability.

[0094] The above is a basic embodiment of the present invention. The technical solution of the present invention is further described below through a preferred embodiment.

[0095] Example 1

[0096] Reference Figure 1 As shown, the method for improving the azimuth estimation capability of a vector hydrophone by extracting a target line spectrum includes the following steps:

[0097] Step 1: Perform Fourier transform on the data of the vector hydrophone scalar channel

[0098] The method of this embodiment uses the spectrum intensity of the vector hydrophone scalar channel as the weighting coefficient of the weighted histogram, and the high-order time-interval (HOT-Lac) characteristic line spectrum extraction only needs to process the vector hydrophone scalar channel signal.

[0099] Step 2: Reference Figure 2 As shown in the figure, based on whether it is possible to use expert experience to extract data without targets and containing interference line spectra in the sonar database of the current environment, the HOT-Lac feature extraction method based on background learning or the integral accumulation HOT-Lac feature extraction method without prior knowledge is selected.

[0100] First, if it is possible to use expert experience to extract a segment of the current environment sonar image database that does not contain target line spectrum but only background data, choose to use the HOT-Lac feature extraction method based on background learning.

[0101] Reference Figure 3 As shown, the HOT-Lac feature extraction method based on background learning is divided into the following steps:

[0102] 1. By manually selecting the mean value U of the background data without a target within d seconds in the database, m groups of data within d seconds are as follows:

[0103]

[0104] 2. Select n groups of data within the time step of t seconds before the current moment, and calculate the difference between them and U for post-processing, as shown in the following formula:

[0105]

[0106] Here, k represents the order.

[0107] Reference Figure 4 and Figure 5 As shown in the figure, the algorithm is verified using a set of data. The data contains interference line spectra, and the target sound source emits three line spectra with different intensities. By comparing the figures before and after processing, it can be seen that the HOT-Lac feature extraction method based on background learning can effectively remove the interference line spectra and retain the target line spectra.

[0108] Second, if it is not possible to use expert experience to extract a section of the current environment sonar image database without threatening targets, which only contains data containing both interference and targets, choose to use the integral accumulation HOT-Lac feature extraction without prior knowledge.

[0109] Reference Figure 6 As shown, the process of HOT-Lac feature extraction method without prior knowledge is:

[0110] 1) The integral accumulation HOT-Lac feature extraction method without prior knowledge is different from the HOT-Lac feature extraction method based on background learning. The amount of calculation is higher than that of the HOT-Lac feature extraction method based on background learning. The integral accumulation HOT-Lac feature extraction method without prior knowledge selects the historical mean U of nm groups of data within the time step s seconds before the current moment (corresponding to m groups), and n is all the data currently stored in the device, as shown in the following formula:

[0111]

[0112] 2) Select t seconds as the time length, and then calculate the difference between v sets of data within t seconds before the current moment and U, and then process them as follows:

[0113]

[0114] The differences between the HOT-Lac feature extraction method based on background learning and the integral accumulation HOT-Lac feature extraction method without prior knowledge are as follows:

[0115] a. Compared with the HOT-Lac feature extraction method based on background learning, the integral accumulation HOT-Lac feature extraction method without prior knowledge requires a period of data accumulation and processing, which may lead to the loss of target line spectrum or the failure to completely eliminate the interference line spectrum.

[0116] b. The integral accumulation HOT-Lac feature extraction method without prior knowledge requires a large amount of data. If the amount of data is insufficient, it is necessary to make appropriate selections for m and v in equations (6) and (7) according to the actual situation.

[0117] c. The integral accumulation HOT-Lac feature extraction method without prior knowledge can be used when there is a target line spectrum. The HOT-Lac feature extraction method based on background learning is mainly used when it is possible to extract background data without a target line spectrum in the current environmental sonar signal.

[0118] Step 3: After line spectrum extraction, vector hydrophone weighted histogram method is used to estimate the direction.

[0119] Compare the weighted histogram method azimuth estimation before and after line spectrum extraction, such as Figure 7 As shown in the figure, the peak is weakened and the peak of the target orientation is highlighted, achieving the purpose of extracting the target line spectrum. It can be seen that the target orientation is highlighted and the peak is reduced.

[0120] The algorithm used in the present invention has a simple structure, does not need to set any parameters in advance, is flexible and fast, and can eliminate interference line spectra without prior knowledge of interference signal frequencies and target signal frequencies. It can still be extracted when the target line spectrum produces a Doppler frequency shift; it can eliminate stable interference signal line spectra in an environment that cannot be eliminated in integral processing, and can effectively suppress interference line spectra, improve azimuth estimation capability, and reduce false alarm rate.

[0121] The present invention also provides a target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone. The target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone can be implemented by executing the process steps of the target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone, that is, those skilled in the art can understand the target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone as a preferred implementation of the target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone.

[0122] Specifically, a target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone includes:

[0123] Module M1: Acquire vector hydrophone data and perform normalization;

[0124] The vector hydrophone data includes scalar channel data;

[0125] Module M2: Based on the vector hydrophone data, obtain the high-order time interval characteristics and extract the target line spectrum;

[0126] Module M3: Calculate the spectrum intensity of the target line spectrum to obtain the target direction.

[0127] The module M1 includes the following submodules:

[0128] Module M1.1: Obtain a frame of data from the vector hydrophone, and form a one-dimensional array after Fourier transformation. Each number in the array represents the amplitude of a frequency point; accumulate multiple frames of data arranged in time sequence after Fourier transformation to form a two-dimensional matrix representing time frequency;

[0129] Module M1.2: Normalize the two-dimensional matrix.

[0130] The module M2 includes the following submodules:

[0131] Module M2.1: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows, and solve the time gap property for it:

[0132]

[0133] Among them, P t Represents pixel value;

[0134] Module M2.2: Use high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain high-order time gap characteristics:

[0135]

[0136] If it is possible to use expert experience to extract a segment of data without target line spectrum and only containing background in the current environment sonar image database, then a high-order temporal intermittent feature extraction method based on background learning is selected;

[0137] If it is not possible to use expert experience to extract a segment of the current environment sonar image database without threatening targets and only containing data with both interference and targets, then the integral accumulation high-order temporal intermittent feature extraction without prior knowledge is chosen.

[0138] The high-order temporal intermittent feature extraction process based on background learning includes:

[0139] Select the mean value U of the background data without a target within d seconds in the database, and m groups of data within d seconds:

[0140]

[0141] Select n groups of data within the time step of t seconds before the current moment, and calculate the difference between them and U for post-processing:

[0142]

[0143] The process of extracting high-order temporal intermittent features by integrating and accumulating without prior knowledge includes:

[0144] Select the historical mean value U of the data group nm within the time step before the current time s seconds:

[0145]

[0146] Where n is all the data currently stored in the device;

[0147] Select t seconds as the time length, and then calculate the difference between v sets of data within t seconds before the current moment and U for post-processing:

[0148]

[0149] The module M3 includes calculating the spectrum intensity of the signal line spectrum after removing the interference line spectrum by high-order time interval, and using the weighted histogram method to weight the spectrum intensity by the acoustic energy flow. The peak of the curve after fitting is the target direction.

[0150] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0151] In the description of the present application, it should be understood that the terms "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0152] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone, characterized in that: include: Step S1: acquiring vector hydrophone data and performing normalization processing; The vector hydrophone data includes scalar channel data; Step S2: based on the vector hydrophone data, obtain high-order time interval characteristics and extract the target line spectrum; Step S3: Calculate the spectrum intensity of the target line spectrum to obtain the target direction.

2. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 1, characterized in that: The step S1 comprises the following sub-steps: Step S1.1: Obtain a frame of data from the vector hydrophone, and form a one-dimensional array after Fourier transformation, where each number in the array represents the amplitude of a frequency point; accumulate multiple frames of data arranged in time sequence after Fourier transformation to form a two-dimensional matrix representing time frequency; Step S1.2: normalize the two-dimensional matrix.

3. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 2, characterized in that: The step S2 comprises the following sub-steps: Step S2.1: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows, and solve the time gap property for it: Among them, P t represents the pixel value, E{·} is the data expectation; Step S2.2: Use the high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain the high-order time gap feature:

4. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 2, characterized in that: If it is possible to use expert experience to extract a segment of data without target line spectrum and only containing background in the current environment sonar image database, then a high-order temporal intermittent feature extraction method based on background learning is selected; If it is not possible to use expert experience to extract a segment of the current environment sonar image database without threatening targets and only containing data with both interference and targets, then the integral accumulation high-order temporal intermittent feature extraction without prior knowledge is chosen.

5. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 4, characterized in that: The high-order temporal intermittent feature extraction process based on background learning includes: Select the mean value U of the background data without a target within d seconds in the database, and m groups of data within d seconds: Select n groups of data within the time step of t seconds before the current moment, and calculate the difference between them and U for post-processing: Here, k represents the order.

6. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 4, characterized in that: The process of extracting high-order temporal intermittent features by integrating and accumulating without prior knowledge includes: Select the historical mean value U of the data group nm within the time step before the current time s seconds: Where n is all the data currently stored in the device; Select t seconds as the time length, and then calculate the difference between v sets of data within t seconds before the current moment and U for post-processing:

7. The target line spectrum extraction method for improving the azimuth estimation capability of a vector hydrophone according to claim 1, characterized in that: The step S3 includes calculating the spectrum intensity of the signal line spectrum after removing the interference line spectrum by high-order time interval, and using the weighted histogram method to weight the spectrum intensity by the acoustic energy flow. The peak of the curve after fitting is the target direction.

8. A target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone, characterized in that: include: Module M1: Acquire vector hydrophone data and perform normalization; The vector hydrophone data includes scalar channel data; Module M2: Based on the vector hydrophone data, obtain the high-order time interval characteristics and extract the target line spectrum; Module M3: Calculate the spectrum intensity of the target line spectrum to obtain the target direction.

9. The target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone according to claim 8, characterized in that: The module M1 includes the following submodules: Module M1.1: Obtain a frame of data from the vector hydrophone, and form a one-dimensional array after Fourier transformation. Each number in the array represents the amplitude of a frequency point; accumulate multiple frames of data arranged in time sequence after Fourier transformation to form a two-dimensional matrix representing time frequency; Module M1.2: Normalize the two-dimensional matrix.

10. The target line spectrum extraction system for improving the azimuth estimation capability of a vector hydrophone according to claim 9, characterized in that: The module M2 includes the following submodules: Module M2.1: Combine the one-dimensional array formed by Fourier transform of the current moment data with the n-row array corresponding to the previous t seconds to form a sub-two-dimensional matrix with n+1 rows, and solve the time gap property for it: Among them, P t represents the pixel value, E{·} is the data expectation; Module M2.2: Use high-order center distance to process the time gap of the n+1-row sub-two-dimensional matrix to obtain high-order time gap characteristics:

Images