A method for constructing a semantic segmentation sample library for cold spring plume seismic response

Abstract

The present application provides a method for constructing a semantic segmentation sample library of cold spring plume seismic response, the construction method comprising: for a target cold spring plume morphological model, based on the bubble parameters included in each group of sample data, obtaining the acoustic wave velocity of the cold spring bubble plume, and fusing it with the seawater acoustic wave velocity distribution map to construct a cold spring plume acoustic wave velocity model; based on the water body velocity parameters, constructing a cold spring plume time domain reflection coefficient model through time domain processing; based on the seismic parameters, using the seismic convolution forward expression to obtain the cold spring plume seismic forward modeling result; based on the noise parameters, performing noise addition and regularization normalization processing to obtain the cold spring plume seismic response result and performing corresponding semantic segmentation processing with the target cold spring plume morphological model to obtain semantic segmentation label samples, so as to construct a cold spring plume seismic response semantic segmentation sample library, thereby improving the construction quality, efficiency and diversity of the cold spring plume seismic response semantic segmentation samples.

Description

Technical Field

[0001] The present application relates to the technical field of bubble plume flow water bodies in submarine cold seep areas, and in particular to a method for constructing a semantic segmentation sample library of cold seep plume flow seismic responses. Background Art

[0002] Seismic exploration is one of the important means of detecting cold spring plumes. By conducting marine seismic exploration in cold spring areas, performing corresponding seismic data processing on water body seismic data, and then performing offset imaging, the imaging results of cold spring plumes in bubbly water bodies can be obtained, that is, the seismic response of water body plumes in cold spring areas.

[0003] At present, the method for constructing a semantic segmentation sample library of cold spring plume seismic response mainly involves manually interpreting the cold spring plume area in the seismic imaging results after imaging the cold spring plume, and then using annotation software to delineate it into target area and background area to obtain the cold spring seismic response and its corresponding semantic segmentation annotation results. Since this method is a manual annotation method, it reduces the quality and efficiency of constructing cold spring plume seismic response samples, and also reduces the diversity of cold spring plume seismic response samples. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a method, device, electronic device and storage medium for constructing a semantic segmentation sample library of cold spring plume seismic response, using multiple groups of sample data in a preset sample data set to respectively construct a cold spring plume acoustic wave velocity model and a cold spring plume time domain reflection coefficient model under each group of sample data, and then synthesize the target cold spring plume morphological model into seismic data through seismic convolution forward calculation, and perform noise addition and regularization normalization processing on it, thereby improving the quality of the obtained cold spring plume seismic response results, and semantically segmenting the cold spring plume seismic response results directly corresponding to the target cold spring plume morphological model, and obtaining the cold spring plume seismic response semantic segmentation label samples under each group of sample data, and then constructing a cold spring plume seismic response semantic segmentation sample library, thereby improving the construction quality and efficiency of the cold spring plume seismic response semantic segmentation samples, and improving the diversity of the cold spring plume seismic response samples.

[0005] The present application embodiment provides a method for constructing a semantic segmentation sample library of cold spring plume seismic response, the construction method comprising:

[0006] For a target cold spring plume morphology model among the pre-constructed multiple cold spring plume morphology models, based on the bubble parameters included in each set of sample data in the preset sample data set, a preset sound velocity expression in a bubble-containing medium is used to calculate, so as to obtain the cold spring bubble plume acoustic wave velocity under each set of sample data, and the cold spring bubble plume acoustic wave velocity is merged with a preset seawater acoustic wave velocity distribution map to construct a cold spring plume acoustic wave velocity model under each set of sample data;

[0007] Based on the water velocity parameters included in the cold spring plume acoustic wave velocity model, a preset reflection coefficient expression is used for calculation, and a cold spring plume time domain reflection coefficient model under each set of sample data is constructed through time domain processing;

[0008] Based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain the cold spring plume seismic forward modeling results under each set of sample data;

[0009] Based on the noise parameters included in each set of sample data, the cold spring plume seismic forward modeling results are subjected to noise addition and regularization normalization processing to obtain the cold spring plume seismic response results under each set of sample data;

[0010] The cold spring plume seismic response results and the target cold spring plume morphological model are subjected to corresponding semantic segmentation processing to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0011] Furthermore, the plurality of cold spring plume morphology models are pre-constructed by the following steps:

[0012] Based on the preset size parameters, construct a morphological model background layer, and set the pixel value of the morphological model background layer to a first preset value;

[0013] Based on the morphological size parameters corresponding to the preset multiple cold spring plume morphologies, multiple cold spring plume areas are constructed in the morphological model background layer, and the pixel value of the cold spring plume area is set to a second preset value to construct multiple cold spring plume morphological models.

[0014] Furthermore, the bubble parameters included in each set of sample data in the preset sample data set are calculated using a preset sound velocity expression in a bubble-containing medium to obtain the sound wave velocity of the cold spring bubble plume under each set of sample data, including:

[0015] Based on the bubble parameters included in each set of sample data in the preset sample data set, using the preset random medium theory and plume flow distribution law, a random disturbance value corresponding to the bubble parameter is generated to achieve random changes in the bubble content value and the bubble radius value included in the bubble parameter;

[0016] Based on the random disturbance value corresponding to the bubble parameter and the preset gas-liquid parameter, the preset sound velocity expression in the bubble-containing medium is used for calculation to obtain the sound wave velocity of the cold spring bubble plume under each set of sample data.

[0017] Furthermore, the cold spring bubble plume acoustic wave velocity and the preset seawater acoustic wave velocity distribution map are merged to construct the cold spring plume acoustic wave velocity model under each set of sample data, including:

[0018] Based on a preset seawater acoustic wave velocity distribution map, constructing a seawater acoustic wave velocity background layer corresponding to the target cold seep plume flow morphology model;

[0019] Based on the target cold seep plume area in the target cold seep plume morphology model, the acoustic wave velocity of the cold seep bubble plume is fused into the seawater acoustic wave velocity background layer to construct the cold seep plume acoustic wave velocity model under each set of sample data.

[0020] Furthermore, the water velocity parameters included in the cold spring plume acoustic wave velocity model are calculated using a preset reflection coefficient expression, and the cold spring plume time domain reflection coefficient model under each set of sample data is constructed through time domain processing, including:

[0021] Based on the cold spring plume acoustic wave velocity model, determining a water velocity parameter included in the cold spring plume acoustic wave velocity model;

[0022] Based on the water velocity value corresponding to each bottom interface included in the water velocity parameter, a preset reflection coefficient expression is used to calculate to obtain the reflection coefficient corresponding to each bottom interface;

[0023] Based on the reflection coefficient, the depth value of the bottom interface of each layer is corresponded to a time point to construct a time domain reflection coefficient model of the cold spring plume under each set of sample data.

[0024] Furthermore, based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to calculate and obtain the cold spring plume seismic forward modeling results under each set of sample data, including:

[0025] Determining the reflection coefficient sequence under each set of sample data in the cold spring plume time domain reflection coefficient model;

[0026] Based on the reflection coefficient sequence and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain a seismic record sequence under each set of sample data, so as to obtain a cold spring plume seismic forward modeling result under each set of sample data.

[0027] Furthermore, based on the noise parameters included in each set of sample data, the cold spring plume seismic forward modeling results are subjected to noise addition and regularization normalization processing to obtain the cold spring plume seismic response results under each set of sample data, including:

[0028] Based on the noise signal-to-noise ratio value in the noise parameter included in each group of sample data, adding random noise corresponding to the noise signal-to-noise ratio value to the cold spring plume seismic forward modeling result to obtain a noise processing result;

[0029] Performing mean calculation and variance calculation on the seismic record data in the noise processing result respectively to obtain the mean value and variance of the seismic record data corresponding to the noise processing result;

[0030] Based on the noise processing result, the mean value of the seismic record data and the variance of the seismic record data, a preset regular standardized expression is used for calculation to obtain the cold spring plume seismic response result under each group of sample data.

[0031] Furthermore, the cold spring plume seismic response result and the target cold spring plume morphological model are subjected to corresponding semantic segmentation processing to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, including:

[0032] extracting model features from the cold spring plume seismic response results according to the morphological features of the target cold spring plume morphological model;

[0033] Matching the model features with the morphological features to determine a corresponding relationship between the cold spring plume seismic response result and the target cold spring plume morphological model;

[0034] Based on the corresponding relationship and the seismic response result of the cold spring plume, the target cold spring plume morphological model is semantically segmented to obtain the semantic segmentation label samples of the seismic response of the cold spring plume under each group of sample data.

[0035] The present application also provides a device for constructing a semantic segmentation sample library of cold spring plume seismic response, the construction device comprising:

[0036] The acoustic wave velocity model construction module is used to calculate the acoustic wave velocity of the cold spring bubble plume under each set of sample data based on the bubble parameters included in each set of sample data in the preset sample data set, and to fuse the acoustic wave velocity of the cold spring bubble plume with the preset seawater acoustic wave velocity distribution map to construct the acoustic wave velocity model of the cold spring plume under each set of sample data;

[0037] A reflection coefficient model building module, used to calculate based on the water velocity parameters included in the cold spring plume acoustic wave velocity model using a preset reflection coefficient expression, and to build a cold spring plume time domain reflection coefficient model for each set of sample data through time domain processing;

[0038] A seismic convolution forward modeling calculation module, used to calculate based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data using a preset seismic convolution forward modeling expression to obtain the cold spring plume seismic forward modeling results under each set of sample data;

[0039] A seismic response processing module, for performing noise addition and regularization normalization processing on the cold spring plume seismic forward modeling result based on the noise parameters included in each set of sample data, so as to obtain the cold spring plume seismic response result under each set of sample data;

[0040] A semantic segmentation sample library construction module is used to perform corresponding semantic segmentation processing on the cold spring plume seismic response results and the target cold spring plume morphological model to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0041] An embodiment of the present application also provides an electronic device, including: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response as described above are performed.

[0042] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response as described above are executed.

[0043] The method for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application includes: for a target cold spring plume morphology model among a plurality of pre-constructed cold spring plume morphology models, based on the bubble parameters included in each group of sample data in a preset sample data set, using a preset sound velocity expression in a bubble-containing medium to calculate, to obtain the cold spring bubble plume acoustic wave velocity under each group of sample data, and fusing the cold spring bubble plume acoustic wave velocity with a preset seawater acoustic wave velocity distribution map to construct a cold spring plume acoustic wave velocity model under each group of sample data; based on the water body velocity parameters included in the cold spring plume acoustic wave velocity model, using a preset reflection coefficient expression to calculate, and constructing the cold spring plume acoustic wave velocity model under each group of sample data through time domain processing. time domain reflection coefficient model; based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each group of sample data, use the preset seismic convolution forward expression to calculate and obtain the cold spring plume seismic forward modeling results under each group of sample data; based on the noise parameters included in each group of sample data, add noise and regularize the cold spring plume seismic forward modeling results to obtain the cold spring plume seismic response results under each group of sample data; perform corresponding semantic segmentation processing on the cold spring plume seismic response results and the target cold spring plume morphological model to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0044] Compared with the method in the prior art of manually interpreting the cold spring plume area in the seismic imaging results after imaging the cold spring plume, and then using annotation software to delineate it into the target area and the background area, multiple groups of sample data in a preset sample data set are used to respectively construct a cold spring plume acoustic wave velocity model and a cold spring plume time domain reflection coefficient model under each group of sample data, and then the target cold spring plume morphological model is synthesized into seismic data through seismic convolution forward calculation, and noise is added and regularized and normalized, thereby improving the quality of the obtained cold spring plume seismic response results, and the cold spring plume seismic response results are directly semantically segmented corresponding to the target cold spring plume morphological model, and the cold spring plume seismic response semantic segmentation label samples under each group of sample data are obtained, and then a cold spring plume seismic response semantic segmentation sample library is constructed, thereby improving the construction quality and efficiency of the cold spring plume seismic response semantic segmentation samples, and improving the diversity of the cold spring plume seismic response samples.

[0045] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A flowchart of a method for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application;

[0048] FIG. 2 (a) is a schematic diagram of a plurality of cold spring plume flow morphology models provided in an embodiment of the present application;

[0049] FIG2( b ) is a schematic diagram of a target cold spring plume flow morphology model provided in an embodiment of the present application;

[0050] Figure 3 A schematic diagram of a calculation process of the acoustic wave velocity of a cold spring bubble plume provided in an embodiment of the present application;

[0051] Figure 4 An example distribution diagram of seawater sound wave velocity provided in an embodiment of the present application;

[0052] Figure 5 A schematic diagram of the fusion of a cold spring plume acoustic wave velocity model provided in an embodiment of the present application;

[0053] Figure 6 A schematic diagram of constructing a time-domain reflection coefficient model of a cold spring plume provided in an embodiment of the present application;

[0054] Figure 7 A calculation example diagram of a cold spring plume seismic forward modeling result provided in an embodiment of the present application;

[0055] Figure 8 This is a calculation example diagram of a cold spring plume seismic response result provided in an embodiment of the present application;

[0056] Fig. 9 A schematic diagram of label samples of a cold spring plume seismic response semantic segmentation sample library provided in an embodiment of the present application;

[0057] Fig.10 An example diagram of a semantic segmentation label sample of a cold spring plume seismic response provided in an embodiment of the present application;

[0058] Fig.11 A schematic diagram of the structure of a device for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application;

[0059] Fig.12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0061] It should be noted that in the area of ​​seafloor cold seep activity, when natural gas migrates upward from the interior of the earth's crust, it enters the seawater through migration channels such as faults and cracks, and then escapes upward in the form of bubbles, forming seafloor cold seep plumes.

[0062] In addition, seismic exploration is one of the important means of detecting cold spring plumes. By conducting marine seismic exploration in cold spring areas, performing corresponding seismic data processing on water body seismic data, and then performing offset imaging, the imaging results of cold spring plumes in bubbly water bodies can be obtained, that is, the seismic response of water body plumes in cold spring areas.

[0063] Furthermore, the semantic segmentation method based on deep learning can identify the cold spring plume area directly from the imaging results of the cold spring plume through a certain amount of sample learning, but the deep learning process requires a large number of semantic segmentation sample libraries of different categories of cold spring plume seismic responses.

[0064] Research has found that at present, the method for constructing a semantic segmentation sample library of the seismic response of cold spring plumes mainly involves manually interpreting the cold spring plume area in the seismic imaging results after imaging the cold spring plumes, and then using annotation software to delineate it into target area and background area to obtain the cold spring seismic response and its corresponding semantic segmentation annotation results; the above method mainly relies on seismic imaging and manual demarcation to construct a semantic segmentation sample library of the seismic response of cold spring plumes, and this method has the following disadvantages.

[0065] First, since the actual cold spring plume seismic data processing methods and migration imaging results rely on the expertise of professional geophysical interpreters, there are certain deviations in the quality of the sample library.

[0066] Second, when manually annotating the cold spring plume seismic data imaging results, the quality of the annotation cannot be guaranteed and the workload is large, which reduces the quality and efficiency of constructing the cold spring plume seismic response samples.

[0067] Third, since actual marine seismic exploration does not pay attention to the seismic records of water bodies containing cold spring plumes, the water body seismic data are generally cut off, resulting in poor diversity in the number of actual cold spring plume seismic samples, which cannot reflect the seismic response characteristics of cold spring plumes under multiple groups of sample data, thereby reducing the diversity of cold spring plume seismic response samples.

[0068] Based on this, an embodiment of the present application provides a method for constructing a semantic segmentation sample library of cold spring plume seismic response, using multiple groups of sample data in a preset sample data set to respectively construct a cold spring plume acoustic wave velocity model and a cold spring plume time domain reflection coefficient model under each group of sample data, and then synthesize the target cold spring plume morphological model into seismic data through seismic convolution forward calculation, and perform noise addition and regularization normalization processing on it, thereby improving the quality of the obtained cold spring plume seismic response results, and semantically segmenting the cold spring plume seismic response results directly corresponding to the target cold spring plume morphological model, and obtaining the cold spring plume seismic response semantic segmentation label samples under each group of sample data, and then constructing a cold spring plume seismic response semantic segmentation sample library, thereby improving the construction quality and efficiency of the cold spring plume seismic response semantic segmentation samples, and improving the diversity of the cold spring plume seismic response samples.

[0069] See also Figure 1 , Figure 1 A flowchart of a method for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application. Figure 1 As shown in, the construction method provided in the embodiment of the present application includes:

[0070] S101. For a target cold spring plume morphology model among a plurality of pre-constructed cold spring plume morphology models, based on the bubble parameters included in each set of sample data in a preset sample data set, a preset sound velocity expression in a bubble-containing medium is used to calculate to obtain the acoustic wave velocity of the cold spring bubble plume under the said each set of sample data, and the acoustic wave velocity of the cold spring bubble plume is merged with a preset seawater acoustic wave velocity distribution map to construct a cold spring plume acoustic wave velocity model under the said each set of sample data.

[0071] In an embodiment of the present application, a sample data set is pre-set for the semantic segmentation sample library of the seismic response of the cold spring plume that needs to be constructed. The sample data set includes multiple groups of sample data, each group of sample data includes but is not limited to bubble parameters, seismic parameters, noise parameters, etc., so as to obtain the semantic segmentation sample library of the seismic response of the cold spring plume using multiple groups of sample data.

[0072] In one embodiment of the present application, in a specific implementation, the step of pre-constructing the plurality of cold spring plume morphology models in step S101 may include:

[0073] S1011. Construct a morphological model background layer based on preset size parameters, and set a pixel value of the morphological model background layer to a first preset value.

[0074] In this step, based on the size parameters of the preset background layer, a morphological model background layer is constructed with a pixel value set to a first preset value; wherein the first preset value can be set to 0.

[0075] S1012. Based on the morphological size parameters corresponding to the preset multiple cold spring plume morphologies, construct multiple cold spring plume areas in the morphological model background layer, and set the pixel value of the cold spring plume area to a second preset value to construct multiple cold spring plume morphological models.

[0076] In the embodiment of the present application, the morphological size parameters include morphological parameters and size parameters; since the bubble plumes in the seabed cold seep area are affected by factors such as ocean currents, the morphologies of the cold seep plumes are different. According to the existing underwater video data and multi-beam data of the cold seep plumes, the morphologies of the cold seep plumes can generally include inclined bodies, gem bodies, cones, flame-shaped and other plume morphologies.

[0077] Among them, the second preset value can be set to 1.

[0078] For example, please refer to Figures 2(a) and 2(b), where Figure 2(a) is a schematic diagram of a variety of cold spring plume flow morphology models provided in an embodiment of the present application, and Figure 2(b) is a schematic diagram of a target cold spring plume flow morphology model provided in an embodiment of the present application.

[0079] As shown in Figure 2 (a), since the cold seep plume morphology includes nine types of plume morphologies, such as inclined body, gem body, cone body, and flame shape, nine types of cold seep plume morphology models were constructed in the morphological model background layer.

[0080] Furthermore, for each of the multiple cold spring plume morphology models, for example, the flame-shaped cold spring plume morphology model shown in FIG. 2 (b) is used as the target cold spring plume morphology model of the sample library for currently constructing the cold spring plume morphology model.

[0081] In one embodiment of the present application, in a specific implementation, the step of calculating the sound wave velocity of the cold spring bubble plume under each set of sample data based on the bubble parameters included in each set of sample data in the preset sample data set using a preset sound velocity expression in the bubble-containing medium in step S101 may include:

[0082] S1013. Based on the bubble parameters included in each set of sample data in the preset sample data set, using the preset random medium theory and plume flow distribution law, generate random disturbance values ​​corresponding to the bubble parameters to achieve random changes in the bubble content value and the bubble radius value included in the bubble parameters.

[0083] It should be noted that seismic exploration refers to the use of the different propagation speeds of seismic waves in different media based on the characteristics of seismic wave propagation underground to infer the properties and structure of underground media; the same principle is also used for the detection of cold spring plumes in water bodies, so it is necessary to determine the acoustic wave velocity of the cold spring bubble plume.

[0084] In the embodiment of the present application, the bubble parameters include a gas content value and a bubble radius value.

[0085] Here, as the bubbles in the cold spring bubble plume continuously rise from the seabed, they are affected by factors such as seawater pressure and methane dissolution. As the seawater depth becomes shallower, the bubble radius and the bubble gas content first increase and then decrease. At a certain moment of the bubble buoyancy, large and small bubbles show the characteristics of sequential stratification, and at the same time have macroscopic characteristics of random distribution in the continuous rising process. Therefore, a certain bottom interface stratification treatment is performed on plumes of different forms in the longitudinal direction, and each layer of the bottom interface corresponds to a gas content value and a bubble radius value. At the same time, random disturbances generated by random medium theory are added to achieve random changes in the bubble content value and the bubble radius value included in the bubble parameters.

[0086] In this step, the gas content value and the bubble radius value included in each set of sample data are substituted into a preset random perturbation expression, and the obtained calculation result is determined as the random perturbation value corresponding to the generated bubble parameter.

[0087] Specifically, the preset random perturbation expression is as follows.

[0088] .

[0089] in, and are the bubble radius and gas content values ​​respectively; and are the background radius and background bubble content of each bottom interface, respectively; and are the bubble radius and gas content values ​​of random disturbance respectively; and are second-order stationary spatial random processes with certain mean and variance respectively.

[0090] S1014. Based on the random disturbance value corresponding to the bubble parameter and the preset gas-liquid parameter, a preset sound velocity expression in the bubble-containing medium is used to calculate to obtain the sound wave velocity of the cold spring bubble plume under each set of sample data.

[0091] In this step, the target cold spring plume area in the target cold spring plume morphology model is extracted, and based on the corresponding randomly disturbed bubble radius value and gas content value in the target cold spring plume area and the preset gas-liquid parameters, and with the sound pressure and radial vibration velocity at the bubble wall as boundary conditions, the preset sound velocity expression in the bubble-containing medium is used for calculation to obtain the cold spring bubble plume sound wave velocity under each group of sample data.

[0092] In the embodiment of the present application, the preset expression of sound velocity in the bubble-containing medium is as follows.

[0093] .

[0094] Among them, C m is the acoustic wave velocity of the cold spring bubble plume, in m / s; K is the bulk modulus of the liquid, in N / m 3 ; K b is the gas bulk modulus, in N / m 3 ; is the liquid density, in kg / m 3 ; is the gas density, in kg / m 3 ;ω is the frequency, in Hz; is the bubble radius, in m; is the surface tension of the liquid, in N / m 3 ; R is the radius when the two-phase mixing zone containing bubbles is assumed to be spherical, in m; is the gas content (bubble volume fraction), that is, the volume fraction of the volume occupied by the gas in the bubble in the spherical area with a radius of R.

[0095] Here, K, K b , ρ, ρ b , R, ω, σ (liquid bulk modulus, gas bulk modulus, liquid density, gas density, radius when the two-phase mixing zone containing bubbles is assumed to be spherical, frequency, liquid surface tension) are all preset gas-liquid parameters; it can be seen that the acoustic wave velocity of the cold spring bubble plume is mainly affected by the gas content value and the bubble radius value.

[0096] For example, see Figure 3 , Figure 3 A schematic diagram of the calculation process of the acoustic wave velocity of a cold spring bubble plume provided in an embodiment of the present application. Figure 3As shown in , after the target cold spring plume area in the target cold spring plume morphology model is extracted, the bubble radius value and gas content value in each group of sample data are randomly perturbed, and finally the cold spring bubble plume sound wave velocity is calculated by calculating the preset sound velocity expression in the bubble-containing medium.

[0097] In one embodiment of the present application, in a specific implementation, the step of fusing the cold spring bubble plume acoustic wave velocity and the preset seawater acoustic wave velocity distribution map in step S101 to construct the cold spring plume acoustic wave velocity model under each set of sample data may include:

[0098] S1015. Based on a preset seawater acoustic wave velocity distribution map, construct a seawater acoustic wave velocity background layer corresponding to the target cold seep plume morphology model.

[0099] It should be noted that, in actual situations, since the cold seep plume is formed in seawater, it is necessary to calculate the acoustic velocity of the entire water body including the plume and its seawater background layer. Therefore, it is necessary to merge the seawater acoustic velocity background layer corresponding to the constructed target cold seep plume morphology model with the acoustic velocity of the cold seep bubble plume.

[0100] In this step, a preset seawater acoustic wave velocity distribution map is used to construct a seawater acoustic wave velocity background layer corresponding to the target cold seep plume flow morphology model.

[0101] For example, see Figure 4 , Figure 4 This is an example distribution diagram of seawater sound wave velocity provided by an embodiment of the present application. Figure 4 As shown in the figure, the horizontal axis of the graph represents the speed of sound (m / s), and the vertical axis represents the depth of seawater (m). It can be seen that with the increase of depth, the speed of seawater sound waves first decreases and then increases.

[0102] S1016. Based on the target cold seep plume area in the target cold seep plume morphology model, the acoustic wave velocity of the cold seep bubble plume is integrated into the seawater acoustic wave velocity background layer to construct a cold seep plume acoustic wave velocity model for each set of sample data.

[0103] See also Figure 5 , Figure 5 A fusion schematic diagram of a cold spring plume sound wave velocity model provided in an embodiment of the present application.

[0104] like Figure 5As shown in , in this step, first, based on the acoustic wave velocity of the cold spring bubble plume, the target cold spring plume area in the target cold spring plume morphology model is velocity-marked; then, the seawater acoustic wave velocity background layer is fused with the target cold spring plume area marked with the acoustic wave velocity of the cold spring bubble plume; finally, the cold spring plume acoustic wave velocity model under each set of sample data is obtained.

[0105] S102, based on the water velocity parameters included in the cold spring plume acoustic wave velocity model, a preset reflection coefficient expression is used for calculation, and a cold spring plume time domain reflection coefficient model for each set of sample data is constructed through time domain processing.

[0106] In an embodiment of the present application, after obtaining the acoustic wave velocity model of the cold spring plume, the reflection coefficient model of the cold spring plume in the time domain is obtained by calculation of the reflection coefficient. Here, since the presence of bubbles in the seawater will change the acoustic properties of the seawater, the response characteristics of the active cold spring on the seismic profile can be identified based on the difference in reflection seismic characteristics between the cold spring bubble plume and the background seawater.

[0107] In one implementation of the present application, during specific implementation, step S102 may include:

[0108] S1021. Based on the cold spring plume acoustic wave velocity model, determine the water velocity parameters included in the cold spring plume acoustic wave velocity model.

[0109] In this step, the water velocity of each bottom interface and the water velocity of the next bottom interface are determined in the cold spring plume acoustic wave velocity model.

[0110] S1022. Based on the water velocity value corresponding to each bottom interface included in the water velocity parameter, a preset reflection coefficient expression is used to perform calculation to obtain the reflection coefficient corresponding to each bottom interface.

[0111] In this step, the water velocity value corresponding to each bottom interface is substituted into a preset reflection coefficient expression, and the reflection coefficient corresponding to each bottom interface is obtained by calculation.

[0112] In the embodiment of the present application, the preset reflection coefficient expression is as follows.

[0113] .

[0114] in, is the reflection coefficient of the bottom interface of the i-th layer; and are the water velocities at the bottom interface of the i+1th layer and the ith layer respectively.

[0115] It should be noted that the reflection coefficient expression is inferred from the normal incidence reflection coefficient expression based on the assumption that the seismic traces on the seismic profile are normal incidence traces.

[0116] Specifically, the expression of the normal incidence reflection coefficient is as follows.

[0117] .

[0118] in, is the reflection coefficient of the bottom interface of the i-th layer; and are the water velocities at the bottom interface of the i+1th layer and the ith layer respectively; and are the densities of the bottom interface of the i+1th layer and the ith layer respectively.

[0119] Here, according to research, when the bubbles from the decomposition of hydrates rise into the seawater, they have little effect on the overall density of the seawater, which means that the density change can be ignored when calculating the reflection coefficient, and the effect of the bubbles on the water velocity can be mainly considered. Therefore, the reflection coefficient expression ignores the density part in the normal incidence reflection coefficient expression.

[0120] S1023. Based on the reflection coefficient, correspond the depth value of the bottom interface of each layer to a time point to construct a time domain reflection coefficient model of the cold spring plume under each set of sample data.

[0121] See also Figure 6 , Figure 6 A schematic diagram of constructing a time domain reflection coefficient model of a cold spring plume provided in an embodiment of the present application. Figure 6 As shown in , in this step, the reflection coefficient and depth value of the bottom interface of the first layer are respectively corresponded to the preset time points, so as to construct the time domain reflection coefficient model of the cold spring plume under each set of sample data through reflection coefficient calculation and deep-time conversion.

[0122] S103, based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain the cold spring plume seismic forward modeling results for each set of sample data.

[0123] In an embodiment of the present application, after constructing a time domain reflection coefficient model of a cold spring plume, a corresponding synthetic seismic record of a bubble-containing plume water body (i.e., a seismic imaging result) is obtained through convolution forward simulation based on a seismic wavelet with a different main frequency from the seismic parameters.

[0124] The seismic parameters include but are not limited to the type, main frequency value and seismic wavelet sequence corresponding to the seismic wavelet.

[0125] In one implementation of the present application, during specific implementation, step S103 may include:

[0126] S1031. Determine a reflection coefficient sequence for each set of sample data in the cold spring plume time domain reflection coefficient model.

[0127] In this step, based on the cold spring plume time domain reflection coefficient model, the reflection coefficient sequence under each group of sample data is determined; the reflection coefficient sequence includes the reflection coefficient value at each time point.

[0128] S1032. Based on the reflection coefficient sequence and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain a seismic record sequence under each set of sample data, so as to obtain a cold spring plume seismic forward modeling result under each set of sample data.

[0129] In this step, based on the reflection coefficient value at each time point, the fluctuation signal and main frequency value at each time point included in the seismic wavelet sequence are substituted into the preset seismic convolution forward expression to obtain the seismic record sequence under each set of sample data, that is, the cold spring plume seismic forward modeling result under each set of sample data is obtained.

[0130] In the embodiment of the present application, the preset seismic convolution forward expression is as follows.

[0131] .

[0132] in, represents a sequence of earthquake records; represents the reflection coefficient sequence; represents the seismic wavelet sequence; Indicates a time offset.

[0133] For example, see Figure 7 , Figure 7 This is an example diagram of the calculation of the cold spring plume seismic forward modeling results provided in the embodiment of the present application. Figure 7 As shown in , when the seismic parameter is a 30 Hz Ricker wavelet sequence, the time domain reflection coefficient model of the cold spring plume is further convolution forward modeled, and the seismic forward modeling results of the cold spring plume under each set of sample data are obtained.

[0134] S104. Based on the noise parameters included in each set of sample data, perform noise addition and regularization normalization processing on the cold spring plume seismic forward modeling results to obtain the cold spring plume seismic response results under each set of sample data.

[0135] After obtaining the cold spring plume seismic forward modeling results through convolution forward modeling, in order to increase the data complexity and authenticity, in an embodiment of the present application, random noise with a certain signal-to-noise ratio is added to the cold spring plume seismic forward modeling results. At the same time, in order to facilitate the use of deep learning methods to extract features of the target cold spring plume morphological model, the cold spring plume seismic forward modeling results after adding noise are also subjected to regularization normalization (Z-score normalization) to eliminate the dimensional influence between different data sets.

[0136] In one implementation of the present application, during specific implementation, step S104 may include:

[0137] S1041. Based on the noise signal-to-noise ratio value in the noise parameters included in each group of sample data, add random noise corresponding to the noise signal-to-noise ratio value to the cold spring plume seismic forward modeling result to obtain a noise processing result.

[0138] In this step, based on the noise signal-to-noise ratio value in the noise parameters included in each set of sample data and the random noise sequence in the noise parameters, random noise corresponding to the noise signal-to-noise ratio value is added to the cold spring plume seismic forward modeling result to obtain the noise processing result.

[0139] Specifically, the expression of the noise processing result is as follows.

[0140] .

[0141] in, represents the noise processing result (i.e., the cold spring plume seismic forward modeling result with random noise corresponding to the noise signal-to-noise ratio value added); represents the reflection coefficient sequence; represents the seismic wavelet sequence; Represents the random noise sequence corresponding to the noise signal-to-noise ratio value.

[0142] S1042, performing mean calculation and variance calculation on the seismic record data in the noise processing result, respectively, to obtain the mean value and variance of the seismic record data corresponding to the noise processing result.

[0143] In this step, for each seismic data in the seismic record data corresponding to the noise processing result, the mean and variance corresponding to each seismic data are calculated respectively.

[0144] S1043. Based on the noise processing result, the mean of the seismic record data and the variance of the seismic record data, a preset regular standardized expression is used to perform calculations to obtain the cold spring plume seismic response results for each set of sample data.

[0145] In this step, the cold spring plume seismic forward modeling results with random noise corresponding to the noise signal-to-noise ratio value added and the calculated mean and variance of the seismic record data are substituted into a preset regular standardized expression to obtain the cold spring plume seismic response results under each group of sample data.

[0146] In the embodiment of the present application, the preset regular standardization expression is as follows.

[0147] .

[0148] in, It represents the seismic response results of cold spring plume; The seismic forward modeling results of cold spring plume with random noise corresponding to the added noise signal-to-noise ratio value are shown; represents the mean value of earthquake record data; Represents the variance of earthquake record data.

[0149] For example, see Figure 8 , Figure 8 This is an example diagram of the calculation of the seismic response results of a cold spring plume provided in an embodiment of the present application. Figure 8 As shown in , the cold spring plume seismic forward modeling results are subjected to noise addition and regularization standardization processing in turn to obtain the cold spring plume seismic response results under each set of sample data.

[0150] S105. Perform corresponding semantic segmentation processing on the cold spring plume seismic response results and the target cold spring plume morphological model to obtain semantic segmentation label samples of the cold spring plume seismic response under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0151] In this step, after performing corresponding semantic segmentation processing on the cold spring plume seismic response results and the target cold spring plume morphological model to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, multiple groups of sample data in the preset sample data set can be used to obtain multiple cold spring plume seismic response semantic segmentation label samples corresponding to the target cold spring plume morphological model to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0152] Furthermore, for each of the multiple cold spring plume morphology models, a corresponding plurality of cold spring plume seismic response semantic segmentation label samples can be obtained through the construction method shown in the embodiments of the present application to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the multiple cold spring plume morphology models.

[0153] For example, see Fig. 9 , Fig. 9 A schematic diagram of label samples of a cold spring plume seismic response semantic segmentation sample library provided in an embodiment of the present application. Fig. 9 As shown in the figure, for 9 cold spring plume morphological models, an example of a set of labeled samples selected from the cold spring plume seismic response semantic segmentation sample library is generated based on the morphological size parameters in the preset sample data set, the bubble content value and bubble radius value included in the bubble parameters, the seawater sound wave velocity, seismic parameters and noise parameters.

[0154] In one embodiment of the present application, in a specific implementation, in step S105, the cold spring plume seismic response result and the target cold spring plume morphological model are subjected to corresponding semantic segmentation processing to obtain the cold spring plume seismic response semantic segmentation label sample under each group of sample data, which may include:

[0155] S1051. Extracting model features from the cold spring plume seismic response results according to the morphological features of the target cold spring plume morphological model.

[0156] S1052: Match the model features with the morphological features to determine a corresponding relationship between the cold spring plume seismic response result and the target cold spring plume morphological model.

[0157] S1053. Based on the corresponding relationship and the cold spring plume seismic response result, semantic segmentation is performed on the target cold spring plume morphological model to obtain a cold spring plume seismic response semantic segmentation label sample under each group of sample data.

[0158] For example, see Fig.10 , Fig.10 This is an example diagram of a cold spring plume seismic response semantic segmentation label sample provided in an embodiment of the present application. Fig.10 As shown in , the cold spring plume seismic response results and the target cold spring plume morphological model are semantically segmented to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data.

[0159] The method for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application utilizes multiple groups of sample data in a preset sample data set to respectively construct a cold spring plume acoustic wave velocity model and a cold spring plume time domain reflection coefficient model under each group of sample data, then synthesizes the target cold spring plume morphological model into seismic data through seismic convolution forward calculation, and performs noise addition and regularization normalization processing on the target cold spring plume seismic response results, thereby improving the quality of the obtained cold spring plume seismic response results, and directly performs semantic segmentation on the cold spring plume seismic response results corresponding to the target cold spring plume morphological model to obtain semantic segmentation label samples of cold spring plume seismic response under each group of sample data, thereby constructing a cold spring plume seismic response semantic segmentation sample library, thereby improving the construction quality and efficiency of the cold spring plume seismic response semantic segmentation samples, and improving the diversity of the cold spring plume seismic response samples.

[0160] See also Fig.11 , Fig.11 A schematic diagram of a structure of a device for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application. Fig.11 As shown in , the construction device 1100 includes:

[0161] The acoustic wave velocity model construction module 1110 is used to calculate the acoustic wave velocity of the cold spring bubble plume under each set of sample data based on the bubble parameters included in each set of sample data in the preset sample data set, and to fuse the acoustic wave velocity of the cold spring bubble plume with the preset seawater acoustic wave velocity distribution map to construct the acoustic wave velocity model of the cold spring plume under each set of sample data;

[0162] A reflection coefficient model building module 1120 is used to calculate based on the water velocity parameters included in the cold spring plume acoustic wave velocity model using a preset reflection coefficient expression, and to build a cold spring plume time domain reflection coefficient model for each set of sample data through time domain processing;

[0163] A seismic convolution forward modeling calculation module 1130 is used to calculate based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data using a preset seismic convolution forward modeling expression to obtain a cold spring plume seismic forward modeling result under each set of sample data;

[0164] A seismic response processing module 1140 is used to perform noise addition and regularization normalization processing on the cold spring plume seismic forward modeling result based on the noise parameters included in each set of sample data, so as to obtain the cold spring plume seismic response result under each set of sample data;

[0165] The semantic segmentation sample library construction module 1150 is used to perform corresponding semantic segmentation processing on the cold spring plume seismic response results and the target cold spring plume morphological model to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

[0166] The device for constructing a semantic segmentation sample library of cold spring plume seismic response provided in an embodiment of the present application uses multiple groups of sample data in a preset sample data set to respectively construct a cold spring plume acoustic wave velocity model and a cold spring plume time domain reflection coefficient model under each group of sample data, then synthesizes the target cold spring plume morphological model into seismic data through seismic convolution forward calculation, and performs noise addition and regularization normalization processing on it, thereby improving the quality of the obtained cold spring plume seismic response results, and directly performs semantic segmentation on the cold spring plume seismic response results corresponding to the target cold spring plume morphological model, and obtains the semantic segmentation label samples of the cold spring plume seismic response under each group of sample data, thereby constructing a cold spring plume seismic response semantic segmentation sample library, thereby improving the construction quality and efficiency of the cold spring plume seismic response semantic segmentation samples, and improving the diversity of the cold spring plume seismic response samples.

[0167] See also Fig.12 , Fig.12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Fig.12 As shown in , the electronic device 1200 includes a processor 1210 , a memory 1220 and a bus 1230 .

[0168] The memory 1220 stores machine-readable instructions executable by the processor 1210. When the electronic device 1200 is running, the processor 1210 communicates with the memory 1220 via the bus 1230. When the machine-readable instructions are executed by the processor 1210, the above-mentioned Figure 1 The steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response in the method embodiment shown, and the specific implementation method can be found in the method embodiment, which will not be repeated here.

[0169] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response in the method embodiment shown, and the specific implementation method can be found in the method embodiment, which will not be repeated here.

[0170] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0171] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0172] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0173] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0174] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0175] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for constructing a semantic segmentation sample library of cold spring plume seismic response, characterized in that: The construction method comprises: For a target cold spring plume morphology model among the pre-constructed multiple cold spring plume morphology models, based on the bubble parameters included in each set of sample data in the preset sample data set, a preset sound velocity expression in a bubble-containing medium is used to calculate, so as to obtain the cold spring bubble plume acoustic wave velocity under each set of sample data, and the cold spring bubble plume acoustic wave velocity is merged with a preset seawater acoustic wave velocity distribution map to construct a cold spring plume acoustic wave velocity model under each set of sample data; Based on the water velocity parameters included in the cold spring plume acoustic wave velocity model, a preset reflection coefficient expression is used for calculation, and a cold spring plume time domain reflection coefficient model under each set of sample data is constructed through time domain processing; Based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain the cold spring plume seismic forward modeling results under each set of sample data; Based on the noise parameters included in each set of sample data, the cold spring plume seismic forward modeling results are subjected to noise addition and Z-score normalization processing to obtain the cold spring plume seismic response results under each set of sample data; The cold spring plume seismic response results and the target cold spring plume morphological model are subjected to corresponding semantic segmentation processing to obtain the cold spring plume seismic response semantic segmentation label samples under each group of sample data, so as to construct a cold spring plume seismic response semantic segmentation sample library corresponding to the target cold spring plume morphological model.

2. The method according to claim 1, characterized in that The various cold seep plume morphology models are pre-built by following the steps below: Based on the preset size parameters, construct a morphological model background layer, and set the pixel value of the morphological model background layer to a first preset value; Based on the morphological size parameters corresponding to the preset multiple cold spring plume morphologies, multiple cold spring plume areas are constructed in the morphological model background layer, and the pixel value of the cold spring plume area is set to a second preset value to construct multiple cold spring plume morphological models.

3. The method according to claim 1, characterized in that: The bubble parameters included in each set of sample data in the preset sample data set are calculated using a preset sound velocity expression in a bubble-containing medium to obtain the cold spring bubble plume sound wave velocity under each set of sample data, including: Based on the bubble parameters included in each set of sample data in the preset sample data set, using the preset random medium theory and plume flow distribution law, a random disturbance value corresponding to the bubble parameter is generated to achieve random changes in the bubble content value and the bubble radius value included in the bubble parameter; Based on the random disturbance value corresponding to the bubble parameter and the preset gas-liquid parameter, the preset sound velocity expression in the bubble-containing medium is used for calculation to obtain the sound wave velocity of the cold spring bubble plume under each set of sample data.

4. The method according to claim 2, characterized in that: The step of fusing the cold seepage bubble plume acoustic wave velocity with a preset seawater acoustic wave velocity distribution map to construct a cold seepage plume acoustic wave velocity model under each set of sample data includes: Based on a preset seawater acoustic wave velocity distribution map, constructing a seawater acoustic wave velocity background layer corresponding to the target cold seep plume flow morphology model; Based on the target cold seep plume area in the target cold seep plume morphology model, the acoustic wave velocity of the cold seep bubble plume is fused into the seawater acoustic wave velocity background layer to construct the cold seep plume acoustic wave velocity model under each set of sample data.

5. The method according to claim 1, characterized in that The water velocity parameters included in the cold spring plume acoustic wave velocity model are calculated using a preset reflection coefficient expression, and the cold spring plume time domain reflection coefficient model under each set of sample data is constructed through time domain processing, including: Based on the cold spring plume acoustic wave velocity model, determining a water velocity parameter included in the cold spring plume acoustic wave velocity model; Based on the water velocity value corresponding to each bottom interface included in the water velocity parameter, a preset reflection coefficient expression is used to calculate to obtain the reflection coefficient corresponding to each bottom interface; Based on the reflection coefficient, the depth value of the bottom interface of each layer is corresponded to a time point to construct a time domain reflection coefficient model of the cold spring plume under each set of sample data.

6. The method according to claim 1, characterized in that The cold spring plume seismic forward modeling result under each set of sample data is obtained by calculating based on the cold spring plume time domain reflection coefficient model and the seismic parameters included in each set of sample data using a preset seismic convolution forward modeling expression, including: Determining the reflection coefficient sequence under each set of sample data in the cold spring plume time domain reflection coefficient model; Based on the reflection coefficient sequence and the seismic parameters included in each set of sample data, a preset seismic convolution forward expression is used to perform calculations to obtain a seismic record sequence under each set of sample data, so as to obtain a cold spring plume seismic forward modeling result under each set of sample data.

7. The method according to claim 1, characterized in that Based on the noise parameters included in each set of sample data, the cold spring plume seismic forward modeling results are subjected to noise addition and Z-score normalization processing to obtain the cold spring plume seismic response results under each set of sample data, including: Based on the noise signal-to-noise ratio value in the noise parameter included in each group of sample data, adding random noise corresponding to the noise signal-to-noise ratio value to the cold spring plume seismic forward modeling result, to obtain a noise processing result; Performing mean calculation and variance calculation on the seismic record data in the noise processing result respectively to obtain the mean value and variance of the seismic record data corresponding to the noise processing result; Based on the noise processing result, the mean value of the seismic record data and the variance of the seismic record data, a preset Z-score standardized expression is used for calculation to obtain the cold spring plume seismic response result under each group of sample data.

8. The method according to claim 1, characterized in that The step of performing corresponding semantic segmentation processing on the cold spring plume seismic response result and the target cold spring plume morphological model to obtain a cold spring plume seismic response semantic segmentation label sample under each group of sample data includes: extracting model features from the cold spring plume seismic response results according to the morphological features of the target cold spring plume morphological model; Matching the model features with the morphological features to determine a corresponding relationship between the cold spring plume seismic response result and the target cold spring plume morphological model; Based on the corresponding relationship and the seismic response result of the cold spring plume, the target cold spring plume morphological model is semantically segmented to obtain the semantic segmentation label samples of the seismic response of the cold spring plume under each group of sample data.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for constructing a semantic segmentation sample library of cold spring plume seismic response according to any one of claims 1 to 8 are executed.

Images