A Lake Topography Prediction Method Based on Machine Learning and Multi-Source Satellite Data

Abstract

This application discloses a lake topography prediction method based on machine learning and multi-source satellite data fusion, belonging to the field of machine learning technology. The method includes: extracting optical features of the target lake from optical satellite imagery; extracting water features of each water cell in the target lake from SWOT satellite data; spatiotemporally aligning and fusing the optical and water features to obtain a fused feature vector; matching the measured water depth points of the actual water depth to the geographical locations of the corresponding cells, and then using the fused feature vectors of the actual water depth and the corresponding cells to form training samples; selecting a machine learning model and initializing its structure or parameters; training the initialized machine learning model using the training samples; predicting the water depth value of the corresponding cell using the trained machine learning model based on the fused feature vectors of each water cell; and determining the topography of the target lake based on the water depth values. This application can efficiently and cost-effectively predict the topography of large-scale lakes.

Description

Technical Field

[0001] This application relates to the field of machine learning technology, and in particular to a method for predicting lake topography based on machine learning and fusing multi-source satellite data. Background Technology

[0002] Lakes are an important component of the Earth's surface system, and their underwater topography (or bathymetry) data is of paramount importance for water resource assessment and management, water storage calculation, aquatic environment simulation, ecosystem research, flood control and disaster reduction, shipping planning, aquaculture, and lake evolution research. While traditional lake topography measurement methods offer high accuracy, they still suffer from high costs and low efficiency. Summary of the Invention

[0003] The main objective of this application is to propose a lake topography prediction method based on machine learning and multi-source satellite data fusion, so as to improve the prediction efficiency of lake topography and reduce costs.

[0004] To achieve the above objectives, one aspect of this application proposes a lake topography prediction method based on machine learning and fusion of multi-source satellite data. The method includes the following steps:

[0005] Optical features of the target lake were extracted from optical satellite imagery;

[0006] Water features of each water cell in the target lake were extracted from SWOT satellite data.

[0007] The optical features and the water features are spatiotemporally aligned and fused to obtain a fused characteristic vector.

[0008] The measured water depth points of the actual water depth are matched to the geographical locations of the corresponding pixels, and then the actual water depth and the fused feature vector of the corresponding pixel are used to form training samples.

[0009] Select a machine learning model and initialize the structure or parameters of the machine learning model;

[0010] The initialized machine learning model is trained using the training samples.

[0011] The trained machine learning model is used to predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel;

[0012] The topography of the target lake is determined based on each of the stated water depth values.

[0013] In some embodiments, extracting the optical features of the target lake from optical satellite imagery includes the following steps:

[0014] The optical features, including band reflectance, band ratio, logarithmic transformation features, and water index of the target lake, are extracted from optical satellite imagery.

[0015] In some embodiments, extracting water features of each water cell in the target lake from SWOT satellite data includes the following steps:

[0016] The water features, including water surface elevation, water surface slope, shoreline distance, and SWOT water body information, are extracted from SWOT satellite data for each water body pixel in the target lake.

[0017] In some embodiments, selecting a machine learning model and initializing the structure or parameters of the machine learning model includes the following steps:

[0018] Based on the characteristics of the training samples and computational resources, the machine learning model is selected from random forest regression model, gradient boosting regression model, artificial neural network model, or convolutional neural network variant model that incorporates neighborhood information, and the structure or parameters of the machine learning model are initialized.

[0019] In some embodiments, training the initialized machine learning model using the training samples includes the following steps:

[0020] The training set is obtained by dividing the training samples;

[0021] The training set is used to learn the parameters of the machine learning model;

[0022] The target hyperparameters of the machine learning model are optimized using cross-validation combined with grid search, random search, or Bayesian optimization strategies.

[0023] In some embodiments, determining the topography of the target lake based on each of the water depth values ​​includes the following steps:

[0024] The water depth values ​​are converted into lake bottom elevations using the water surface elevations of the water body pixels in the SWOT satellite data.

[0025] Generate a raster map of the lakebed elevation;

[0026] Spatial smoothing filtering, outlier removal and repair, and contour line generation and mapping are performed on the raster image to obtain the topographic map of the target lake.

[0027] In some embodiments, prior to extracting the optical features of the target lake from the optical satellite imagery, the method further includes the following steps:

[0028] The optical satellite imagery is subjected to radiometric correction, geometric correction and registration, water body extraction, and removal of clouds, cloud shadows and other interference.

[0029] Before extracting the water features of each water cell in the target lake from the SWOT satellite data, the method further includes the following steps:

[0030] Outlier removal, spatiotemporal matching and grid unification, and water body boundary optimization were performed on the SWOT satellite data.

[0031] To achieve the above objectives, another aspect of this application proposes a lake topography prediction device based on machine learning and multi-source satellite data fusion, the device comprising:

[0032] An optical feature extraction unit is used to extract the optical features of the target lake from optical satellite imagery;

[0033] The water feature extraction unit is used to extract the water features of each water cell in the target lake from SWOT satellite data.

[0034] The feature fusion unit is used to perform spatiotemporal alignment and fusion of the optical features and the water features to obtain a fused feature vector.

[0035] The sample construction unit is used to match the measured water depth points of the real water depth to the geographical location of the corresponding pixel, and then to form training samples by combining the real water depth and the fused feature vector of the corresponding pixel.

[0036] A model initialization unit is used to select a machine learning model and initialize the structure or parameters of the machine learning model.

[0037] A model training unit is used to train the initialized machine learning model using the training samples.

[0038] A water depth prediction unit is used to predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel using the trained machine learning model.

[0039] A terrain determination unit is used to determine the terrain of the target lake based on each of the water depth values.

[0040] To achieve the above objectives, another aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0041] To achieve the above objectives, another aspect of the embodiments of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0042] The embodiments of this application include at least the following beneficial effects:

[0043] This application extracts optical features of a target lake from optical satellite imagery; extracts water features of each water cell in the target lake from SWOT satellite data; spatiotemporally aligns and fuses the optical and water features to obtain a fused feature vector; matches measured water depth points to the geographical locations of corresponding cells, and then uses the fused feature vectors of the measured water depths and corresponding cells to form training samples; selects a machine learning model and initializes its structure or parameters; trains the initialized machine learning model using the training samples; uses the trained machine learning model to predict the water depth value of each water cell based on the fused feature vectors of each water cell; and determines the topography of the target lake based on the water depth values. This application, by using a trained machine learning model to predict the topography of the entire lake based on extracted feature values, can efficiently and cost-effectively predict the underwater topography of large-scale lakes, providing important data support for water resource management and environmental protection. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A flowchart illustrating a lake topography prediction method based on machine learning and multi-source satellite data, provided in an embodiment of this application;

[0046] Figure 2 An example flowchart of a lake topography prediction method based on machine learning and multi-source satellite data is provided for embodiments of this application.

[0047] Figure 3 A schematic diagram of a lake topography prediction device based on machine learning and multi-source satellite data fusion provided in this application embodiment;

[0048] Figure 4 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0050] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0051] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0053] Before providing a detailed description of the embodiments of this application, some related technologies involved in the embodiments of this application will be described first, as follows:

[0054] With the development of remote sensing technology, large-scale, periodic lake monitoring using satellite data has become possible. In recent years, in particular, the emergence of high-resolution, multi-source satellites (such as the Sentinel and Landsat series) and satellite missions specifically designed for surface hydrological observation (such as the SWOT satellite) has provided an unprecedented data foundation for lake topography inversion. Meanwhile, machine learning technology, with its powerful nonlinear fitting and complex pattern recognition capabilities, has demonstrated enormous potential in processing remote sensing big data and addressing Earth science problems. Therefore, combining multi-source satellite data with advanced machine learning algorithms to efficiently and accurately predict lake topography has become a research hotspot in the fields of hydrology and remote sensing.

[0055] Currently, there are many methods for measuring lake topography:

[0056] (1) Traditional underwater topographic surveying:

[0057] Traditional underwater topographic surveying is mainly divided into two categories: direct measurement and acoustic sounding. Direct measurement involves using a graduated sounding rod to measure depth directly in shallow water; another method is the plumb bob sounding, where a weighted rope is lowered to the bottom and the depth is determined by reading the rope length. Acoustic sounding methods typically rely on shipboard platforms (including manually operated vessels or unmanned vessels). For example, a single-beam echo sounder emits sound waves vertically from the ship's hull to the seabed, calculating the depth at a single point based on the round-trip time of the sound waves, thus obtaining the topographic profile directly below the ship's path. A multi-beam echo sounder, for instance, simultaneously emits multiple sound beams arranged in a fan shape, measuring the depth of a wide strip along the ship's path in a single measurement, enabling the acquisition of high-precision, full-coverage underwater topographic data. Side-scan sonar is also used to generate acoustic images of the underwater topography, aiding in the identification of underwater objects, seabed types, and the understanding of overall topographic features.

[0058] (2) Satellite optical remote sensing water depth inversion based on empirical models:

[0059] This type of method utilizes images acquired by satellite optical sensors (such as the multispectral imager carried by Landsat and Sentinel-2) to estimate water depth by analyzing the selective absorption and scattering characteristics of water bodies on different spectral bands. The core principle is that light attenuates with increasing depth in water, especially in short-wavelength bands like blue and green, where penetration is relatively strong, while red and near-infrared light are rapidly absorbed. Empirical models aim to establish a mathematical relationship between the apparent reflectance of satellite imagery (or atmospherically corrected surface reflectance) and actual water depth. This typically requires a certain number of measured water depth points as calibration data for regression analysis. Common models include single-band log-linear models (assuming a linear relationship between water depth and the logarithm of reflectance in a specific band) and multi-band log-band ratio models (such as the model proposed by Stumpf et al., which uses the logarithmic ratio of blue-green band reflectance to mitigate the influence of water optical parameters and some bottom sediment variations).

[0060] (3) Remote sensing water depth inversion based on physical model (semi-analytical model):

[0061] Compared to empirical models, remote sensing depth retrieval methods based on physical or semi-analytical models attempt to describe the radiative transfer process of light in water bodies more deeply. These models not only consider water depth but also explicitly incorporate inherent optical parameters (IOPs) of the water body (such as absorption and backscattering coefficients) and seabed reflectivity (substrate type and reflectivity) into their framework. The basic idea is to separate the contributions of water depth, water components (such as chlorophyll, suspended sediment, and colored soluble organic matter (CDOM)), and seabed reflectivity from the total reflectance signal above the water surface observed by satellite by solving the inverse problem of the complex underwater radiative transfer equation (or its simplified form). For example, water depth can be inverted under given conditions of water body IOPs and seabed spectral libraries through iterative optimization, lookup table (LUT) matching, or spectral decomposition. These methods are theoretically more robust, more adaptable to environmental changes, and can potentially retrieve water body optical parameters simultaneously.

[0062] (4) Water depth estimation based on a small amount of satellite altimetry data combined with imagery:

[0063] This method utilizes discrete but precise water surface elevation (WSE) anchor points provided by high-precision satellite altimetry technology (such as the ICESat-2 laser altimeter satellite), and combines them with water body boundaries (shorelines) delineated by optical or radar satellite imagery acquired quasi-synchronously to estimate water storage changes and local water depth characteristics of lakes and other water bodies. The ICESat-2 photon-counting lidar can penetrate parts of the water body, and in shallow water areas, it can even directly detect the bottom, thus providing the true water depth for some areas; even if it cannot penetrate to the bottom, the precise WSE it provides is a valuable elevation benchmark. When multiple periods of water level elevation and corresponding water body extents are obtained, an area-elevation curve (i.e., a lake-marsh curve or reservoir capacity curve) can be constructed for the lake, thereby estimating the total water storage changes at different water levels. For a specific shoreline location, its elevation is the water level elevation at that time, which can be considered as a "zero-depth" reference for that point. By combining this elevation benchmark, the relative water depth estimated by other remote sensing water depth inversion methods (such as optical empirical models) can be converted into absolute water depth with actual elevation significance.

[0064] (5) Preliminary exploration of water body characteristic research using SWOT satellite data alone:

[0065] The SWOT satellite is designed to provide high-precision elevation, extent, and slope of global surface water bodies. Current research mainly focuses on using SWOT data for water level monitoring, dynamic analysis of water body extent changes, and river flow estimation. Its direct application to the accurate inversion of detailed underwater topography of lakes (rather than just water surface elevation), especially in its fusion with other types of satellite data (such as multispectral imagery), is still in its early stages of exploration.

[0066] (6) Applying machine learning to water depth inversion from a single remote sensing data source:

[0067] In recent years, machine learning algorithms have received increasing attention and application in the field of remote sensing water depth inversion due to their powerful nonlinear modeling capabilities and the advantage of automatically learning features from complex data. Numerous studies have attempted to utilize various machine learning models (such as random forests, support vector machines / regression, gradient boosting trees, artificial neural networks, and even deep learning models like convolutional neural networks) to combine single-source optical satellite imagery (e.g., Landsat series or Sentinel-2 imagery only) for water depth prediction. These methods typically use the raw band reflectance of the imagery, calculated spectral indices (such as various water body indices and color indices), and texture features as input features, and measured water depth data as training labels. The model is trained through supervised learning to establish a highly nonlinear mapping relationship between image features and water depth. Compared to traditional empirical models, machine learning methods often better handle the impact of complex water conditions and changes in seabed sediment, sometimes achieving higher prediction accuracy without requiring pre-defined fixed mathematical formulas.

[0068] The shortcomings of existing technologies:

[0069] (1) Traditional underwater topographic surveying:

[0070] Although the accuracy is high, the operation is costly in terms of materials and manpower, has limited coverage, low efficiency, and poor timeliness, making it difficult to quickly and repeatedly conduct dynamic monitoring. Especially in shallow water areas, and in some urban landscape lakes, measurements are affected by obstacles such as passing tourist boats and aquatic plants.

[0071] (2) Empirical / semi-analytical optical remote sensing water depth inversion method:

[0072] ① High dependence on water quality: The accuracy is severely affected by the optical properties of the water body (such as turbidity, chlorophyll concentration, and substrate type), and the model has poor universality. It usually requires localized calibration for different lakes or different periods.

[0073] ② High requirements for atmospheric correction: Atmospheric scattering and absorption have a significant impact on water body signals, and accurate atmospheric correction is crucial, but it is often difficult to achieve perfectly.

[0074] ③Limited depth detection range: In deep or turbid waters, the light signal attenuates rapidly, making detection difficult. In extremely shallow waters, sediment reflection has a significant impact.

[0075] ④ Reliance on measured data: Most empirical models and some semi-analytical models still require a certain amount of measured water depth data for model parameter calibration and verification, which limits their effective application in areas with no or little data.

[0076] ⑤ Limitations of using SWOT data alone for topographic inversion: SWOT directly measures water surface elevation and extent, not underwater topography. Although its data can be used to infer changes in water volume or constrain water boundaries, it is quite difficult to directly and accurately invert detailed lakebed topography, and the spatial resolution of SWOT (e.g., its rasterized water surface elevation product) is still somewhat limited for fine topographic characterization.

[0077] ⑥ Machine learning methods based on a single remote sensing data source: Although machine learning improves modeling capabilities, the information provided by a single data source is limited in dimensionality. For example, it is difficult to accurately distinguish the differences in reflected signals caused by changes in water depth and changes in water composition using only optical imagery, and it is also impossible to directly obtain accurate absolute water surface elevation as a benchmark. This limits further improvement in model accuracy and robustness under complex hydrological conditions.

[0078] ⑦ Insufficient data fusion: Existing technologies have limited ability to effectively integrate the unique advantages of structured hydrological information such as spectral information from optical images and high-precision water surface elevation and water body extent provided by SWOT satellites, failing to fully leverage the collaborative potential of multi-source data to overcome the limitations of a single data source.

[0079] To address the problems of high cost, low efficiency, significant susceptibility to water quality, poor universality, limited deep-water exploration capabilities, and failure to fully integrate the advantages of multi-source satellite data in the aforementioned existing technologies, this application aims to solve the following technical problems:

[0080] (1) How to effectively integrate the rich spectral information provided by satellite multispectral / hyperspectral band images with the high-precision water surface elevation, water body range and water surface slope data provided by SWOT satellite, so as to more comprehensively characterize the water-light-topography features of lakes.

[0081] (2) How to utilize the powerful nonlinear modeling capabilities of machine learning to learn from the features of fused multi-source data and predict the underwater topography of lakes with high accuracy, thereby reducing the dependence on the assumptions of traditional empirical models and improving the universality and prediction accuracy of the model.

[0082] (3) How to use the precise water surface elevation provided by SWOT data as a benchmark to help distinguish the signal variations in optical images caused by changes in water depth and water composition, and to help constrain the boundary conditions and elevation references for terrain prediction.

[0083] (4) A lake topography prediction technology is provided that is more efficient and less expensive than traditional measurement methods, and more accurate and more applicable than single remote sensing data source methods. It is particularly suitable for rapid topographic mapping of lakes where there is a lack of measured data or large-scale lakes.

[0084] Based on this, this application proposes a lake topography prediction method based on multi-source satellite data and machine learning. First, satellite band images and SWOT satellite data are acquired and preprocessed to extract multi-dimensional features related to lake topography. Next, the model is trained and its parameters are adjusted using measured water depth data from unmanned surface vessels. Finally, the trained machine learning model is used to predict the topography of the entire lake based on the extracted feature values. This application can efficiently and cost-effectively predict the underwater topography of large-scale lakes, providing important data support for water resource management and environmental protection.

[0085] This application provides a lake topography prediction method based on machine learning and multi-source satellite data fusion, relating to the field of machine learning technology. The lake topography prediction method based on machine learning and multi-source satellite data fusion provided in this application can be applied to a terminal, a server, or software running on a terminal or server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application implementing a lake topography prediction method based on machine learning and multi-source satellite data fusion, but is not limited to the above forms.

[0086] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0087] Reference Figure 1This application provides a lake topography prediction method based on machine learning and multi-source satellite data fusion. This method may include, but is not limited to, S1 to S8, as follows:

[0088] S1: Extract the optical features of the target lake from optical satellite imagery;

[0089] S2: Extract water features of each water cell in the target lake from SWOT satellite data;

[0090] S3: Spatiotemporally align and fuse the optical features and the water features to obtain a fused characteristic vector;

[0091] S4: Match the measured water depth points of the real water depth to the geographical location of the corresponding pixels, and then use the real water depth and the fused feature vector of the corresponding pixel to form training samples.

[0092] S5: Select a machine learning model and initialize the structure or parameters of the machine learning model;

[0093] S6: Train the initialized machine learning model using the training samples;

[0094] S7: Using the trained machine learning model, predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel;

[0095] S8: Determine the topography of the target lake based on each of the stated water depth values.

[0096] Optionally, extracting the optical features of the target lake from optical satellite imagery includes the following steps:

[0097] The optical features, including band reflectance, band ratio, logarithmic transformation features, and water index of the target lake, are extracted from optical satellite imagery.

[0098] Optionally, the step of extracting water features of each water cell in the target lake from SWOT satellite data includes the following steps:

[0099] The water features, including water surface elevation, water surface slope, shoreline distance, and SWOT water body information, are extracted from SWOT satellite data for each water body pixel in the target lake.

[0100] Optionally, selecting a machine learning model and initializing the structure or parameters of the machine learning model includes the following steps:

[0101] Based on the characteristics of the training samples and computational resources, the machine learning model is selected from random forest regression model, gradient boosting regression model, artificial neural network model, or convolutional neural network variant model that incorporates neighborhood information, and the structure or parameters of the machine learning model are initialized.

[0102] Optionally, training the initialized machine learning model using the training samples includes the following steps:

[0103] The training set is obtained by dividing the training samples;

[0104] The training set is used to learn the parameters of the machine learning model;

[0105] The target hyperparameters of the machine learning model are optimized using cross-validation combined with grid search, random search, or Bayesian optimization strategies.

[0106] Optionally, determining the topography of the target lake based on each of the water depth values ​​includes the following steps:

[0107] The water depth values ​​are converted into lake bottom elevations using the water surface elevations of the water body pixels in the SWOT satellite data.

[0108] Generate a raster map of the lakebed elevation;

[0109] Spatial smoothing filtering, outlier removal and repair, and contour line generation and mapping are performed on the raster image to obtain the topographic map of the target lake.

[0110] Optionally, before extracting the optical features of the target lake from the optical satellite imagery, the method further includes the following steps:

[0111] The optical satellite imagery is subjected to radiometric correction, geometric correction and registration, water body extraction, and removal of clouds, cloud shadows and other interference.

[0112] Before extracting the water features of each water cell in the target lake from the SWOT satellite data, the method further includes the following steps:

[0113] Outlier removal, spatiotemporal matching and grid unification, and water body boundary optimization were performed on the SWOT satellite data.

[0114] The following section will provide a detailed introduction and explanation of the solutions in the embodiments of this application, using specific application examples.

[0115] Reference Figure 2This application aims to propose a method for predicting underwater topography in lakes. The key technical solution includes deep fusion of multi-source satellite remote sensing data (optical imagery and SWOT satellite observations) combined with advanced machine learning algorithms. The basic idea is to fully utilize the optical response characteristics of satellite imagery to water composition and water column attenuation effects, and combine this with multi-dimensional information provided by the SWOT satellite, such as centimeter-level water surface elevation (WSE), high-precision water body extent, and water surface slope, to construct a more comprehensive feature set with richer information and more significant physical constraints. Subsequently, a machine learning model is used to mine and learn the complex nonlinear mapping relationship between these geophysical features and actual water depth, ultimately achieving high-precision, wide-coverage inversion of underwater topography in lakes.

[0116] The detailed steps of this embodiment are as follows:

[0117] ① Data acquisition.

[0118] Optical satellite imagery acquisition:

[0119] Collect archived or newly acquired multispectral / hyperspectral satellite imagery covering the target lake area. Data sources may include, but are not limited to, publicly available data such as Landsat series (e.g., OLI / TIRS), Sentinel-2 (MSI), Sentinel-3 (OLCI), or high-resolution satellite imagery such as Planet, WorldView, and GeoEye.

[0120] SWOT satellite data acquisition:

[0121] Acquire SWOT (Surface Water and Ocean Topography) satellite data products that are as consistent as possible with the timing of optical imagery or fall within the same hydrological period. Key data products include: high-precision water surface elevation (WSE) raster data, water body classification data, and potential water surface slope (WSS) products provided by L2_HR_KaRIn (Ka-band Radar Interferometer High Rate).

[0122] Auxiliary data acquisition (optional):

[0123] Digital Elevation Models (DEMs): Such as ASTER GDEM and SRTM DEM, used for shoreline delineation, surrounding terrain analysis, and result correction. Historical Measured Water Depth Data: A small amount of high-precision historical water depth data obtained through sonar sounding, LiDAR, and other methods, primarily used as training labels for machine learning models and / or for independent accuracy verification. Hydrological Station Observation Data: Synchronous or quasi-synchronous water level observation records from hydrological stations in the target lake or adjacent areas, used for cross-validation or bias correction of the WSE data provided by SWOT analysis.

[0124] ② Data preprocessing.

[0125] Optical image preprocessing:

[0126] Radiometric correction converts raw DN (Digital Number) values ​​to TOA (Top Atmosphere) reflectance or radiance values. Atmospheric correction eliminates the influence of atmospheric scattering and absorption on ground reflectance, obtaining accurate surface reflectance. Mature algorithms or tools such as FLAASH, QUAC, Sen2Cor, and LaSRC can be used. Geometric correction and registration perform high-precision geometric correction on the image to ensure accurate spatial registration with SWOT data, DEM, and other auxiliary data, with errors controlled within one pixel. Cloud, cloud shadow, and other interference removal uses quality assessment bands or specific algorithms (such as Fmask, CFmask) to identify and mask invalid pixels such as clouds, cloud shadows, and dense fog. Accurate water body extraction uses a combination of improved Normalized Difference Water Index (MNDWI), Automated Water Extraction Index (AWEI), object-oriented classification methods, or deep learning segmentation models to extract high-precision lake water body extents.

[0127] SWOT data preprocessing:

[0128] Outlier handling: Remove obvious outliers or isolated points from WSE, WSS, and other data. Spatiotemporal matching and grid unification: Reproject SWOT rasterized products (such as WSE) to a geographic coordinate system consistent with optical imagery and resample to a uniform analytical grid resolution (e.g., consistent with optical imagery resolution or setting an optimal analytical scale). Water body boundary optimization: Utilize the high-precision water body extent classification data provided by SWOT to further optimize or constrain water body boundaries extracted from optical imagery.

[0129] ③ Feature extraction and multi-source data fusion.

[0130] Feature extraction based on optical images:

[0131] Band reflectance is directly calculated using the preprocessed surface reflectance values ​​of each band (blue, green, red, near-infrared, shortwave infrared, etc.). Band ratios are calculated to address water depth sensitivity, such as the green / blue band ratio, red / green band ratio, and near-infrared / green band ratio, to mitigate the influence of the underwater matrix and some water components. Logarithmic transformation features are used, employing logarithmically transformed single bands or band ratios (e.g., ln(blue), ln(green), ln(red), ln(NIR / Green), ln(Blue / Green)). These features are widely used in classic semi-theoretical / empirical water depth inversion models and can linearize the relationship between water depth and radiation signals. Water body indices, such as MNDWI and NDWI, may exhibit a certain correlation with water depth under specific conditions (e.g., extremely shallow, clear water).

[0132] Feature extraction based on SWOT data:

[0133] High-precision water surface elevation (WSE), the high-precision WSE value corresponding to each water body pixel, is one of the core inputs of this method, providing a crucial reference plane for water depth inversion. Water surface slope (WSS), the slope of the water surface in the X and Y directions, directly provided by SWOT data or calculated. WSS reflects local hydrodynamic conditions and may indirectly indicate changes in underwater topography. Shoreline distance, calculated based on the precise shoreline extracted from SWOT or optical imagery, is the Euclidean distance from each water body pixel to the nearest shoreline. This feature utilizes the prior knowledge that water depth generally increases with distance from the shore. SWOT water body information, water body pixel classification or water body extent information provided by SWOT, is used to accurately define the effective area for model training and prediction.

[0134] Feature vector construction and fusion:

[0135] Various features extracted from optical images and SWOT data will be spatially aligned and fused at a uniform pixel (or grid unit) scale to construct a high-dimensional feature vector for each water pixel.

[0136] ④ Machine learning model construction and water depth inversion.

[0137] Training sample construction:

[0138] Label data (real water depth): Ideally, measured water depth data (such as sonar or LiDAR measurements) covering a representative portion of the study area are used as the real labels. The measured water depth points are precisely matched to the geographical locations of their corresponding pixels, and the fused feature vector of that pixel is extracted to form training sample pairs of (feature vector, measured water depth).

[0139] Machine learning model selection and configuration:

[0140] Use machine learning models suitable for handling high-dimensional inputs and complex nonlinear regression tasks. Alternative models include: Random Forest Regression (RFR), Gradient Boosting Regression (GBR), such as XGBoost, LightGBM, and CatBoost, Support Vector Regression (SVR), Artificial Neural Networks (ANNs), such as Multilayer Perceptron (MLP); or variants of Convolutional Neural Networks (CNNs) that incorporate neighborhood information (if local pixel neighborhoods are used as image patch input). Select a suitable model based on data characteristics and computational resources, and perform preliminary structural design or parameter configuration.

[0141] Model training and optimization:

[0142] Dataset partitioning involves dividing the constructed training sample set into a training set and independent validation / test sets according to a certain ratio (e.g., 70 / 30 or 80 / 20). Model training uses the training set to learn the parameters of the selected machine learning model. Hyperparameter optimization employs cross-validation techniques, combined with strategies such as grid search, random search, or Bayesian optimization, to systematically optimize the model's key hyperparameters in order to obtain the best generalization performance.

[0143] ⑤ Generation and output of underwater topography in lakes.

[0144] Model-based water depth prediction:

[0145] The trained and optimized optimal machine learning model is applied to the fused feature vectors of all water pixels within the target lake area to predict the water depth value for each pixel. Using the high-precision water surface elevation (WSE_SWOT) provided for the corresponding pixel in SWOT analysis, the model-predicted water depth (Depth_predicted) is converted into lakebed elevation (Bed_Elevation).

[0146] Bed_Elevatio=WSE_SWOT-Depth_predicted;

[0147] Output:

[0148] Generate a raster map of the lake bottom elevation model.

[0149] Post-processing of results:

[0150] The generated lakebed elevation or depth map is then subjected to necessary post-processing, such as spatial smoothing filtering, outlier removal and repair, and contour line generation and mapping.

[0151] ⑥ Accuracy verification and evaluation.

[0152] Independent validation dataset:

[0153] The final lake topography product was rigorously evaluated for accuracy using independent measured water depth data (“test set”) that was not involved in model training and optimization.

[0154] Quantitative evaluation indicators:

[0155] Root mean square error, mean absolute error, coefficient of determination, mean relative error, or percentage deviation.

[0156] In summary, this embodiment includes the following technical solutions:

[0157] Key technical points:

[0158] (1) Multi-source heterogeneous data fusion strategy: How to effectively integrate pixel-level optical spectral features with higher-level structural hydrological information (such as WSE grid, water body range vector / grid, water surface slope) provided by SWOT at the feature level or model level to maximize information complementarity.

[0159] (2) Innovative feature extraction and application based on SWOT data: How to transform SWOT data such as WSE, water body extent, and water surface slope into the most effective input features for machine learning models to predict lake bottom topography. For example, using WSE as an absolute elevation reference to constrain relative water depth, or using multi-temporal WSE and water body extent changes to infer the reservoir capacity characteristics of the lake to assist in topographic inversion.

[0160] (3) Optimization of machine learning models for fusion features: Select or design a machine learning model that can effectively handle multiple types of features such as spectrum, elevation, slope, and distance, and perform targeted structural optimization, loss function design or training strategy adjustment to improve the learning efficiency and prediction accuracy of the model.

[0161] (4) Adaptability and transferability of the model in the case of sparse or no ground truth: How can a lake topography prediction model with a certain generalization ability still be trained by means of semi-supervised learning, transfer learning or pseudo-labels generated by using SWOT data when the measured water depth data is extremely limited?

[0162] More specifically:

[0163] (1) A method for predicting lake topography, which involves acquiring satellite (multispectral / hyperspectral) band images and SWOT satellite data (including at least water surface elevation and / or water body range) of the target lake, preprocessing and extracting features from the two types of data, fusing the extracted band image features with the SWOT data features, and using the fused feature set to predict the underwater topography of the lake through a machine learning model.

[0164] (2) Specific methods of feature fusion: Protect the technical solution of combining features such as spectral reflectance and water depth correlation index of satellite band images with features such as water surface elevation, water surface slope and distance to shoreline of SWOT at the pixel (or grid) level to form a multi-dimensional feature vector for use as input to machine learning models.

[0165] (3) Specific applications of machine learning models: Protect the use of specific categories or improved machine learning models (such as random forests, gradient boosters, neural networks, etc. optimized for such fused data) to achieve terrain prediction by learning the mapping relationship between fused features and water depth (or lake bottom elevation).

[0166] (4) Methods of using SWOT data to constrain models or calibration results: Protect the method of using the high-precision water surface elevation provided by SWOT as part of the training label (e.g., if the predicted lake bottom elevation is, then the lake bottom elevation = SWOT WSE - predicted water depth), or as the absolute elevation benchmark of the predicted results for calibration.

[0167] The lake topography prediction system based on the above methods is an integrated system comprising a data acquisition module, a preprocessing module, a feature extraction and fusion module, a machine learning model training and prediction module, and a topography result generation and output module. In cases where measured water depths are scarce or sparse, SWOT data is used to assist in generating training samples or improving model robustness.

[0168] The beneficial effects of this embodiment:

[0169] In terms of originality: This embodiment addresses the bottlenecks of traditional underwater topographic surveying methods for lakes, such as high cost, low efficiency, poor timeliness, and difficulty in applying them to vast, complex, or sparsely populated waters. Focusing on critical needs such as sustainable water resource utilization, precise water environment management, and flood control and disaster reduction, it proposes an intelligent underwater topographic inversion technology for lakes that deeply integrates multi-type satellite remote sensing observation data with advanced machine learning algorithms. This technology aims to improve the efficiency, accuracy, and coverage of underwater topographic mapping. It innovatively constructs a collaborative inversion framework combining optical satellite image features (reflecting water optical properties and water column attenuation) and unique SWOT satellite observation data (high-precision water surface elevation (WSE), water body extent, and water surface slope). By constructing a more information-rich and physically constrained multi-source feature set and utilizing machine learning models to mine the complex nonlinear relationships between these features and underwater topography, it achieves a revolutionary breakthrough over traditional methods. This method overcomes the limitations of previous methods that relied on insufficient information from single remote sensing data sources or overly strong physical model assumptions. It provides a novel technical approach for rapidly and accurately acquiring underwater topographic information of large-scale lakes, demonstrating significant originality and technological leadership.

[0170] In terms of value: The technical framework proposed in this embodiment closely addresses the urgent need to improve the acquisition of basic geographic information for lakes, and its value is reflected in multiple aspects. First, by efficiently and cost-effectively utilizing multi-source satellite remote sensing data to replace or supplement traditional measurement methods, the economic and time costs of underwater topographic mapping of lakes are greatly reduced, making it possible to conduct topographic surveys of lakes lacking actual measurement data. Second, the high-precision, high-timeliness underwater topographic data (DBEM / DBM) produced by this technology are indispensable key input parameters for hydrological, hydrodynamic, water environment, and ecological models, significantly improving the simulation accuracy and prediction reliability of related models. This provides solid data support for precise reservoir scheduling, water balance analysis, flood risk assessment, water quality simulation, ecosystem health assessment, waterway planning, and water-related engineering design. This technology has the advantages of being non-contact, large-scale, and repeatable, making it particularly suitable for monitoring remote areas, environmentally sensitive areas, or dynamically changing water bodies. It provides core technical support for achieving refined, intelligent management and sustainable development of lake water resources, and has significant strategic importance and broad application prospects for addressing water security challenges under the background of climate change.

[0171] In terms of practical effectiveness: The lake underwater topography prediction technology based on multi-source satellite data fusion and machine learning proposed in this embodiment has undergone thorough methodological design and (future or already conducted) case verification, demonstrating strong practical application performance. This technology effectively overcomes the bottleneck of insufficient information from a single remote sensing data source, providing a direct vertical benchmark constraint for water depth inversion through high-precision water surface elevation information provided by SWOT satellites, significantly improving prediction accuracy. The introduction of machine learning models enables it to adaptively learn complex relationships from the data, reducing dependence on prior water body optical parameters and enhancing the method's universality for lakes with different water quality conditions. The underwater topography results produced by this method (such as RMSE, MAE, R...) 2 Based on various metrics, this technology is expected to achieve or surpass the accuracy levels of some existing remote sensing inversion methods, and boasts unparalleled advantages in data acquisition costs, mapping efficiency, and coverage. The technology has a clear workflow, and key steps (such as data preprocessing, feature extraction, model training, and terrain generation) are easily automated and standardized. It possesses enormous potential to move from research to operational applications, enabling rapid response to emergency monitoring needs (such as terrain assessment within flood-prone areas), and demonstrates significant practical benefits and operability.

[0172] In terms of systemic aspects: The underwater topography prediction technology for lakes proposed in this embodiment is not a simple superposition of single technologies, but a complete technical system encompassing multi-source heterogeneous data acquisition, collaborative preprocessing, deep feature engineering, intelligent model construction, and final product generation and verification. This system emphasizes the collaborative observation advantages of optical remote sensing and radar remote sensing (SWOT's KaRIn), systematically designing feature extraction and fusion strategies targeting water column optical characteristics, water surface geometry, and shoreline spatial relationships to maximize information utilization. The application of machine learning models constitutes the core of intelligent inversion, achieving end-to-end mapping from high-dimensional features to underwater topography. Furthermore, this method, through direct linkage with WSE data provided by SWOT, ensures that the predicted "water depth" can be accurately converted into "lake bottom elevation" with a clear elevation benchmark, forming a systematic three-dimensional spatial information cognition scheme from the water surface to the bottom. This technology system not only focuses on the accuracy of terrain prediction, but also takes into account the automation potential and applicability of the method. It provides a key link for building digital twin lakes and realizing routine remote sensing monitoring and intelligent management of lakes. It is a systematic and innovative solution in the field of lake information science.

[0173] Reference Figure 3 This application also provides a lake topography prediction device based on machine learning and multi-source satellite data fusion, which can implement the above-mentioned lake topography prediction method based on machine learning and multi-source satellite data fusion. The device includes:

[0174] An optical feature extraction unit is used to extract the optical features of the target lake from optical satellite imagery;

[0175] The water feature extraction unit is used to extract the water features of each water cell in the target lake from SWOT satellite data.

[0176] The feature fusion unit is used to perform spatiotemporal alignment and fusion of the optical features and the water features to obtain a fused feature vector.

[0177] The sample construction unit is used to match the measured water depth points of the real water depth to the geographical location of the corresponding pixel, and then to form training samples by combining the real water depth and the fused feature vector of the corresponding pixel.

[0178] A model initialization unit is used to select a machine learning model and initialize the structure or parameters of the machine learning model.

[0179] A model training unit is used to train the initialized machine learning model using the training samples.

[0180] A water depth prediction unit is used to predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel using the trained machine learning model.

[0181] A terrain determination unit is used to determine the terrain of the target lake based on each of the water depth values.

[0182] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0183] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method of this application. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0184] It is understood that the content of the above method embodiments is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the methods of this application, and the beneficial effects achieved are the same as those achieved by the methods of this application.

[0185] Please see Figure 4 , Figure 4 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0186] The processor 401 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0187] The memory 402 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 402 and is called and executed by the processor 401.

[0188] Input / output interface 403 is used to implement information input and output;

[0189] The communication interface 404 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0190] Bus 405 transmits information between various components of the device (e.g., processor 401, memory 402, input / output interface 403, and communication interface 404);

[0191] The processor 401, memory 402, input / output interface 403 and communication interface 404 are connected to each other within the device via bus 405.

[0192] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of this application.

[0193] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0194] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0195] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0196] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0197] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0198] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0199] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0200] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0201] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0202] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0203] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0204] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0205] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A lake topography prediction method based on machine learning and fusion of multi-source satellite data, characterized in that, The method includes the following steps: Optical features of the target lake were extracted from optical satellite imagery; Water features of each water body pixel in the target lake are extracted from SWOT satellite data; wherein, the water features include water surface elevation, water surface slope, shoreline distance and SWOT water body information of each water body pixel in the target lake extracted from SWOT satellite data; The optical features and the water features are spatiotemporally aligned and fused to obtain a fused characteristic vector; wherein, spatial alignment and fusion are performed at a uniform pixel or grid unit scale to construct a fused characteristic vector for each water body pixel; The measured water depth points of the actual water depth are matched to the geographical locations of the corresponding pixels, and then the actual water depth and the fused feature vector of the corresponding pixel are used to form training samples. Select a machine learning model and initialize the structure or parameters of the machine learning model; The initialized machine learning model is trained using the training samples. The trained machine learning model is used to predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel; The topography of the target lake is determined based on each of the stated water depth values; The process of training the initialized machine learning model using the training samples includes the following steps: The training set is obtained by dividing the training samples; The training set is used to learn the parameters of the machine learning model; The target hyperparameters of the machine learning model are optimized using cross-validation combined with grid search, random search, or Bayesian optimization strategies. Before extracting the optical features of the target lake from the optical satellite imagery, the method further includes the following steps: The optical satellite imagery is subjected to radiometric correction, geometric correction and registration, water body extraction, and removal of clouds, cloud shadows and other interference. Before extracting the water features of each water cell in the target lake from the SWOT satellite data, the method further includes the following steps: Outlier removal, spatiotemporal matching and grid unification, and water body boundary optimization were performed on the SWOT satellite data. Determining the topography of the target lake based on each of the water depth values ​​includes the following steps: The water depth values ​​are converted into lake bottom elevations using the water surface elevations of the water body pixels in the SWOT satellite data. ； in, The elevation of the lake bottom. This refers to the water surface elevation. This represents the water depth. Generate a raster map of the lakebed elevation; Spatial smoothing filtering, outlier removal and repair, and contour line generation and mapping are performed on the raster image to obtain the topographic map of the target lake.

2. The lake topography prediction method based on machine learning and multi-source satellite data according to claim 1, characterized in that, The extraction of optical features of the target lake from optical satellite imagery includes the following steps: The optical features, including band reflectance, band ratio, logarithmic transformation characteristics, and water index of the target lake, are extracted from optical satellite imagery.

3. The lake topography prediction method based on machine learning and multi-source satellite data according to claim 1, characterized in that, The process of selecting a machine learning model and initializing its structure or parameters includes the following steps: Based on the characteristics of the training samples and computational resources, the machine learning model is selected from random forest regression model, gradient boosting regression model, artificial neural network model, or convolutional neural network variant model that incorporates neighborhood information, and the structure or parameters of the machine learning model are initialized.

4. A lake topography prediction device based on machine learning and multi-source satellite data, characterized in that, The apparatus is used to implement the lake topography prediction method based on machine learning and fusion of multi-source satellite data as described in claim 1, the apparatus comprising: An optical feature extraction unit is used to extract the optical features of the target lake from optical satellite imagery; The water feature extraction unit is used to extract the water features of each water cell in the target lake from SWOT satellite data. The feature fusion unit is used to perform spatiotemporal alignment and fusion of the optical features and the water features to obtain a fused feature vector. The sample construction unit is used to match the measured water depth points of the real water depth to the geographical location of the corresponding pixel, and then to form training samples by combining the real water depth and the fused feature vector of the corresponding pixel. A model initialization unit is used to select a machine learning model and initialize the structure or parameters of the machine learning model. A model training unit is used to train the initialized machine learning model using the training samples. A water depth prediction unit is used to predict the water depth value of the corresponding pixel based on the fusion characteristic vector of each water pixel using the trained machine learning model. A terrain determination unit is used to determine the terrain of the target lake based on each of the water depth values.

5. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 3.

Images