Large-scale water body underwater terrain inversion method, equipment and medium

Abstract

The invention provides a large-scale water body underwater terrain inversion method and device and a medium, and the method comprises the steps: carrying out the partitioning of a to-be-inverted water body region, and obtaining a deep water region and a land-water conversion shallow water region; based on a spatial interpolation method, carrying out inversion on the land-water transformation shallow water area to obtain a land-water transformation shallow water area topographic image; a deepwater area water depth inversion model constructed based on the wave band reflectivity information and the underwater topographic elevation data performs inversion on the deepwater area to obtain a deepwater area underwater topographic image; and carrying out image splicing fusion on the land and water transformation shallow water region topographic image and the deep water region underwater topographic image to obtain the underwater topography of the water body to be inverted. In the face of a complex and changeable water body environment, a water body area to be inverted is divided into a deep water area and a land and water conversion shallow water area, and different and more targeted inversion methods can be adopted according to the characteristics of water bodies in different areas.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater terrain simulation, and specifically provides a method, equipment and medium for inversion of underwater terrain of a large water body. Background Art

[0002] High-quality underwater topographic data is the basis for hydrological simulation, flood control calculation, shipping scheduling and digital twin basin construction. For a long time, underwater topographic data acquisition has relied on traditional measurement methods such as artificial sounding, sonar sounding, and radar sounding. However, these methods are expensive, require on-site operations, and have limited measurement ranges. They are not suitable for the rapid acquisition of underwater topographic data of large water bodies, especially water bodies of great strategic significance, where it is even more difficult to directly obtain real topographic data.

[0003] Therefore, the field needs a new large-scale water body underwater terrain inversion solution to solve the above problems. Summary of the invention

[0004] In order to overcome the above-mentioned defects, the present invention is proposed to provide a solution or at least partially solve the problem of being unable to accurately and effectively invert the underwater topography of large water bodies.

[0005] In a first aspect, the present invention provides a method for inverting underwater topography of a large water body, comprising: partitioning the water body area to be inverted to obtain deep water areas and water-land conversion shallow water areas; inverting the water-land conversion shallow water areas based on a spatial interpolation method to obtain a water-land conversion shallow water area topography image; inverting the deep water area based on a deep water area depth inversion model constructed based on band reflectivity information and underwater topography elevation data to obtain an underwater topography image of the deep water area; stitching and fusing the water-land conversion shallow water area topography image and the deep water area underwater topography image to obtain the underwater topography of the water body to be inverted.

[0006] In one technical solution of the above-mentioned large-scale water body underwater topography inversion method, the water body area to be inverted is divided into zones to obtain deep water areas and shallow water areas converted from water to land, including: obtaining water level change characteristics of the water body area to be inverted; and determining deep water areas and shallow water areas converted from water to land based on the water level change characteristics.

[0007] In one technical solution of the above-mentioned large-scale water body underwater topography inversion method, the water-land transformation shallow water area is inverted based on the spatial interpolation method to obtain the topographic image of the water-land transformation shallow water area, and the process includes: using the improved normalized difference water body index to extract the water body boundary of the water-land transformation shallow water area to distinguish the shallow water area and the land; obtaining the water level line of the shallow water area; based on the water level line of the shallow water area, using the Kriging interpolation method to invert the underwater topography map of the shallow water area.

[0008] In a technical solution of the above-mentioned large water body underwater topography inversion method, the method includes at least the following steps to obtain a deep water area water depth inversion model: obtaining the relationship between band reflectivity information and underwater topography elevation data; and constructing a deep water area water depth inversion model based on the relationship between band reflectivity information and underwater topography elevation data.

[0009] In a technical solution of the above-mentioned method for inverting underwater terrain of large water bodies, the deep water area water depth inversion model includes at least one or more of a single band model, a band ratio model, a neural network model and a random forest model. The process of constructing a deep water area water depth inversion model based on the relationship between band reflectivity information and underwater terrain elevation data includes: determining the optimal water depth factor based on the correlation between the band reflectivity and the underwater terrain elevation, performing regression analysis on the optimal water depth factor, fitting the single band and band ratio water depth factors to obtain a single band model and a band ratio model; using the band reflectivity information as input features of the neural network model and the random forest model to learn and predict the underwater terrain elevation, and obtaining the neural network model and the random forest model.

[0010] In one technical solution of the above-mentioned large-scale water body underwater terrain inversion method, the optimal water depth factor is determined based on the correlation between the band reflectivity and the underwater terrain elevation, a regression analysis is performed on the optimal water depth factor, and the single-band and band ratio water depth factors are fitted to obtain the single-band model and the band ratio model. The process includes at least: calculating the correlation between the single-band reflectivity and the underwater terrain elevation, and taking the single band with the highest correlation with the underwater terrain elevation as the single-band optimal water depth factor; performing a regression analysis on the single-band optimal water depth factor, and fitting the single-band optimal water depth factor using linear function, exponential function, logarithmic function, power function, and polynomial function respectively, and selecting the one with the best fitting effect to obtain the single-band model.

[0011] In one technical solution of the above-mentioned large-scale water body underwater terrain inversion method, the optimal water depth factor is determined based on the correlation between the band reflectivity and the underwater terrain elevation, the optimal water depth factor is subjected to regression analysis, and the single-band and band ratio water depth factors are fitted to obtain the single-band model and the band ratio model. The process also includes: obtaining different band ratio combinations, calculating the correlation between all band ratios and the underwater terrain elevation, and finding the band ratio combination with the highest correlation as the band ratio optimal water depth factor; performing regression analysis on the band ratio optimal water depth factor, and fitting the band ratio optimal water depth factor using linear function, exponential function, logarithmic function, power function, and polynomial function respectively, and selecting the one with the best fitting effect to obtain the band ratio model.

[0012] In a second aspect, an electronic device includes a processor and a storage device, the storage device is suitable for storing multiple program codes, and the program codes are suitable for being loaded and run by the processor to execute a large water body underwater topography inversion method of any one of the technical solutions of the above-mentioned large water body underwater topography inversion method.

[0013] In a third aspect, the present invention provides a computer-readable storage medium storing a plurality of program codes, wherein the program codes are suitable for being loaded and run by a processor to execute a large water body underwater topography inversion method of any one of the technical solutions of the above-mentioned large water body underwater topography inversion method.

[0014] The above one or more technical solutions of the present invention have at least one or more of the following beneficial effects:

[0015] In the technical scheme for implementing the present invention, the present invention provides a method for inverting underwater topography of a large water body, including: partitioning the water body area to be inverted to obtain deep water areas and shallow water areas converted from water to land; inverting the shallow water area converted from water to land based on the spatial interpolation method to obtain a terrain image of the shallow water area converted from water to land; inverting the deep water area based on a water depth inversion model of the deep water area constructed based on band reflectivity information and underwater terrain elevation data to obtain an underwater terrain image of the deep water area; performing image splicing and fusion of the terrain image of the shallow water area converted from water to land and the underwater terrain image of the deep water area to obtain the underwater topography of the water body to be inverted. Compared with the prior art, the method for inverting underwater topography of a large water body provided by the present invention has the following beneficial effects: in the face of a complex and changeable water environment, by partitioning the water body area to be inverted into a deep water area and a shallow water area converted from water to land, different and more targeted inversion methods can be used according to the characteristics of water bodies in different regions. For deep water areas, based on their unique water optical properties and relatively stable water depth, a special model based on band reflectivity information and underwater terrain elevation data is used for inversion; while for shallow water areas with land-water transition, spatial interpolation is used for inversion in view of their frequent changes in water level and dynamic changes in land-water boundaries. This partitioning process avoids the errors caused by using a single method to deal with complex and diverse water conditions, thereby more accurately restoring the underwater terrain of different areas and overall improving the inversion accuracy of underwater terrain of large water bodies to be inverted. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The disclosure of the present invention will become more easily understood with reference to the accompanying drawings. It is easy for those skilled in the art to understand that these drawings are only for illustrative purposes and are not intended to limit the scope of protection of the present invention. In addition, similar numbers in the figures are used to represent similar components, among which:

[0017] Figure 1 1 is a schematic flow chart of the main steps of a method for inverting underwater terrain of a large water body according to an embodiment of the present invention;

[0018] Figure 2 It is a schematic diagram of a process for partitioning the water area to be inverted to obtain deep water areas and shallow water areas for water-land conversion according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] Some embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0020] Embodiment 1

[0021] See attached Figure 1 , Figure 1 FIG. 1 is a schematic flow chart of the main steps of a method for inverting underwater terrain of a large water body according to an embodiment of the present invention. Figure 1 As shown, the large water body underwater terrain inversion method in the embodiment of the present invention mainly includes the following steps S1 to S4.

[0022] Step S1, partition the water area to be inverted to obtain a deep water area and a shallow water area converted from water to land.

[0023] In this embodiment, the deep water area is usually deep, and the optical properties of the water body, underwater terrain changes, etc. are relatively complex. The propagation of light in the water body and the law of interaction with the bottom of the water are significantly different from those in the shallow water area. The shallow water area of ​​water-land transformation is the transition zone between the water body and the land. The water level changes and terrain are significantly affected by the interaction between water and land. For example, it will be affected by factors such as tides and water flow at the river estuary. The water depth is relatively shallow and changes relatively frequently. Through such zoning, it can lay the foundation for the subsequent use of different and more targeted inversion methods to better adapt to the characteristics of different regions to accurately invert the underwater terrain.

[0024] Step S2: Invert the shallow water area converted from water to land based on a spatial interpolation method to obtain a terrain image of the shallow water area converted from water to land, for example, Kriging interpolation method, inverse distance weighted interpolation method, etc. may be used.

[0025] Step S3: inverting the deep water area using a deep water area depth inversion model constructed based on the band reflectivity information and the underwater terrain elevation data to obtain an underwater terrain image of the deep water area.

[0026] In this embodiment, under different water environments and different water depth conditions, the water body reflects light in different bands. For example, in the visible light range, the blue light band, green light band, red light band, etc., the substances in the water body (such as suspended sediment, plankton, etc.), the water depth itself and other factors will affect the reflectivity values ​​of the water body to these bands of light. Through satellite remote sensing and other technical means, the reflectivity information of each band on the surface of the water body can be obtained, and these multi-band reflectivity data are combined to form the band reflectivity information.

[0027] Underwater terrain elevation data: It reflects the ups and downs of the terrain under the water in deep water areas, that is, the vertical distance from a certain position on the bottom of the water to the water surface. There are many ways to obtain this data, for example, it can be obtained by conducting underwater terrain measurements in deep water areas on the spot through professional depth sounding equipment (such as depth sounders); it can also refer to existing hydrogeological data, which may record the underwater terrain elevation conditions of some areas measured in the past; it can also use some local underwater terrain models constructed in the early stage to extract relevant data.

[0028] Since there is an intrinsic connection between the band reflectivity information and the underwater terrain elevation data, for example, as the water depth increases, the absorption and scattering of light by the water body will gradually increase, which will lead to changes in the reflectivity of each band. This change is regular and may show a linear relationship or a nonlinear relationship (such as a quadratic function, an exponential function, etc.). Based on the analysis, statistics, and fitting of a large amount of sample data (both the acquired band reflectivity data and the corresponding underwater terrain elevation data), a mathematical model that can reflect the relationship between the two is constructed using mathematical methods. This model is the deep water depth inversion model.

[0029] Step S4, image splicing and fusion of shallow water area topography image and deep water area underwater topography image of land-water conversion to obtain underwater topography of water body to be inverted. After the previous steps, shallow water area topography image and deep water area underwater topography image of land-water conversion are obtained respectively, but they are generated separately for different regions. To fully display the underwater topography of the entire water body to be inverted, image splicing and fusion are required. This step can eliminate the boundary difference between the two regional images, so that they are naturally and coherently connected together to form a unified image that can reflect the whole picture of underwater topography of the entire water body. In the splicing and fusion process, it is necessary to consider the consistency of the two regional images in terms of coordinate system, resolution, data format, etc., and realize seamless splicing through technical means such as image registration, resampling, and data fusion. For example, by finding the same-name objects or pixels with the same characteristics at the boundary of the two regional images for accurate coordinate matching, and then according to a certain algorithm, the data of the overlapping part is fused to ensure natural transition, and finally obtain a complete underwater topography image of the water body to be inverted, providing a comprehensive and accurate data basis for subsequent analysis and application.

[0030] In one embodiment, Figure 2 As shown, step S1, partitioning the water area to be inverted to obtain the deep water area and the shallow water area converted from water to land, including:

[0031] Step S11, obtaining the water level variation characteristics of the water body area to be inverted; Step S12, determining the deep water area and the shallow water area converted from water to land based on the water level variation characteristics.

[0032] In this embodiment, the water level change characteristics may include water level stability and change amplitude, wherein the water level change amplitude in the deep water area is relatively small, and the water level in the shallow water area changes frequently, and it is easy to have the phenomenon of alternation between water and land. Due to its shallow water depth, a small amount of water volume change will cause a large fluctuation in the water level, thereby causing the land to be submerged or re-exposed. And because its water body capacity is small, it is sensitive to water volume changes.

[0033] Or, in an alternative method, the water depth information can be directly obtained using detection equipment, and the water area to be inverted can be divided into zones according to the water depth to obtain deep water areas and shallow water areas converted from water to land;

[0034] Alternatively, in an alternative approach, the spectral reflectance characteristics are used to partition the water area to be inverted. Areas of different water depths and water compositions differ in spectral reflectance characteristics. For example, in the visible light band, the reflectance of the green light band in shallow water areas may be relatively high due to the influence of bottom reflection; while in deep water areas, the reflectance of the blue light band may be relatively high due to the absorption and scattering of light. The reflectance of water bodies in different bands is measured by satellite remote sensing or on-site spectrometers, and these spectral features are used to distinguish between deep water areas and shallow water areas. In either case, the water area to be inverted can be partitioned to obtain deep water areas and shallow water areas converted from water to land.

[0035] In one embodiment, step S2, the process of inverting the shallow water area converted from water to land based on the spatial interpolation method to obtain the terrain image of the shallow water area converted from water to land includes:

[0036] The improved normalized difference water index is used to extract the water body boundary of the shallow water area of ​​water-land transformation, distinguish the shallow water area and land; obtain the water level line of the shallow water area; based on the water level line of the shallow water area, the Kriging interpolation method is used to invert the underwater topography map of the shallow water area.

[0037] In this embodiment, the formula of the improved normalized difference water index is:

[0038]

[0039] Where: MNDWI is the improved normalized difference water index; R Green is the green band reflectivity; R MIR is the reflectivity in the mid-infrared band.

[0040] There are many ways to obtain the water level in shallow water areas. One common way is to use water level monitoring equipment installed on site, such as water gauges set up around shallow water areas or at appropriate locations, to directly read water level values ​​at different times and then determine the water level. You can also use remote sensing image analysis to observe the traces of the boundary between the water body and the land on the image, and calculate the location of the water level based on the known elevation benchmark and other information in the area. You can also refer to historical water level data and the hydrological model simulation results of the area to comprehensively determine the water level in shallow water areas.

[0041] In one embodiment, the method includes at least the following steps to obtain a deep water area water depth inversion model: obtaining the relationship between band reflectivity information and underwater terrain elevation data; and constructing a deep water area water depth inversion model based on the relationship between band reflectivity information and underwater terrain elevation data.

[0042] In this embodiment, a large number of sample points containing both band reflectivity information and underwater terrain elevation data are collected. These sample points can be obtained by measuring at different locations and at different times. The relationship between the two is found through statistical analysis and mathematical modeling methods. After the relationship between the band reflectivity information and the underwater terrain elevation data is determined, an inversion model can be constructed.

[0043] In one embodiment, the deep water depth inversion model includes at least one or more of a single band model, a band ratio model, a neural network model, and a random forest model. The process of constructing the deep water depth inversion model based on the relationship between the band reflectivity information and the underwater terrain elevation data includes:

[0044] The optimal water depth factor is determined based on the correlation between the band reflectivity and the underwater terrain elevation, and a regression analysis is performed on the optimal water depth factor. The single-band and band ratio water depth factors are fitted to obtain a single-band model and a band ratio model.

[0045] In this embodiment, the reflectivity of different bands reflects the water depth to different degrees. By collecting a large amount of measured data (including reflectivity data of each band and corresponding underwater terrain elevation data), statistical analysis methods (such as Pearson correlation coefficient, Kendall correlation coefficient and Spearman correlation coefficient) are used to calculate correlation coefficients, etc., to determine which band or bands have the strongest correlation between reflectivity and elevation. The reflectivity of the band with the strongest correlation or the index derived from it is determined as the optimal water depth factor. For example, after analyzing the data of a deep water area of ​​a lake, it is found that the correlation coefficient between the reflectivity of the green light band and the underwater terrain elevation reaches 0.641, which is stronger than the correlation of other bands. Then the reflectivity of the green light band may be determined as the optimal water depth factor. Using different ratio combinations of 6 bands and the underwater terrain elevation of the training sample points for correlation analysis, it is found that the band ratio combination with the highest correlation with the underwater terrain elevation is the green band / blue band, and the correlation coefficient is 0.653. Regression analysis was performed on the optimal water depth factor of the single-band model and the optimal water depth factor of the band ratio. The optimal water depth factor was fitted using linear function, exponential function, logarithmic function, power function and polynomial function respectively. It was found that the linear function had the best fitting effect in the single-band water depth inversion model, with a determination coefficient of 0.740; the exponential function had the best fitting effect in the band ratio water depth inversion model, with a determination coefficient of 0.721.

[0046] The band reflectivity information is used as the input feature of the neural network model and the random forest model to learn and predict the underwater terrain elevation, and the neural network model and the random forest model are obtained.

[0047] In this embodiment, the neural network is a nonlinear multi-layer forward feedback network, which is divided into three layers: input layer, hidden layer and output layer. The input layer is used to receive input data, and the hidden layer performs complex nonlinear transformation on the input data through multiple neurons. The output layer is used to generate prediction results. The input layer of the constructed neural network model includes 6 nodes, corresponding to the 6 bands of remote sensing images. The hidden layer is 4 layers, and the number of nodes in each layer is 128, 64, 32, and 16 respectively. The radial basis function (RBF) is selected as the activation function to enhance the nonlinear mapping ability of the model. The output layer includes 1 node for outputting the predicted underwater terrain elevation value. 200 iterations were performed during the model training process, and the learning efficiency was 0.1. Through repeated experiments and parameter adjustment, the optimal model was finally determined. The mean square error (MSE) was selected as the loss function, and the Adam optimizer was used for optimization. The validation set was used to monitor the model performance during the training process. After the training is completed, the model is applied to the remote sensing images of the entire deep water area to predict the underwater terrain elevation of the deep water area and generate an underwater terrain map.

[0048] Random forest model: The optimal parameter settings were determined through multiple experiments: the number of classifiers is 160, the minimum number of samples for leaf nodes is 1, the maximum number of features is 6, and the maximum depth of the decision tree is 20. The optimization of these parameters can effectively improve the stability and prediction accuracy of the model. During the model training process, the random forest algorithm randomly extracts multiple subsets from the training set through the Bootstrap sampling method, and trains a decision tree on each subset. Each tree independently predicts the underwater terrain elevation, and finally generates a deep-water underwater terrain map after obtaining a comprehensive prediction result through majority voting or averaging.

[0049] Furthermore, the water depth range interval can be pre-set, for example, the deep water area is divided into shallow water area (such as 10-50 meters deep), medium deep water area (50-200 meters) and deep water area (greater than 200 meters). During the model training process, different models (single-band model, band ratio model, neural network model or random forest model) are trained for each water depth range, or an integrated model is trained in which each sub-model focuses on a specific water depth range. When prediction is required, the model with the best training effect within the range is automatically selected for prediction based on the water depth range corresponding to the input reflectivity data. Specifically, a meta-model can be constructed to automatically select a suitable model. The meta-model takes the input band reflectivity information and water depth range as input, and decides which model should be selected in the current situation by learning the historical model selection records and model performance data. For example, the meta-model can be a simple decision tree or a small neural network that determines which model (single-band model, band ratio model, neural network model, or random forest model) is most likely to provide the most accurate prediction under the current water depth and reflectivity conditions based on the input features.

[0050] Embodiment 2

[0051] The present invention also provides an electronic device. In an embodiment of a device according to the present invention, the device includes a processor and a storage device, the storage device can be configured to store a program for executing the large water body underwater topography inversion method of the above method embodiment, and the processor can be configured to execute the program in the storage device, which includes but is not limited to the program for executing the large water body underwater topography inversion method of the above method embodiment. For ease of explanation, only the parts related to the embodiment of the present invention are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present invention. The control device can be a control device device formed by various electronic devices.

[0052] Embodiment 3

[0053] The present invention also provides a computer-readable storage medium. In a computer-readable storage medium embodiment according to the present invention, the computer-readable storage medium can be configured to store a program for executing the large water body underwater topography inversion method of the above method embodiment, and the program can be loaded and run by a processor to implement the above large water body underwater topography inversion method. For ease of explanation, only the parts related to the embodiment of the present invention are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present invention. The computer-readable storage medium can be a storage device formed by various electronic devices. Optionally, the computer-readable storage medium in the embodiment of the present invention is a non-temporary computer-readable storage medium.

[0054] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the original technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for inverting underwater terrain of a large water body, characterized in that: include: The water area to be inverted is divided into zones to obtain deep water areas and shallow water areas transformed from water to land; Inverting the land-water transformation shallow water area based on the spatial interpolation method to obtain a land-water transformation shallow water area terrain image; Inverting the deep water area using a deep water area depth inversion model constructed based on band reflectivity information and underwater terrain elevation data to obtain an underwater terrain image of the deep water area; The shallow water area topographic image and the deep water area underwater topographic image are stitched and fused to obtain the underwater topography of the water body to be inverted.

2. The method according to claim 1, characterized in that The water area to be inverted is divided into two zones, and the deep water area and the shallow water area converted from water to land are obtained, including: Obtaining water level variation characteristics of the water body area to be inverted; The deep water area and the shallow water area transformed from land to water are determined based on the water level change characteristics.

3. The method according to claim 2, characterized in that The process of inverting the shallow water area transformed from water to land based on the spatial interpolation method to obtain the terrain image of the shallow water area transformed from water to land includes: The improved normalized difference water index is used to extract the water boundary of the shallow water area of ​​the water-land transformation, and distinguish the shallow water area and the land; Obtaining a water level line of the shallow water area; Based on the water level line of the shallow water area, the Kriging interpolation method is used to invert the underwater topographic map of the shallow water area.

4. The method according to claim 2, wherein the method comprises at least the following steps of obtaining a water depth inversion model of a deep water area: Obtain the relationship between band reflectivity information and underwater terrain elevation data; A deep water depth inversion model is constructed based on the relationship between the band reflectivity information and the underwater terrain elevation data.

5. The method according to claim 4, characterized in that The deep water area water depth inversion model at least includes one or more of a single band model, a band ratio model, a neural network model and a random forest model. The process of constructing the deep water area water depth inversion model based on the relationship between the band reflectivity information and the underwater terrain elevation data includes: Determine the optimal water depth factor based on the correlation between the band reflectivity and the underwater terrain elevation, perform regression analysis on the optimal water depth factor, and fit the single-band and band ratio water depth factors to obtain a single-band model and a band ratio model; The band reflectivity information is used as the input feature of the neural network model and the random forest model to learn and predict the underwater terrain elevation, and the neural network model and the random forest model are obtained.

6. The method according to claim 5, characterized in that The process of determining the optimal water depth factor based on the correlation between the band reflectivity and the underwater terrain elevation, performing regression analysis on the optimal water depth factor, fitting the single-band and band ratio water depth factors, and obtaining the single-band model and the band ratio model at least includes: Calculate the correlation between single-band reflectivity and underwater terrain elevation, and take the single-band with the highest correlation with underwater terrain elevation as the single-band optimal water depth factor; A regression analysis was performed on the single-band optimal water depth factor, and the single-band optimal water depth factor was fitted using linear function, exponential function, logarithmic function, power function, and polynomial function respectively. The single-band model was obtained by selecting the one with the best fitting effect.

7. The method according to claim 5, characterized in that The process of determining the optimal water depth factor based on the correlation between the band reflectivity and the underwater terrain elevation, performing regression analysis on the optimal water depth factor, fitting the single band and band ratio water depth factors, and obtaining the single band model and the band ratio model also includes: Obtain different band ratio combinations, calculate the correlation between all band ratios and underwater terrain elevation, and find the band ratio combination with the highest correlation as the optimal water depth factor for the band ratio; The optimal water depth factor of the band ratio is subjected to regression analysis, and the optimal water depth factor of the band ratio is fitted using linear function, exponential function, logarithmic function, power function and polynomial function respectively. The band ratio model can be obtained by selecting the one with the best fitting effect.

8. An electronic device comprising a processor and a storage device, the storage device being suitable for storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by a processor to execute the large water body underwater terrain inversion method according to any one of claims 1 to 7.

9. A computer-readable storage medium storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by a processor to execute the large water body underwater terrain inversion method according to any one of claims 1 to 7.

Images