Braid river delta training image generation method and device

Abstract

The invention provides a braided river delta training image generation method and device, and the method comprises the steps: S1, constructing a river channel center line of an underwater diversion river channel of a braided river delta; s2, constructing a riverway bifurcation based on the riverway midline; s3, depicting a main river channel and a secondary river channel based on the river channel bifurcation; and S4, writing the main river channel and the secondary river channel into a three-dimensional grid, and outputting a three-dimensional grid model image of the braided river delta with a preset scale according to a preset river channel phase proportion. The invention aims to solve the problem of high difficulty in acquiring a braid river delta training image, and how to more truly represent spatial structure characteristics of a split river channel under a petal river delta training image.

Description

Technical Field

[0001] The invention belongs to the field of geological exploration and relates to a method and device for generating a braided river delta training image. Background Art

[0002] Multi-point geostatistics, a geological modeling method first proposed in 1993, struggled to be used in practical modeling due to its low simulation efficiency. It was not until the introduction of the "search tree" structure in 2002 that the method gained widespread application. This method is more suitable for fluvial sedimentary systems than traditional two-point geostatistics. Despite its widespread application in reservoir modeling, multi-point geostatistics suffers from a significant limitation: the difficulty in obtaining training images, particularly 3D training images.

[0003] The training image is a priori geological model that replaces the variogram used in traditional two-point geostatistics as a tool for measuring reservoir heterogeneity and plays a crucial role in reservoir prediction. The training image is a digital image that can represent the actual reservoir structure, geometry, and distribution pattern. It is a conceptual model that reflects prior geological concepts and other geological characteristics of the reservoir.

[0004] To date, there is no mature, unified method for establishing training images for multi-point geostatistical methods. Acquisition of training images relies heavily on geologists' guesswork, resulting in significant uncertainty. Furthermore, two-dimensional training images are difficult to represent the spatial structure of three-dimensional river channels, and acquiring three-dimensional training images is even more challenging. This limits the application of multi-point geostatistical methods for characterizing braided river channels. Braided river deltas are formed by sediment deposited by braided rivers as they enter the sea or lake. Currently, no mature, unified method has been found for establishing training images for braided river deltas. Summary of the Invention

[0005] The present invention provides a method and device for generating braided river delta training images to solve the problem of difficulty in obtaining braided river delta training images and how to more realistically characterize the spatial structural characteristics of distributary channels under the petal-shaped river delta training images.

[0006] The present invention adopts the following technical solutions:

[0007] The present invention provides a method for generating a braided river delta training image, comprising:

[0008] Step S1: constructing the centerline of the underwater distributary channel of the braided river delta;

[0009] Step S2: constructing a river bifurcation based on the river centerline;

[0010] Step S3: Delineating the main river channel and the secondary river channel based on the river channel bifurcation;

[0011] Step S4: writing the main channel and the secondary channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset channel phase ratio.

[0012] Optionally, step S1: constructing the centerline of an underwater distributary channel in a braided river delta includes:

[0013] Constructing a vector line of the center line of the river channel according to the initial point, azimuth, control point spacing, maximum offset distance, and total length of the river channel;

[0014] The vector lines are smoothed and encrypted to obtain a river centerline node set.

[0015] Optionally, step S2: constructing a river bifurcation based on the river centerline includes:

[0016] A plurality of nodes are selected from the center line of the river channel, and the river channel bifurcations are constructed with each node as a starting point according to the average bifurcation distance, bifurcation angle, and direction of the bifurcation river channel.

[0017] Optionally, step S3: delineating a main river channel and a secondary river channel based on the river channel bifurcation includes:

[0018] Determining the main river channel in the river bifurcation according to the river channel width and thickness;

[0019] And the secondary river channel in the river channel bifurcation is determined according to the frequency of river channel bifurcation and the average bifurcation distance.

[0020] Optionally, step S3: delineating the main river channel and the secondary river channel based on the river channel bifurcation further includes:

[0021] By traversing the river bifurcations one by one, it is determined whether the current river bifurcation vector and the previous river bifurcation vector have an intersection. If an intersection exists, the two river bifurcation vectors form a river cutoff.

[0022] Optionally, step S4: writing the main river channel and the secondary river channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio includes:

[0023] The size of the actual image of the braided river delta is associated with a network model that matches the number of grids and the size of the grids to obtain an associated three-dimensional grid model of the braided river delta.

[0024] Optionally, step S4 further includes:

[0025] Assigning and distinguishing the main channel phase, secondary channel phase and background phase in the associated three-dimensional grid model of the braided river delta to obtain the differentiated three-dimensional grid model of the braided river delta;

[0026] Determining, according to a preset river channel phase ratio, a three-dimensional grid model image of the braided river delta of a preset scale that meets the requirements of the three-dimensional grid model of the braided river delta after the assignment and differentiation;

[0027] Output the three-dimensional mesh model image.

[0028] The present invention also provides a braided river delta training image generation device, comprising:

[0029] The river centerline construction module is used to construct the river centerline of the underwater distributary channel of the braided river delta;

[0030] A river bifurcation construction module, configured to construct a river bifurcation based on the river centerline;

[0031] A river channel division module, used for delineating a main river channel and a secondary river channel based on the river channel bifurcation;

[0032] The output module is used to write the main river channel and the secondary river channel into a three-dimensional grid, and output a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio.

[0033] The present invention also provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for generating a braided river delta training image described in any one of the above embodiments are implemented.

[0034] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method for generating a braided river delta training image as described in any one of the above embodiments.

[0035] The beneficial effects of the present invention are:

[0036] The present invention mainly constructs a three-dimensional grid model of a braided river delta, which mainly depicts the simulation of two types of underwater distributary channels of braided river deltas, namely the main channel and the secondary channel. It can also output the three-dimensional grid model image of the braided river delta required by the user according to the parameters input by the user, solving the problem of difficulty in obtaining training images of braided river deltas. The braided river delta training images provided by the present invention can more realistically characterize the spatial structural characteristics of the underwater distributary channels of braided river deltas. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart of a method for generating a braided river delta training image provided by the present invention;

[0038] Figure 2A schematic diagram of river centerline initialization provided by the present invention;

[0039] Figure 3 A schematic diagram of a river centerline densification provided by the present invention;

[0040] Figure 4 A schematic diagram of a river bifurcation diagram provided by the present invention;

[0041] Figure 5 A schematic diagram of a main and secondary river channel and a river channel cutoff diagram provided by the present invention;

[0042] Figure 6 A schematic diagram of a grid splitting plane provided by the present invention;

[0043] Figure 7 A schematic diagram of a unit grid assignment on a plane provided by the present invention;

[0044] Figure 8 A schematic diagram of unit grid assignment on a river section provided by the present invention;

[0045] Figure 9 A schematic diagram of simulation results of Example 1 provided by the present invention;

[0046] Figure 10 A schematic diagram of simulation results of Example 2 provided by the present invention;

[0047] Figure 11 A schematic diagram of simulation results of Example 3 provided by the present invention;

[0048] Figure 12 Schematic diagram of a braided river delta training image generation device provided by the present invention. DETAILED DESCRIPTION

[0049] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] The present invention provides a method for generating a braided river delta training image. Figure 1 The method shown includes:

[0051] Step S1: constructing the centerline of the underwater distributary channel of the braided river delta;

[0052] Optionally, step S1: constructing the centerline of an underwater distributary channel in a braided river delta includes:

[0053] Constructing a vector line of the center line of the river channel according to the initial point, azimuth, control point spacing, maximum offset distance, and total length of the river channel;

[0054] The vector lines are smoothed and encrypted to obtain a river centerline node set.

[0055] In one embodiment, the present invention uses a point set to represent vector lines. This involves collecting a set of spatial point coordinates along the centerline of an underwater distributary channel in a braided river delta. These coordinates are then used to construct a vector line, quantitatively representing the centerline. Once the vector line is determined, the spatial position and basic shape of a river channel can be determined.

[0056] The generation of river centerline includes five parameters, namely initial point, azimuth, control point spacing, maximum offset distance, and total length. Figure 2 The following example illustrates the meaning of each parameter. The initial point is the centerline of the river, i.e., the first point of the vector line. The hollow circle in the figure represents the position of the initial point. The azimuth angle θ represents the main direction of the river channel extension. With the initial point as the coordinate origin, the east (E) direction is 0 degrees, and the angle corresponding to the main direction rotated counterclockwise is the azimuth angle θ. The control point spacing represents the offset spacing of the river channel control nodes in the main direction. Figure 2 The distance between the scale marks in the main direction is the control point spacing. A control point is generated at regular intervals along the main direction of the river. The maximum offset distance represents the maximum distance from a control point in the main direction, constraining control points to lie within a certain range on either side of the main direction axis. Along the main direction of the river, a control point is generated at every control point spacing. The distance a control point deviates from the main direction is sampled between the negative and positive maximum offset distances, typically generated using a random function.

[0057] The total length indicates the maximum extent of the river channel. When the distance between the generated river channel control point and the initial point is greater than the total length of the river channel, the generation of the next river channel control point will be stopped. At this point, the initial control node of the river channel centerline is generated. Based on the initial control node of the river channel centerline, the vector line is smoothed and encrypted to obtain the river channel centerline node set as shown below: Figure 3 As shown, the node set is the river vector line finally required.

[0058] Step S2: constructing a river bifurcation based on the river centerline;

[0059] Optionally, step S2: constructing a river bifurcation based on the river centerline includes:

[0060] A plurality of nodes are selected from the center line of the river channel, and the river channel bifurcations are constructed with each node as a starting point according to the average bifurcation distance, bifurcation angle, and direction of the bifurcation river channel.

[0061] In one embodiment, deltaic deposits are usually formed at the mouth of a river, diverging radially. A river channel will bifurcate at the mouth of the sea, forming an underwater diversion channel. The river channel continues to bifurcate to form a deltaic channel network. The channel bifurcation contains three parameters: average bifurcation distance, bifurcation angle, and direction of the bifurcation channel. Take multiple nodes in the center line of the river channel, take each node as the starting point, and construct the channel bifurcation according to the average bifurcation distance, bifurcation angle, and direction of the bifurcation channel. Figure 4 As shown in the figure, the angle between the direction of Channel 1 and the direction of Channel 2 is the bifurcation angle. The specific value of the bifurcation angle is specified by the user to control the parameter input. The directions of Channel 1 and Channel 2 are obtained by offsetting the main direction of the unbranched channel counterclockwise and clockwise by half the bifurcation angle, respectively. The average bifurcation distance is the average distance between the first and second bifurcations of the channel and is used to control the frequency of channel bifurcation. A shorter average distance indicates a more frequent channel bifurcation, while a longer average distance indicates a less frequent channel bifurcation.

[0062] Step S3: Delineating the main river channel and the secondary river channel based on the river channel bifurcation;

[0063] Optionally, step S3: delineating the main river channel and the secondary river channel based on the river channel bifurcation includes:

[0064] Determining the main river channel in the river bifurcation according to the river channel width and thickness;

[0065] And the secondary river channel in the river channel bifurcation is determined according to the frequency of river channel bifurcation and the average bifurcation distance.

[0066] In one embodiment, two types of river channels are defined, namely, the main river channel and the secondary river channel. The main river channel is larger than the secondary river channel. Specifically, the main river channel and the secondary river channel can be distinguished in terms of the width and thickness of the river channel. When a river flows into the sea or a lake, the river channel will branch step by step to form a delta river network. When the river channel branches, the scales of the two branched river channels may be different, thus forming the main and secondary river channels. Figure 5 As shown in the figure, when the river channel is close to the estuary, the river channel is large and the main channel is usually developed; when the river channel gradually bifurcates and the distance from the estuary becomes farther and farther, the river channel scale decreases and the bifurcated river channel becomes a secondary channel. Figure 5 The solid line represents the main river channel vector line, and the dotted line represents the secondary river channel vector line.

[0067] Two parameters are defined for describing a secondary channel: the starting location of the secondary channel and the probability of its occurrence. The starting location of the secondary channel is pre-set; a secondary channel will only appear if the channel's starting point exceeds this location. Secondly, the probability of a secondary channel appearing at each starting point is calculated. This determines the probability of a secondary channel appearing at each bifurcation. A corresponding random function is set here. If the probability of a secondary channel appearing at a point is greater than 0.5, a secondary channel appears at that point; if it is less than 0.5, no secondary channel appears at that point.

[0068] Furthermore, the following constraints are imposed on the rules for river bifurcations: Rivers can only bifurcate from large to small. That is, a main river can bifurcate into a main channel or a secondary channel, and a secondary channel can only bifurcate into a secondary channel. A river can only bifurcate into two secondary sub-channels at a time.

[0069] Each river channel has a certain probability of forking at a certain distance. The distance and probability are input by the user through parameters to control the frequency of river channel forking. The parameters of the main river channel and the secondary river channel are different and are input by the user separately.

[0070] Optionally, step S3: delineating the main river channel and the secondary river channel based on the river channel bifurcation further includes:

[0071] By traversing the river bifurcations one by one, it is determined whether the current river bifurcation vector and the previous river bifurcation vector have an intersection. If an intersection exists, the two river bifurcation vectors form a river cutoff.

[0072] It should be noted that there may be a situation where the river channel is cut off between the main channel and the secondary channel. The estuary is used as the initial point of the first river channel, and the river channel gradually extends and bifurcates along the given main direction. If a bifurcated river channel vector line intersects, the river channel needs to be cut off. The truncation rule is that the later generated river channel will pinch out towards the existing river channel. After the river channel bifurcates, the river channel before the bifurcation will be frozen and added to the channel set. The last control node of the river channel before the bifurcation will be used as the initial point of the sub-channel, and the bifurcated river channel will continue to be simulated downward. When a river channel exceeds the preset delta range during the bifurcation process, the river channel will stop bifurcation, freeze the river channel and add it to the channel set. When all river channels extend beyond the preset range, the simulation ends, and a set of river channels, that is, a set of river channel vector lines, will be obtained. This set quantitatively expresses the channel network structure of the braided river delta. By combining the channel vector line set with the channel geological parameters and discretizing them into the three-dimensional grid of the training image, a digital channel grid model can be obtained, and the final required braided river delta training image can be obtained.

[0073] Step S4: writing the main channel and the secondary channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset channel phase ratio.

[0074] Optionally, step S4: writing the main river channel and the secondary river channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio includes:

[0075] The size of the actual image of the braided river delta is associated with a network model that matches the number of grids and the size of the grids to obtain an associated three-dimensional grid model of the braided river delta.

[0076] In one embodiment, a complete braided river delta channel network is simulated, and the training image is the output of a three-dimensional grid model of a braided river delta, and the channel vector lines need to be associated with the grid model. Figure 6 As shown in the figure, a 3D grid model is defined within the braided river delta. All river channel vector lines within the 3D grid are calculated with the 3D grid, and the channel information is written to the 3D grid. The actual size of the training image is determined by the number of grid cells and the grid size, which are defined by the user.

[0077] Optionally, step S4 further includes: determining a three-dimensional grid model image of the braided river delta of a preset scale that meets the requirements of the three-dimensional grid model of the braided river delta after the assignment and differentiation according to a preset river phase ratio; and outputting the three-dimensional grid model image.

[0078] Assigning and distinguishing the main channel phase, secondary channel phase and background phase in the associated three-dimensional grid model of the braided river delta to obtain the differentiated three-dimensional grid model of the braided river delta;

[0079] Specifically, the braided river 3D training image contains three phase types, namely the main channel phase, the secondary channel phase, and the background phase. After defining the 3D grid, we need to assign a value to each cell grid. According to the positional relationship between the cell grid and the main and secondary channels, when the cell is in the main channel, it is assigned to the main channel phase; when the cell is in the secondary channel, it is assigned to the secondary channel phase; otherwise, it is assigned to the background phase. Figure 7 As shown, the solid line represents the center line of the river channel, and the distance between the dotted line and the solid line represents the maximum width of the river channel on the plane. The width is input by the user through parameters. When the center point of the grid is within the dotted line, it is considered that the grid is inside the river channel, and the unit grid is assigned the river channel phase. Depending on the type of the current river channel, the specific assignment is determined to be the main river channel or the secondary river channel. Since the training image is a three-dimensional grid model, after the unit grid is judged on the plane, it is necessary to further judge based on the position of the unit grid on the cross section. Figure 8As shown in the figure, a river channel has a flat top and convex bottom in cross section. The specific shape is determined by the geological parameters of the channel, including the width (top width) and thickness (maximum thickness). For each cell that is located within the channel in both the plane and cross section, the cell is assigned the channel phase. The specific assignment is determined by the current channel type, whether it is a primary channel or a secondary channel.

[0080] Further determine the specifics based on the position of the unit grid on the cross section:

[0081] A river channel appears flat on top and convex on the bottom in cross-section. Its specific shape is determined by the channel's geological parameters, including its width (top width) and thickness (maximum thickness). For each cell, if both the plane and cross-section view lie within the channel, the cell is assigned the channel facies.

[0082] Regarding the Fluvsim modeling method, the target-based modeling method (Fluvsim) designed by CV Deutsch can effectively characterize the genetic relationships between different reservoir structural units and is an important fluvial reservoir modeling method. Unlike traditional target-based methods, the Fluvsim method has the following characteristics: 1. A clear and reversible coordinate hierarchy system; 2. Input parameters that reflect geological significance to control the target body geometry; 3. Accurate control of vertical phase ratios; 4. True asymmetric channel geometry; 5. True non-fluctuating channel top surface. The ideal channel model portrayed by the Fluvsim method includes three sedimentary facies types in addition to the background facies: channel fill, natural levee, and breach fan. The model diagram fully reflects the vertical and horizontal superposition and combination of various sedimentary bodies. The target body geometry of each facies type can be controlled by parameter files. The parameters are generally determined through geological research, mainly including relevant data such as field outcrops, seismic data, well logging, sedimentary simulation experiments, and geological knowledge bases.

[0083] The calculation method for judging whether the grid is within the river channel based on the Fluvsim profile modeling method is as follows:

[0084]

[0085] Among them, w(y) is the width of the river, c v (y) is the local curvature, is the maximum curvature.

[0086] When a(y)≤0.5, the depth of the channel base below the channel top is calculated as:

[0087]

[0088] In the formula, b(y)=-ln(2) / ln(a(y)),w∈[0,w(y)]w∈[0,w(y)],

[0089] When a(y)>0.5, the depth of the channel base below the channel top is calculated as:

[0090]

[0091] In the formula, c(y)=-ln(2) / ln(1-a(y))

[0092] The above judgment method can be used to determine whether the grid is within the river channel, thereby assigning a value to the grid.

[0093] Define a 3D network model for the training image. Once a complete braided river delta channel vector line is obtained, a phase determination can be performed on each cell in the 3D grid model. After traversing all cells, a single-phase delta channel training image is created. The generation of the 3D training image requires that the channel phase ratio be met. Only when the channel cell ratio within the 3D grid reaches a preset ratio will the resulting 3D network model meet the required training image. Therefore, parameters must be adjusted and braided river delta channel networks of varying sizes repeatedly simulated. Each time a complete braided river delta vector line is generated, the 3D grid model is assigned a value until the channel ratio reaches the preset ratio, at which point the 3D grid model image is output.

[0094] Based on the braided river delta training image generation method embodiment provided by the present invention, the present invention also provides a set of simulation result embodiments:

[0095] Example 1: Set the delta range to a rectangular area of ​​4000 meters in length and 5000 meters in width. The algorithm parameters are set as follows: the main river bifurcation angle is 20-30 degrees; the average bifurcation distance of the main river is 500 meters; the secondary river begins to appear at a distance of 1500 meters from the starting point of the first river (the estuary); the secondary river generation probability is 0.5; the secondary river bifurcation angle is 20-30 degrees, and the average bifurcation distance of the secondary river is 500 meters. Only one set of braided river delta channel networks is simulated, and the planar situation of the channel network is input and displayed. The simulation results are as follows: Figure 9 shown.

[0096] Example 2: Set the delta range to a rectangular area with a length of 4000 meters, a width of 5000 meters, and a thickness of 10 meters. The algorithm parameters are set as follows: the bifurcation angle of the main river channel is 20-30°; the average bifurcation distance of the main river channel is 500 meters, the starting position of the secondary river channel: 1500 meters from the starting point of the first river channel (estuary), the probability of secondary river channel generation is 0.5; the bifurcation angle of the secondary river channel is 20-30°, and the average bifurcation distance of the secondary river channel is 500 meters. The main river channel is 120 meters wide and has a maximum thickness of 8 meters, and the secondary river channel is 60 meters wide and has a maximum thickness of 4 meters. Only one set of braided river delta channel networks is simulated, and the channel network is a collection of channel centerlines. Define the number of three-dimensional grids as 250*200*10, the grid size is 5*5*1, and the grid range covers the entire delta range. According to the braided river delta channel network, all grids are assigned values ​​to obtain a braided river delta training image. The simulation results are as follows Figure 10 shown.

[0097] Example 3: Set the delta range to a rectangular area of ​​4000 meters in length, 5000 meters in width and 10 meters in thickness. The algorithm parameters are set as follows: the bifurcation angle of the main river channel is 20-30°; the average bifurcation distance of the main river channel is 500 meters, the starting position of the secondary river channel is 1500 meters from the starting point of the first river channel (estuary), and the probability of secondary river channel generation is 0.5; the bifurcation angle of the secondary river channel is 20-30°, and the average bifurcation distance of the secondary river channel is 500 meters. The main river channel is 120 meters wide and has a maximum thickness of 8 meters, and the secondary river channel is 60 meters wide and has a maximum thickness of 4 meters. The number of three-dimensional grids of the training image is defined as 460*90*10, and the grid size is 5*5*1. The channel phase ratio is preset to 80%, and the braided river delta channel network is repeatedly simulated and assigned to the defined training image grid until the channel phase ratio reaches 80%. Finally, the braided river delta training image is obtained. The simulation results are as follows Figure 11 shown.

[0098] The present invention also provides a braided river delta training image generation device, such as Figure 12 As shown, the device includes:

[0099] The river centerline construction module is used to construct the river centerline of the underwater distributary channel of the braided river delta;

[0100] A river bifurcation construction module, configured to construct a river bifurcation based on the river centerline;

[0101] A river channel division module, used for delineating a main river channel and a secondary river channel based on the river channel bifurcation;

[0102] The output module is used to write the main river channel and the secondary river channel into a three-dimensional grid, and output a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio.

[0103] It should be noted that the present device embodiment can implement a braided river delta training image generation method that is completely consistent with the above method embodiment, which will not be described in detail here.

[0104] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of a braided river delta training image generation method described in any one of the above embodiments when executing the computer program.

[0105] The present invention also provides a computer-readable storage medium, which stores a computer program, characterized in that when the computer program is executed by a processor, it implements the steps of the method for generating a braided river delta training image as described in any one of the above embodiments.

Claims

1. A method for generating a braided river delta training image, characterized in that: include: Step S1: constructing the centerline of the underwater distributary channel of the braided river delta; Step S2: constructing a river bifurcation based on the river centerline; Step S3: Delineating the main river channel and the secondary river channel based on the river channel bifurcation; Step S4: writing the main channel and the secondary channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset channel phase ratio.

2. The method for generating a braided river delta training image according to claim 1, wherein: Step S1: Constructing the centerline of the underwater distributary channel of the braided river delta includes: Constructing a vector line of the center line of the river channel according to the initial point, azimuth, control point spacing, maximum offset distance, and total length of the river channel; The vector lines are smoothed and encrypted to obtain a river centerline node set.

3. The method for generating a braided river delta training image according to claim 1, wherein: Step S2: constructing a river bifurcation based on the river centerline includes: A plurality of nodes are selected from the center line of the river channel, and the river channel bifurcations are constructed with each node as a starting point according to the average bifurcation distance, bifurcation angle, and direction of the bifurcation river channel.

4. The method for generating a braided river delta training image according to claim 1, wherein: Step S3: Delineating the main river channel and the secondary river channel based on the river channel bifurcation includes: Determining the main river channel in the river bifurcation according to the river channel width and thickness; And the secondary river channel in the river channel bifurcation is determined according to the frequency of river channel bifurcation and the average bifurcation distance.

5. The method for generating a braided river delta training image according to claim 4, wherein: Step S3: Delineating the main river channel and the secondary river channel based on the river channel bifurcation further includes: By traversing the river bifurcations one by one, it is determined whether the current river bifurcation vector and the previous river bifurcation vector have an intersection. If an intersection exists, the two river bifurcation vectors form a river cutoff.

6. The method for generating a braided river delta training image according to claim 1, wherein: Step S4: writing the main river channel and the secondary river channel into a three-dimensional grid, and outputting a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio, including: According to the size of the actual image of the braided river delta and the matching The network models of grid number and grid size are associated to obtain the associated three-dimensional grid model of braided river delta.

7. The method for generating a braided river delta training image according to claim 6, wherein: Also includes: Assigning and distinguishing the main channel phase, secondary channel phase and background phase in the associated three-dimensional grid model of the braided river delta to obtain the differentiated three-dimensional grid model of the braided river delta; Determining a three-dimensional grid model image of the braided river delta of a preset scale that meets the requirements of the three-dimensional grid model of the braided river delta after differentiation according to a preset river channel phase ratio; Output the three-dimensional mesh model image.

8. A braided river delta training image generation device, characterized in that: include: The river centerline construction module is used to construct the river centerline of the underwater distributary channel of the braided river delta; A river bifurcation construction module, configured to construct a river bifurcation based on the river centerline; A river channel division module, used for delineating a main river channel and a secondary river channel based on the river channel bifurcation; The output module is used to write the main river channel and the secondary river channel into a three-dimensional grid, and output a three-dimensional grid model image of a braided river delta of a preset scale according to a preset river channel phase ratio.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for generating a braided river delta training image as claimed in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for generating a braided river delta training image as claimed in any one of claims 1 to 7 are implemented.

Images