A reservoir capacity curve reconstruction method and system based on multi-source data fusion

Abstract

The present application relates to the technical field of reservoir capacity curve reconstruction, in particular to a reservoir capacity curve reconstruction method and system based on multi-source data fusion, comprising the following steps: obtaining multi-source data fusion elevation information, reconstructing the water level-storage capacity relationship curve, and outputting the reservoir capacity curve reconstruction graphical expression result. In the present application, the water body boundary surface information is obtained by fusing the unmanned aerial vehicle remote sensing image to identify the boundary pixel area, the abnormal water level line number is identified by combining the difference index comparison, the complete three-dimensional fusion data set is constructed by matching the water point cloud and the underwater elevation data, the grid network is generated according to the unified projection and the abnormal elevation points are corrected, the continuous trajectory line sequence is established and integrated to form the two-dimensional boundary expression, which can quickly identify the reservoir capacity change in the dynamic environment, effectively avoid the incomplete boundary extraction and the data precision decline caused by the complex terrain and the tedious operation in the traditional measurement method, and improve the comprehensive performance of the reservoir capacity curve reconstruction in the spatial coverage, data updating and expression precision.

Description

Technical Field

[0001] This invention relates to the field of reservoir capacity curve reconstruction technology, and in particular to a reservoir capacity curve reconstruction method and system based on multi-source data fusion. Background Technology

[0002] The field of reservoir capacity curve reconstruction technology mainly involves the establishment and correction of reservoir capacity curves in hydrological and water resources management. This includes expressing the relationship between reservoir capacity and different water levels, diversifying data acquisition methods, and applying mathematical models in curve fitting and reconstruction. Typically, spatial analysis and numerical modeling are performed based on information sources such as measured data, remote sensing imagery, and DEM (Digital Elevation Model). The functional relationship between water level and reservoir capacity is constructed through interpolation, fitting, or inversion. Traditional reservoir capacity curve reconstruction methods involve obtaining the water surface boundary line at a specific water level through manually measured cross-sectional data, and then estimating the reservoir capacity value corresponding to each water level using area-elevation relationships and surface accumulation addition. The main steps include field measurement of representative cross-sections in the reservoir area, manual drawing of contour maps, calculation of the area at each elevation, and volume integration based on vertical projection. This method is cumbersome and easily affected by terrain complexity and observation errors, making it difficult to adapt to the needs of reconstructing dynamically changing reservoir capacity.

[0003] In existing technologies, the reconstruction of reservoir capacity curves relies on manually measured cross-sections to obtain boundary line information below the water level. In complex terrain areas, this is often limited by the accessibility of the site and the density of cross-sections, making it difficult to accurately reflect the true shape of the reservoir water body boundary. In addition, this method requires manual drawing and calculation of area-elevation relationships, which is cumbersome, greatly affected by human factors, and difficult to control measurement accuracy. Observational errors are easily amplified during the calculation process. Under dynamic hydrological conditions, it cannot effectively respond to rapid changes in reservoir capacity, making it difficult for the reconstruction results to meet the dual requirements of high timeliness and high accuracy, thus posing significant limitations in the application of water resource scheduling and management. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for reconstructing the capacity curve based on multi-source data fusion.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for reconstructing a storage capacity curve based on multi-source data fusion, comprising the following steps:

[0006] S1: Acquire UAV remote sensing images and ground control point layout layers, identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets.

[0007] S2: Extract the area value from the water body boundary surface layer dataset, match it with the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, and output the area mismatch boundary identification information.

[0008] S3: Based on the water level line number area in the area mismatch boundary identification information, match the reservoir surface elevation point cloud sequence and underwater elevation scatter data, perform unified geographic projection transformation, extract the distance between adjacent elevation points for filtering, and construct a set of surface and underwater fused elevation data.

[0009] S4: Extract the above-water and underwater fused elevation data set, construct an equally spaced regular grid, generate contour line closed graphics for grid cell boundary points, perform interpolation correction on deviation points, and output a standard coordinate isoelastic grid layer.

[0010] S5: Based on the standard coordinate isoelastic grid layer, and according to the correspondence between the two-dimensional grid and the water level, establish a sequence of spatial node trajectory lines, integrate them in order of water level, and output the reconstructed graphical representation of the reservoir capacity curve.

[0011] As a further embodiment of the present invention, the water body boundary surface layer dataset includes boundary closed polygons and water body area information; the area mismatch boundary identification information includes difference index values, water level line numbers, and area comparison status; the water surface and underwater integrated elevation data set includes unified coordinate points, elevation difference indexes, and filler point density information; the standard coordinate isopleth elevation grid layer includes raster boundary contour lines, corrected interpolation points, and elevation grid cells; and the reservoir capacity curve reconstruction graphic representation result includes two-dimensional boundary trajectories, numbered water level line positions, and spatial point matrix trajectory line sequences.

[0012] As a further aspect of the present invention, the step of obtaining the water body boundary surface layer dataset specifically includes:

[0013] S111: Acquire UAV remote sensing imagery and ground control point layout layers, read all pixel grayscale values ​​in the image histogram, perform difference calculations for grayscale value changes, search for areas where the grayscale difference between adjacent pixels is greater than the preset grayscale change threshold, record the corresponding spatial coordinates, and generate a set of grayscale change boundary pixels.

[0014] S112: Based on the set of gray-scale abrupt boundary pixels, extract the adjacent coordinate index information of each boundary pixel, determine whether the adjacent indices are continuous and form a closed boundary path, and if the connection condition is met, then sequentially splice the boundary points to construct a spatial closed structure and establish a set of boundary contour polygon regions.

[0015] S113: Based on the set of polygon regions with boundary contours, perform statistical operations on the number of pixels in each polygon region, and multiply it with the actual ground area parameter of a single pixel to generate the surface cover area information of each region and establish a water body boundary surface layer dataset.

[0016] As a further aspect of the present invention, the step of obtaining the area mismatch boundary identification information specifically comprises:

[0017] S211: Obtain the water body boundary surface layer dataset, extract the surface coverage area value of each planar patch in the layer, and index the patch number to correspond with the water level line number in the original reservoir capacity curve data. Retrieve the water surface area value recorded by the corresponding water level line in the original reservoir capacity curve, establish a pairing relationship between the two sets of area values, and generate a water surface area pairing matrix.

[0018] S212: Based on the water surface area pairing matrix, calculate the numerical difference between each pair of area values, and compare it item by item with the preset area comparison benchmark difference limit. Mark all water level line number segments whose numerical difference is greater than the preset area comparison benchmark difference limit to obtain the area anomaly number set.

[0019] S213: Based on the set of area anomaly numbers, locate the corresponding water level line number in the water body boundary surface layer, extract the spatial contour data of the corresponding boundary area, assign area mismatch markers, aggregate all marked boundary areas, and output area mismatch boundary identification information.

[0020] As a further aspect of the present invention, the step of acquiring the integrated elevation data set of above-water and underwater areas specifically comprises:

[0021] S311: Based on the water level line number area in the area mismatch boundary identification information, match the reservoir area lidar echo data and underwater elevation scatter data collected by the depth sounding device under the corresponding number, extract the spatial location coordinates and elevation values ​​of the two types of data respectively, and perform a unified geographic projection parameter conversion to establish a unified projection elevation point set.

[0022] S312: Based on the unified elevation point set of the projection, extract the spatial coordinates of adjacent point pairs, calculate the horizontal distance, filter point pairs whose adjacent distance is less than the point spacing filtering threshold, and determine whether the difference between the corresponding elevation values ​​is less than a set value. Aggregate the point pairs that meet the conditions to generate a set of continuous elevation point pairs.

[0023] S313: Based on the set of continuous elevation points, interpolate to generate intermediate points in space, calculate the elevation interpolation of the intermediate points according to the distance weight, add all interpolated points to the original elevation point set, merge all elevation point information, and establish a fused elevation data set for both above and below water.

[0024] As a further aspect of the present invention, the step of obtaining the standard coordinate isoelastic grid layer specifically comprises:

[0025] S411: Based on the three-dimensional spatial coordinates and elevation values ​​in the above-water and underwater fusion elevation data set, reconstruct all points into geographic coordinate data frames under a unified coordinate system to generate a projection-consistent coordinate dataset.

[0026] S412: Based on the projection-consistent coordinate dataset, construct a regular grid structure at a set interval, extract the boundary nodes of each grid cell, perform interpolation calculations according to the elevation values ​​corresponding to the nodes, and connect points with the same elevation values ​​to form a closed shape, thereby obtaining a set of continuous equal-elevation closed shapes.

[0027] S413: Based on the elevation values ​​and spatial coordinate information of each boundary point in the continuous contour closed graphic set, the spatial distance between the grid deviation point and the neighboring point is measured and calculated. Elevation value interpolation correction is performed in a distance-weighted manner. The corrected elevation value of the boundary point is obtained by calculation. The correction result is mapped to the original grid cell, the contour line boundary is updated, and a standard coordinate isoelastic grid layer is established.

[0028] As a further aspect of the present invention, the formula for calculating the boundary point correction elevation value is as follows:

[0029] ;

[0030] in, The boundary point elevation correction value represents the raster deviation point. Indicates the first Elevation values ​​of neighboring points, Indicates the deviation point and the first Euclidean distance between neighboring points Indicates the distance-weighted index. This represents the total number of neighboring points.

[0031] As a further aspect of the present invention, the step of obtaining the graphical representation result of the reconstructed storage capacity curve is specifically as follows:

[0032] S511: Based on the spatial position coordinates and corresponding elevation values ​​of each grid cell in the standard coordinate isoelastic grid layer, and combined with the set water level line numbering partition information, extract the boundary point positions of all grid cells under each water level line number, and connect the boundary points in sequence according to the two-dimensional grid spatial order to construct an initial point path set and generate a two-dimensional spatial point path line sequence.

[0033] S512: Based on the two-dimensional spatial dot matrix trajectory line sequence, sort the numbers in ascending order according to the water level line numbering, identify the closed correspondence between adjacent trajectory lines, perform sequential aggregation operation to integrate each segment trajectory, obtain the complete trajectory boundary structure under all numbered segments, and establish a two-dimensional boundary trajectory set.

[0034] S513: Based on all closed boundary structures in the set of two-dimensional boundary trajectories, extract the spatial position sequence and sorting number of each trajectory node, reconstruct and map adjacent boundary sequences and establish a linear connection index, combine all closed trajectories, transform them into two-dimensional graphic representation entities according to the trajectory structure, and output the reconstructed graphic representation result of the reservoir capacity curve.

[0035] A system for reconstructing storage capacity curves based on multi-source data fusion includes:

[0036] The remote sensing boundary extraction module is used to perform S1: acquire UAV remote sensing images and ground control point layout layers, identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets.

[0037] The area difference analysis module is used to perform S2: extract the area value from the water body boundary surface layer dataset, match the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, and output the area mismatch boundary identification information.

[0038] The elevation data fusion module is used to perform S3: based on the water level line number area in the area mismatch boundary identification information, it matches the reservoir surface elevation point cloud sequence and underwater elevation scatter data, performs unified geographic projection transformation, extracts the distance between adjacent elevation points for filtering, and constructs a set of surface and underwater fused elevation data.

[0039] The iso-grid standardization module is used to perform S4: extract the above-water and underwater integrated elevation data set, construct an equally spaced regular grid, generate contour line closed graphics for grid cell boundary points, perform interpolation correction on deviation points, and output a standard coordinate iso-elevation grid layer;

[0040] The curve reconstruction modeling module is used to execute S5: based on the standard coordinate isoelastic grid layer, according to the correspondence between the two-dimensional grid and the water level, establish a sequence of spatial node trajectory lines, integrate them in order of water level, and output the result of the reservoir capacity curve reconstruction graphic representation.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In this invention, water body boundary information is obtained by fusing UAV remote sensing images to identify boundary pixel regions, and abnormal water level line numbers are identified by comparing difference indicators. A complete three-dimensional fusion dataset is constructed by matching surface point cloud and underwater elevation data. A raster is generated based on a unified projection and abnormal elevation points are corrected. A continuous trajectory line sequence is established and integrated to form a two-dimensional boundary representation. This achieves rapid identification and high-precision representation of reservoir capacity changes in a dynamic environment, effectively avoiding the problems of incomplete boundary extraction and decreased data accuracy caused by complex terrain and cumbersome operation in traditional measurement methods. This improves the comprehensive performance of reservoir capacity curve reconstruction in terms of spatial coverage, data update, and representation accuracy. Attached Figure Description

[0043] Figure 1 This is a flowchart of the main steps of the present invention;

[0044] Figure 2 This is a flowchart illustrating the process of acquiring the water body boundary surface layer dataset according to the present invention.

[0045] Figure 3 This is a flowchart illustrating the process of obtaining area mismatch boundary identification information according to the present invention.

[0046] Figure 4 This is a flowchart illustrating the process of acquiring the combined surface and underwater elevation data set according to the present invention.

[0047] Figure 5 This is a flowchart illustrating the process of obtaining the standard coordinate iso-elevation grid layer of this invention.

[0048] Figure 6 This is a flowchart illustrating the process of obtaining the graphical representation result of the reconstructed library capacity curve in this invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0051] Please see Figure 1A method for reconstructing storage capacity curves based on multi-source data fusion includes the following steps:

[0052] S1: Acquire UAV remote sensing images and ground control point layout layers, read image histogram grayscale differences to identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets.

[0053] S2: Extract the area value from the water body boundary surface layer dataset, match it with the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, quantitatively compare it with the preset area comparison benchmark difference limit, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, establish a set of difference marker sections, and output the area mismatch boundary identification information.

[0054] S3: Based on the water level line number area in the area mismatch boundary identification information, match the water surface elevation point cloud sequence formed by the lidar echo in the reservoir area with the underwater elevation scatter data collected by the depth sounding device, perform unified geographic projection transformation, extract the distance between adjacent elevation points for filtering, and perform point cloud densification and completion processing on areas with elevation difference less than 0.5 meters to construct a water surface and underwater fused elevation data set;

[0055] S4: Extract the three-dimensional spatial coordinates and elevation values ​​from the above-water and underwater fusion elevation data set, construct an equally spaced regular grid based on the set coordinate datum WGS-84, generate contour line closed graphics for grid cell boundary points, and perform interpolation correction on deviation points to output a standard coordinate isoelastic grid layer.

[0056] S5: Based on the spatial position of each grid cell in the standard coordinate isoelastic grid layer and the corresponding water level line number area, establish a spatial node trajectory line sequence according to the correspondence between the two-dimensional grid and the water level, integrate them in order of water level to form a complete two-dimensional boundary trajectory set, and output the reconstructed reservoir capacity curve graphic representation result.

[0057] The water body boundary surface layer dataset includes closed boundary polygons and water body area information. The area mismatch boundary identification information includes difference index values, water level line numbers, and area comparison status. The integrated elevation data set above and below water includes unified coordinate points, elevation difference indexes, and filler point density information. The standard coordinate isopleth elevation grid layer includes raster boundary contour lines, corrected interpolation points, and elevation grid cells. The reservoir capacity curve reconstruction graphic representation results include two-dimensional boundary trajectories, numbered water level line positions, and spatial point matrix trajectory line sequences.

[0058] Please see Figure 2 Step S1 is as follows:

[0059] S111: Acquire UAV remote sensing imagery and ground control point layout layers, read all pixel grayscale values ​​in the image histogram, perform difference calculations for grayscale value changes, search for areas where the grayscale difference between adjacent pixels is greater than the preset grayscale change threshold, record the corresponding spatial coordinates, and generate a set of grayscale change boundary pixels.

[0060] In the process of acquiring UAV remote sensing imagery and ground control point (GCP) layers, as the primary step in building the data foundation for reconstructing the reservoir capacity curve, the first step is to execute flight mission planning, select a flight platform, and equip it with a full-frame aerial survey camera. The flight altitude is set at 120 meters, the forward overlap rate is set at 80%, and the lateral overlap rate is set at 70% to ensure multi-angle coverage and geometric accuracy of the image data. The deployment of ground control points (GCPs) follows the principle of "gridization + boundary control," with no fewer than 15 control points evenly distributed around the perimeter and center of the survey area. RTK-GPS receivers (such as high-precision GNSS receivers) are used to collect the three-dimensional coordinates (X, Y, Z) of each control point. The coordinate system adopted is CGCS2000, and the elevation datum is the 1985 National Elevation Datum.

[0061] After data acquisition, the raw image data was imported into the photogrammetric workstation. Aerial triangulation was performed using the coordinates of the ground control points to generate an orthophoto (DOM) with a ground resolution (GSD) better than 5 cm. This high-resolution image provides pixel-level observational data support for subsequent accurate extraction of water area and, consequently, inference and reconstruction of the reservoir capacity curve. Subsequently, histogram reading was performed. The orthophoto was converted into a single-channel grayscale matrix with a matrix size of [missing information]. ,in Traverse each coordinate in the matrix. Extract the grayscale value of the pixel at that location. The value range is [0, 255].

[0062] This algorithm performs interpolation calculations based on grayscale value changes to identify features at the boundary between land and water. Instead of directly calling edge detection functions, it uses a sliding window algorithm. A defined... A sliding window with the current cell Centered on the axis, calculate its gradient magnitudes in the horizontal and vertical directions. Horizontal gradient Vertical gradient Overall gradient magnitude Through formula The calculation yielded the result.

[0063] Set the grayscale mutation detection threshold as The threshold was determined based on a variant of the OTSU (Otsu's method) experiment. In the experiment, typical land-water interface sample areas were selected, and gradient histograms were statistically analyzed. It was found that the mean gradient in non-boundary areas was approximately 5, with a standard deviation of 2, while the mean gradient in boundary areas exceeded 45. To ensure high confidence, a threshold was set... The program compares and calculates them one by one. and ,when When the pixel is identified as a potential boundary point, its spatial coordinates are extracted. The data is then stored in a dynamic array, ultimately generating a set of grayscale abrupt boundary pixels.

[0064] S112: Based on the set of boundary pixels with gray-scale abrupt changes, extract the adjacent coordinate index information of each boundary pixel, determine whether the adjacent indices are continuous and form a closed boundary path, and if the connection condition is met, then sequentially splice the boundary points to construct a spatial closed structure and establish a set of polygonal regions of the boundary outline.

[0065] Based on the generated set of grayscale abrupt boundary pixels, a boundary tracking and closure detection procedure is initiated. This is a crucial step in ensuring the accuracy of calculating the closed water surface area corresponding to a single water level in the reservoir capacity curve reconstruction. First, the discrete points in the set are mapped back to a two-dimensional raster space to create a binary mask image, where boundary pixels have a value of 1 and the background has a value of 0. An 8-neighborhood connectivity search algorithm is then used to extract the neighbor coordinate index information of each boundary pixel. For any boundary point... Scan the eight surrounding pixels in a clockwise direction. Find neighboring points with a value of 1. .

[0066] The logic for determining whether adjacent indices form a closed boundary path is as follows: Initialize a path stack, and set the starting point... Push it onto the stack and mark it as "visited". During the tracing process, if the current point... If there are unvisited connection points within the search radius, add them to the path stack; if the search returns to the previous path, add them to the path stack. If the path length is greater than the set minimum perimeter threshold (e.g., 50 pixels to exclude noise points), then the connection condition is satisfied.

[0067] If the connection conditions are met, the boundary points are sequentially pieced together. For broken boundaries (i.e., cases where the search is interrupted but not closed), morphological closing operations are used for repair. Specifically, a circular structuring element with a radius of 3 pixels is defined. First, the binary image is expanded to fill gaps smaller than 3 pixels, and then an erosion operation is performed to restore the original shape, thereby constructing a physically continuous spatial closed structure.

[0068] Finally, the raster coordinate sequence is converted into vector polygon data. The Douglas-Peucker algorithm is used to thin out the boundary points, with a distance tolerance of 0.1 meters, to remove redundant nodes while preserving shape features. The processed node sequence is then encapsulated into Polygon objects (WKT format) according to the OGC standard and assigned a unique ID, thus establishing a set of boundary contour polygon regions.

[0069] S113: Based on the set of polygon regions with boundary contours, perform statistical operations on the number of pixels in each polygon region, and multiply it with the actual ground area parameter of a single pixel to generate the surface cover area information of each region and establish a water body boundary surface layer dataset.

[0070] Based on the set of polygon regions with boundary contours, a statistical operation is performed on the number of pixels within each polygon region. Using the inverse process of the scan-line fill algorithm, the total number of pixels contained within each polygon object is counted. .

[0071] Obtain the actual ground area parameter for a single pixel. Since the image has undergone orthorectification, the ground resolution (GSD) remains consistent across the entire map. If the GSD is 0.05 meters (i.e., 5 centimeters), then the actual ground area of ​​a single pixel is... square meters.

[0072] Perform a product operation to generate land cover area information for each region. For the [specific region / region]... A polygonal region, its surface coverage area The calculation formula is:

[0073] ;

[0074] For example, if a water feature contains 1,200,000 pixels, then its area is... square meters.

[0075] To ensure the integrity of the geographic attributes of the data, the calculated area value was written as the attribute field "Area_m2" into the attribute table of the vector data. Simultaneously, projection information (e.g., PROJCS["CGCS2000_3_Degree_Gauss_Kruger_Zone_37"]) was defined for the layer based on the previously collected GCP coordinates. Finally, the vector data containing geometric and attribute information was saved in Shapefile or GeoJSON format to establish a water body boundary layer dataset, providing high-precision measured water surface area node data for reservoir capacity curve reconstruction.

[0076] Please see Figure 3 Step S2 is as follows:

[0077] S211: Obtain the water body boundary surface layer dataset, extract the surface cover area value of each isometric patch in the layer, and index the patch number to correspond with the water level line number in the original reservoir capacity curve data. Retrieve the water surface area value recorded by the corresponding water level line in the original reservoir capacity curve, establish the pairing relationship between the two sets of area values, and generate a water surface area pairing matrix.

[0078] The water body boundary surface layer dataset is obtained, and the attribute table is read through the GIS data parsing interface to extract the surface cover area value and corresponding patch ID of each isometric patch in the layer. This aims to align the latest measured data with the historical model, which is a prerequisite for determining whether to initiate the reservoir capacity curve reconstruction process.

[0079] Simultaneously, the original reservoir capacity curve data is loaded. This data is typically stored in CSV or Excel format, containing "water level elevation (m)" and "water surface area (m)". A mapping table is needed to establish the correspondence between the map patch ID and the water level line ID. To achieve accurate indexing, a mapping mechanism between map patch IDs and water level line IDs needs to be established. The naming rules for map patch IDs should include water level information; for example, a map patch ID of "WB_145_05" represents a water level elevation of 145.5 meters.

[0080] The water level value obtained by analyzing the map patch number is retrieved from the original reservoir capacity curve data. Since the reservoir capacity curve is usually the result of discrete point measurements (e.g., measured every 0.5 meters), if the water level value corresponding to the map patch does not have an exact match in the reservoir capacity table, the theoretical water surface area is calculated using linear interpolation.

[0081] Assume that the storage capacity table contains records: and If the water level corresponding to the patch is The retrieved original reservoir capacity curve contains the water surface area value recorded at the corresponding water level line. The calculation is as follows:

[0082] ;

[0083] Establish a pairing relationship between two sets of area values, and then combine the measured areas. (From S113) and theoretical area (From the reservoir capacity curve) Construct a data structure according to row correspondence, generating a water surface area pairing matrix. The matrix form is as follows: The columns are [water level line number, measured area, theoretical area].

[0084] S212: Based on the water surface area pairing matrix, calculate the numerical difference between each pair of area values, and compare it item by item with the preset area comparison benchmark difference limit. Mark all water level line number segments whose numerical difference is greater than the preset area comparison benchmark difference limit to obtain the set of area anomaly numbers.

[0085] Based on the water surface area pairing matrix, difference analysis is performed on each row of data to accurately identify the specific water level intervals that cause distortion in the existing model, providing targeted basis for local or global reservoir capacity curve reconstruction. The numerical difference between each pair of area values ​​is calculated. and relative error ratio .

[0086] A preset limit for the difference in area comparison benchmarks was set. This limit was set based on the statistical characteristics of historical hydrological data. By fitting a normal distribution to the water surface area measurement errors during the dry and wet seasons of the past 5 years, it was found that the error fluctuation within the 95% confidence interval was within a certain range. Within [a certain range]. Considering the high-precision characteristics of UAV mapping, a relative error limit is set. As a rigid threshold for anomaly detection, a limit is also set for the absolute area difference. (Protective threshold for small areas).

[0087] The logic for item-by-item comparison is: if and If so, the data pair is determined to be abnormal;

[0088] For example, a water level line numbered L-145 corresponds to a measured area of ​​51,000 square kilometers. Theoretical area 48,000 ;

[0089] Calculate the difference: ;

[0090] Calculate the ratio: ;

[0091] because and The record is marked.

[0092] Table 1 shows some of the water surface area pairing and judgment results. Through this judgment, the system can automatically filter out distorted nodes that urgently need to be reconstructed from the reservoir capacity curve:

[0093] Table 1. Determination of Differences in Water Surface Area

[0094] Water level line number Measured area ( ) Theoretical area ( ) Difference ( ) Error rate (%) Judgment Result L-145 51000 48000 3000 6.25 abnormal L-146 52100 52000 100 0.19 normal L-147 45000 50000 5000 10.00 abnormal

[0095] Traverse the entire matrix and extract the water level numbers corresponding to all rows whose judgment results are "abnormal" to obtain the set of area abnormality numbers.

[0096] S213: Based on the set of area anomaly numbers, locate the corresponding water level line number in the water body boundary surface layer, extract the spatial contour data of the corresponding boundary area, assign area mismatch markers, aggregate all marked boundary areas, and output area mismatch boundary identification information.

[0097] Based on the set of area anomaly numbers, perform a reverse spatial query in the GIS environment. Using the water level line numbers (such as L-145, L-147) in the set as index keys, traverse the water body boundary layer dataset to locate the corresponding water level line number in the layer's patch number information.

[0098] Extract the spatial contour data of the corresponding boundary region. This includes the coordinate sequence of all vertices of the polygon. To facilitate subsequent processing, an "area mismatch marker" is assigned to each abnormal patch while extracting geometric information.

[0099] Aggregate all marked boundary areas. Merge scattered anomalous patch objects into a single geographic feature layer. During this process, record the centroid coordinates and boundary rectangle extent of each anomalous area to facilitate the directional retrieval of subsequent LiDAR and sonar data. The final output data structure includes metadata such as the geometry of the anomalous area, its original number, and area difference value, i.e., the output area mismatch boundary identification information, which serves as the direct input range for subsequent physical spatial scanning and reservoir capacity curve reconstruction calculations.

[0100] Please see Figure 4 Step S3 is as follows:

[0101] S311: Based on the water level line number area in the area mismatch boundary identification information, match the lidar echo data of the reservoir area under the corresponding number with the underwater elevation scatter data collected by the depth sounding device, extract the spatial location coordinates and elevation values ​​of the two types of data respectively, and perform a unified geographic projection parameter conversion to establish a unified projection elevation point set.

[0102] Based on the water level line numbering areas in the area mismatch boundary identification information, in order to achieve accurate reservoir capacity curve reconstruction, it is necessary to reconstruct the underwater and above-water topography from a three-dimensional physical perspective. Therefore, a multi-source data fusion process is initiated. For each marked mismatch area, data of the corresponding time period and spatial range are retrieved from the database.

[0103] Matching LiDAR echo data from the reservoir area: LAS format point cloud data acquired using an airborne LiDAR system, primarily covering the area above the water surface and along the bank slope. Spatial coordinates were extracted. .

[0104] Matching underwater elevation scatter data collected by a depth sounder: XYZ text data collected by a multibeam echo sounder, covering the area below the water surface, and its coordinates extracted. .

[0105] Since the two sets of raw data may be based on different coordinate frames (e.g., LiDAR uses the WGS84 ellipsoidal height, while the depth sounder uses the local theoretical lowest tide level), a unified geographic projection parameter transformation must be performed. The target coordinate system is set to the CGCS2000 projection coordinate system, and the elevation datum is unified to the 1985 National Elevation Datum.

[0106] For LiDAR data, a Bursa-Wolf model with 7 parameters (3 translations, 3 rotations, and 1 scale) is used for transformation:

[0107] ;

[0108] For depth sounding data, real-time water level elevation needs to be superimposed. water depth value Convert to absolute elevation .

[0109] After the conversion is completed, the two types of point clouds are merged into the same data container to establish a unified projection elevation point set.

[0110] S312: Based on the unified elevation point set of projection, extract the spatial coordinates of adjacent point pairs, calculate the horizontal distance, filter point pairs whose adjacent distance is less than the point spacing filtering threshold, and determine whether the difference between the corresponding elevation values ​​is less than the set value (0.5 meters). Aggregate the point pairs that meet the conditions to generate a set of continuous elevation point pairs.

[0111] Based on a unified projection elevation point set, this study aims to identify the transition zone between surface and underwater data. Because LiDAR signals may be absorbed by water at the land-water interface, and sonar signals have blind spots in extremely shallow water (<0.5m), a physical gap typically exists between the two types of data at the interface. Filling this gap is essential to ensuring the continuity of volume integrals during reservoir capacity curve reconstruction.

[0112] Extract the spatial coordinates of adjacent point pairs. Use the KD-Tree algorithm to construct a spatial index for the point cloud, accelerating the nearest neighbor search. Traverse the edge points in the LiDAR point set and find the nearest point in the sounding point set.

[0113] Calculate horizontal distance Filter point pairs whose adjacent distance is less than the point spacing filtering threshold. This threshold is set to twice the data acquisition density. For example, if the average point spacing is 1 meter, the threshold is set to 2.0 meters to ensure that only adjacent measurement points are connected.

[0114] It then determines whether the difference between corresponding elevation values ​​is less than a set value (0.5 meters). That is, it calculates... .

[0115] The 0.5-meter setting is based on a comprehensive error analysis of the reservoir wave run-up and the depth sounder's draft. Experiments show that, under calm water conditions, the effective elevation continuity error at the water-land interface should be controlled within 0.5 meters. Exceeding this value may indicate that the data belong to different terrain features (such as steep banks or cliffs) and should not be directly connected.

[0116] The conditions will be met ( and (point pairs) Perform aggregation, record the index pairs, and generate a set of continuous elevation point pairs.

[0117] S313: Based on the set of continuous elevation point pairs, interpolate intermediate points in space, calculate the elevation interpolation of intermediate points according to distance weight, add all interpolated points to the original elevation point set, merge all elevation point information, and establish a fused elevation data set for both above and below water.

[0118] Based on the set of continuous elevation point pairs, interpolation completion is performed to eliminate data holes, providing a seamless terrain surface model for reservoir capacity curve reconstruction. For each pair of points in the set... ,in From the water From underwater.

[0119] Interpolation in space generates intermediate points The plane coordinates of its intermediate point. The calculation is as follows:

[0120] , The elevation value of the midpoint is set to the linear average of the original elevation differences, that is: .

[0121] This step involves constructing a model of the transition zone where land and water meet. To increase smoothness, if... and For larger horizontal distances, multi-point interpolation (such as dividing the points into three equal parts) can be used. In this embodiment, single-point median interpolation is used.

[0122] All generated interpolation points Add to the original elevation point set (including original LiDAR points and sounding points). Perform deduplication to remove redundant points with completely overlapping spatial coordinates. Merge all elevation point information to form a complete dataset covering the above-water slope, the water-land transition zone, and underwater topography, establishing a fused above-water and underwater elevation data set.

[0123] Please see Figure 5 Step S4 is as follows:

[0124] S411: Based on the three-dimensional spatial coordinates and elevation values ​​in the fusion elevation data set of above-water and underwater, reconstruct all points into geographic coordinate data frames under a unified coordinate system according to the set coordinate datum WGS-84, and generate a projection-consistent coordinate dataset.

[0125] Based on the three-dimensional spatial coordinates and elevation values ​​in the above-water and underwater fusion elevation data set, although the projection has been unified before, in order to meet the needs of WebGL or specific three-dimensional visualization engines, it is necessary to reconfirm or convert to the globally universal WGS-84 coordinate system (EPSG: 4326).

[0126] Perform projection transformation according to the set coordinate datum WGS-84. Utilize the transformation pipeline provided by the PROJ library to project the plane coordinates... Convert to latitude and longitude Elevation value Keep it unchanged or correct the orthographic height according to a geoid model (such as EGM96).

[0127] All points are reconstructed into geographic coordinate data frames under a unified coordinate system. The data frame structure is designed as [ID, Longitude, Latitude, Elevation, Source_Flag]. The Source_Flag indicates the data source (0=LiDAR, 1=Sonar, 2=Interpolated).

[0128] A projection-consistent coordinate dataset is generated and stored in binary LAS or compressed CSV format to ensure read and write efficiency under large data volumes, laying a standardized foundation for the subsequent generation of digital elevation models for reservoir capacity curve reconstruction.

[0129] S412: Based on the projection-consistent coordinate dataset, a regular grid structure is constructed according to a set interval. The boundary nodes of each grid cell are extracted, and interpolation calculations are performed according to the elevation values ​​corresponding to the nodes. Points with the same elevation values ​​are connected to form a closed shape, resulting in a set of continuous equal-elevation closed shapes.

[0130] Based on a projection-consistent coordinate dataset, a digital elevation model (DEM) is constructed, which serves as the core volume carrier for reconstructing reservoir capacity curves.

[0131] Construct a regular grid structure with a set spacing. Set the grid resolution to 1 meter. 1 meter. Calculate the spatial bounding box of the dataset. Number of rows and columns in the grid .

[0132] The boundary nodes (i.e., the four corner points of the grid) of each raster cell are extracted. Interpolation calculations are performed based on the elevation values ​​corresponding to the nodes. Since the original point cloud is discretely distributed, the natural neighbor interpolation method is used to calculate the elevation values ​​of the grid nodes. This method is based on the geometric properties of the Voronoi diagram, which can effectively preserve terrain features and avoid artificial oscillations.

[0133] Connecting points with the same elevation value forms a closed shape. A moving square algorithm is used to extract contour lines, with a contour interval of 1 meter. The algorithm traverses each grid cell and, based on the relationship between the elevation values ​​of the four corner points and the target contour line value (higher or lower), determines the contour line division method within that grid by looking up a table.

[0134] For example, if the top left and top right corners are higher than the target value, and the bottom left and bottom right corners are lower than the target value, then the contour lines cross the grid. Connect all the segmented lines according to their topological relationships to obtain a continuous closed set of contour lines.

[0135] S413: Based on the elevation values ​​and spatial coordinate information of each boundary point in the continuous contour closed graphic set, the spatial distance between the raster deviation point and its neighboring points is measured and calculated. Elevation value interpolation correction is performed using a distance-weighted method, employing the following formula:

[0136] ;

[0137] The corrected elevation values ​​of boundary points are calculated, the corrected results are mapped back to the original raster cells, the contour line boundaries are updated, and a standard coordinate isoelastic grid layer is established; among these steps... The boundary point elevation correction value represents the raster deviation point. Indicates the first Elevation values ​​of neighboring points, Indicates the deviation point and the first Euclidean distance between neighboring points Indicates the distance-weighted index. This represents the total number of neighboring points;

[0138] Based on the elevation values ​​and spatial coordinate information of each boundary point in the continuous contour closed graphic set, in order to correct the smoothing distortion that may occur during the interpolation process and to ensure that the elevation-area relationship on which the reservoir capacity curve is reconstructed has high-precision physical authenticity, the contour node is calibrated a second time using the original high-precision point cloud.

[0139] The spatial distance between raster offset points (i.e., nodes on contour lines) and neighboring points (points in the original point cloud) is measured and calculated. The search radius is set. Meters, search for the original point within that radius as the "nearest point".

[0140] Elevation value interpolation correction is performed using a distance-weighted method, employing the following formula:

[0141] ;

[0142] in:

[0143] : The corrected elevation value of the raster deviation point to be corrected.

[0144] : No. The original elevation values ​​of the neighboring points.

[0145] Deviation point With the The Euclidean distance between the nearest neighboring points.

[0146] : Distance-weighted index. This parameter controls the degree to which distance affects the weighting. Experiments show that when When the value is too low (e.g., 1), terrain details are blurred; when the value is too high (e.g., 3), it is greatly affected by noise. Through cross-validation, this embodiment sets... That is, inverse distance squared weighting is used to balance smoothness and local features.

[0147] The total number of neighboring points.

[0148] Assuming the point to be corrected Three original points were found in the surrounding area. The data is as follows:

[0149] Point 1: ;

[0150] Point 2: ;

[0151] Point 3: ;

[0152] set up .

[0153] Calculate the weighted denominator :

[0154] Point 1: ;

[0155] Point 2: ;

[0156] Point 3: ;

[0157] Sum of denominators: .

[0158] Calculate the numerator :

[0159] Point 1: ;

[0160] Point 2: ;

[0161] Point 3: ;

[0162] Sum of numerators: .

[0163] Calculate the correction value :

[0164] ;

[0165] Table 2 shows the calculation differences under different distance weights, verifying the results. Reasonableness (result closest to the local mode):

[0166] Table 2. Experimental data comparing distance-weighted indices.

[0167] experimental group Value Calculation results ( ) Deviation analysis A 1 150.50 Slightly off to the farthest point B (This example) 2 150.50 Balanced, preserving near-point features C 3 150.50 Strong bias towards the near point

[0168] The corrected elevation values ​​of boundary points are calculated and mapped back to the original raster cells to update the contour line boundaries. This result demonstrates that introducing distance-weighted correction from the original point cloud effectively eliminates smoothing errors during rasterization, making the contour lines more closely resemble the actual terrain. Finally, a standard coordinate iso-elevation grid layer is established, thus completing the preparation of high-precision terrain data for reservoir capacity curve reconstruction.

[0169] Please see Figure 6 The S5 steps are as follows:

[0170] S511: Based on the spatial coordinates and corresponding elevation values ​​of each grid cell in the standard coordinate isoelastic grid layer, combined with the set water level line numbering and partitioning information, extract the boundary point positions of all grid cells under each water level line number, and connect the boundary points in sequence according to the two-dimensional grid spatial order to construct an initial point path set and generate a two-dimensional spatial point trajectory line sequence.

[0171] Based on the spatial coordinates and corresponding elevation values ​​of each grid cell in the standard coordinate iso-elevation grid layer, combined with the set water level line numbering zoning information (for example, dividing water level intervals by 0.5-meter intervals).

[0172] Extract the boundary point positions of all raster cells under each water level line number. Specifically, for water level... Filter out all grid correction boundary points with an elevation value equal to 150m.

[0173] Based on the spatial order of the two-dimensional raster, the boundary points are sequentially connected to construct an initial set of point paths. Due to the ordered nature of the raster data (row and column indexing), the Moore's Neighborhood Tracking algorithm is used to search for adjacent contour points in a clockwise direction, starting from the first matching point in the upper left corner, and recording the path index sequence. Discrete point sequences are organized into line objects to generate a two-dimensional spatial point matrix trajectory line sequence. This sequence actually depicts the corrected water level-area relationship geometry and is the geometric expression of the "area" parameter in the reservoir capacity curve reconstruction.

[0174] S512: Based on the two-dimensional spatial matrix trajectory line sequence, the numbering is sorted in ascending order according to the water level line numbering, the closed correspondence between adjacent trajectory lines is identified, the sequential aggregation operation is performed to integrate each segment trajectory, the complete trajectory boundary structure under all numbered segments is obtained, and a two-dimensional boundary trajectory set is established.

[0175] Based on the two-dimensional spatial dot matrix trajectory line sequence, the water level lines are numbered in ascending order (e.g., 145m, 145.5m, 146m...).

[0176] Identify the closed correspondence between adjacent trajectory lines. In actual extraction, due to complex terrain, contour lines at the same elevation may have breaks or multiple islands. A topology checking algorithm is used to determine the Euclidean distance between the endpoints of the line segments. If the distance between the endpoints of two trajectory segments is less than a set threshold (0.2 meters), they are considered as different segments of the same contour line.

[0177] Perform sequential aggregation to integrate the segmented trajectories. Line segments identified as belonging to the same group are geometrically joined together. For any small gaps that are not yet closed, force a closure operation to ensure that each contour line is a geometrically closed loop structure.

[0178] Obtain the complete trajectory boundary structure for all numbered segments and establish a two-dimensional boundary trajectory set. Each element in this set represents the theoretical water surface boundary at a specific water level. The area enclosed by these boundaries will be directly used to recalculate the reservoir capacity volume at each water level, thereby completing the numerical iterative update in the reservoir capacity curve reconstruction.

[0179] S513: Based on all closed boundary structures in the two-dimensional boundary trajectory set, extract the spatial position sequence and sorting number of each trajectory node, reconstruct and map adjacent boundary sequences and establish a linear connection index, combine all closed trajectories, transform them into two-dimensional graphic representation entities according to the trajectory structure, and output the reconstructed graphic representation result of the storage capacity curve.

[0180] Based on all closed boundary structures in the two-dimensional boundary trajectory set, extract the spatial position sequence and sorting number of each trajectory node.

[0181] Reconstruct and map adjacent boundary sequences and establish linear connectivity indexes. To achieve the transformation from two-dimensional to three-dimensional entities (or quasi-three-dimensional graphics), triangular mesh connections or lofting relationships need to be established between contour lines of adjacent elevations.

[0182] In this embodiment, as a visualization and verification step for the reconstruction of the reservoir capacity curve, a lofting logic is used to generate the reservoir capacity volume graphic. The corresponding nodes between two adjacent contour lines (such as 150m and 150.5m) are calculated, and a connection index is established according to the shortest distance principle.

[0183] All closed trajectories are combined and transformed into two-dimensional graphic representations of entities according to their trajectory structures. Using computer-aided design (CAD) libraries (such as ezdxf) or GIS drawing engines, the repaired contour line sequence is drawn as vector graphics. Each contour line layer is treated as a layer or block, with accompanying elevation attributes.

[0184] The output is a reconstructed reservoir capacity curve graphical representation. This result is not merely a curve chart, but a contour map with precise spatial coordinates, visually displaying the water surface morphology at different water levels and corresponding to the corrected water level-area-volume relationship. Using this graphical representation, technicians can extract the water surface area corresponding to any water level and calculate the cumulative reservoir capacity using the platform volume formula, thereby generating a completely new reservoir capacity curve table reflecting the current actual topography, achieving a complete reservoir capacity curve reconstruction. Furthermore, after anomaly removal and elevation correction, its accuracy surpasses that of traditional interpolation results.

[0185] Experimental data show that the reconstructed graphics generated by this method and the reconstructed reservoir capacity curves based on it, when compared with the actual measured cross sections, have an average elevation error controlled within 8 cm. The reservoir capacity calculation error is reduced by 12.5% ​​compared with the traditional trapezoidal platform method, effectively solving the problem of distorted water level-reservoir capacity relationship caused by mismatch between water and land boundary data.

[0186] A system for reconstructing storage capacity curves based on multi-source data fusion includes:

[0187] The remote sensing boundary extraction module is used to perform S1: acquire UAV remote sensing images and ground control point layout layers, identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets.

[0188] The area difference analysis module is used to perform S2: extract the area value from the water body boundary surface layer dataset, match the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, and output the area mismatch boundary identification information.

[0189] The elevation data fusion module is used to execute S3: based on the water level line number area in the area mismatch boundary identification information, it matches the reservoir's surface elevation point cloud sequence and underwater elevation scatter data, performs unified geographic projection transformation, extracts the distance between adjacent elevation points for filtering, and constructs a set of surface and underwater fused elevation data.

[0190] The iso-mesh standardization module is used to perform S4: extract the combined elevation data set above and below water, construct an equally spaced regular grid, generate contour line closed graphics for grid cell boundary points, perform interpolation correction on deviation points, and output a standard coordinate iso-elevation grid layer;

[0191] The curve reconstruction modeling module is used to execute S5: based on the standard coordinate isoelastic grid layer, and according to the correspondence between the two-dimensional grid and the water level, it establishes a sequence of spatial node trajectory lines, integrates them in order of water level, and outputs the result of the reservoir capacity curve reconstruction graphic representation.

[0192] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for reconstructing reservoir capacity curves based on multi-source data fusion, characterized in that, Includes the following steps: S1: Acquire UAV remote sensing images and ground control point layout layers, identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets. S2: Extract the area value from the water body boundary surface layer dataset, match it with the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, and output the area mismatch boundary identification information. S3: Based on the water level line number area in the area mismatch boundary identification information, match the reservoir surface elevation point cloud sequence and underwater elevation scatter data, perform unified geographic projection transformation, extract the distance between adjacent elevation points for filtering, and construct a set of surface and underwater fused elevation data. S4: Extract the above-water and underwater fused elevation data set, construct an equally spaced regular grid, generate contour line closed graphics for grid cell boundary points, perform interpolation correction on deviation points, and output a standard coordinate isoelastic grid layer. S5: Based on the standard coordinate isoelastic grid layer, and according to the correspondence between the two-dimensional grid and the water level, establish a spatial node trajectory line sequence, integrate it according to the water level order, and output the reconstructed graph representation of the reservoir capacity curve. The specific steps for obtaining the area mismatch boundary identification information are as follows: S211: Obtain the water body boundary surface layer dataset, extract the surface coverage area value of each planar patch in the layer, and index the patch number to correspond with the water level line number in the original reservoir capacity curve data. Retrieve the water surface area value recorded by the corresponding water level line in the original reservoir capacity curve, establish a pairing relationship between the two sets of area values, and generate a water surface area pairing matrix. S212: Based on the water surface area pairing matrix, calculate the numerical difference between each pair of area values, and compare it item by item with the preset area comparison benchmark difference limit. Mark all water level line number segments whose numerical difference is greater than the preset area comparison benchmark difference limit to obtain the area anomaly number set. S213: Based on the set of area anomaly numbers, locate the corresponding water level line number in the water body boundary surface layer, extract the spatial contour data of the corresponding boundary area, assign area mismatch markers, aggregate all marked boundary areas, and output area mismatch boundary identification information. The specific steps for obtaining the above-water and underwater integrated elevation data set are as follows: S311: Based on the water level line number area in the area mismatch boundary identification information, match the reservoir area lidar echo data and underwater elevation scatter data collected by the depth sounding device under the corresponding number, extract the spatial location coordinates and elevation values ​​of the two types of data respectively, and perform a unified geographic projection parameter conversion to establish a unified projection elevation point set. S312: Based on the unified elevation point set of the projection, extract the spatial coordinates of adjacent point pairs, calculate the horizontal distance, filter point pairs whose adjacent distance is less than the point spacing filtering threshold, and determine whether the difference between the corresponding elevation values ​​is less than a set value. Aggregate the point pairs that meet the conditions to generate a set of continuous elevation point pairs. S313: Based on the set of continuous elevation points, interpolate to generate intermediate points in space, calculate the elevation interpolation of the intermediate points according to the distance weight, add all interpolated points to the original elevation point set, merge all elevation point information, and establish a fused elevation data set for both above and below water.

2. The method for reconstructing reservoir capacity curves based on multi-source data fusion according to claim 1, characterized in that: The water body boundary surface layer dataset includes boundary closed polygons and water body area information. The area mismatch boundary identification information includes difference index values, water level line numbers, and area comparison status. The water surface and underwater integrated elevation data set includes unified coordinate points, elevation difference indexes, and filler point density information. The standard coordinate isopleth elevation grid layer includes raster boundary contour lines, corrected interpolation points, and elevation grid cells. The reservoir capacity curve reconstruction graphic representation result includes two-dimensional boundary trajectories, numbered water level line positions, and spatial point matrix trajectory line sequences.

3. The method for reconstructing reservoir capacity curves based on multi-source data fusion according to claim 1, characterized in that, The specific steps for obtaining the water body boundary surface layer dataset are as follows: S111: Acquire UAV remote sensing imagery and ground control point layout layers, read all pixel grayscale values ​​in the image histogram, perform difference calculations for grayscale value changes, search for areas where the grayscale difference between adjacent pixels is greater than the preset grayscale change threshold, record the corresponding spatial coordinates, and generate a set of grayscale change boundary pixels. S112: Based on the set of gray-scale abrupt boundary pixels, extract the adjacent coordinate index information of each boundary pixel, determine whether the adjacent indices are continuous and form a closed boundary path, and if the connection condition is met, then sequentially splice the boundary points to construct a spatial closed structure and establish a set of boundary contour polygon regions. S113: Based on the set of polygon regions with boundary contours, perform statistical operations on the number of pixels in each polygon region, and multiply it with the actual ground area parameter of a single pixel to generate the surface cover area information of each region and establish a water body boundary surface layer dataset.

4. The method for reconstructing reservoir capacity curves based on multi-source data fusion according to claim 1, characterized in that, The specific steps for obtaining the standard coordinate isoelastic grid layer are as follows: S411: Based on the three-dimensional spatial coordinates and elevation values ​​in the above-water and underwater fusion elevation data set, reconstruct all points into geographic coordinate data frames under a unified coordinate system to generate a projection-consistent coordinate dataset. S412: Based on the projection-consistent coordinate dataset, construct a regular grid structure at a set interval, extract the boundary nodes of each grid cell, perform interpolation calculations according to the elevation values ​​corresponding to the nodes, and connect points with the same elevation values ​​to form a closed shape, thereby obtaining a set of continuous equal-elevation closed shapes. S413: Based on the elevation values ​​and spatial coordinate information of each boundary point in the continuous contour closed graphic set, the spatial distance between the grid deviation point and the neighboring point is measured and calculated. Elevation value interpolation correction is performed in a distance-weighted manner. The corrected elevation value of the boundary point is obtained by calculation. The correction result is mapped to the original grid cell, the contour line boundary is updated, and a standard coordinate isoelastic grid layer is established.

5. The method for reconstructing reservoir capacity curves based on multi-source data fusion according to claim 4, characterized in that, The formula for calculating the corrected elevation value of the boundary point is as follows: ； in, The boundary point elevation correction value represents the raster deviation point. Indicates the first Elevation values ​​of neighboring points, Indicates the deviation point and the first Euclidean distance between neighboring points Indicates the distance-weighted index. This represents the total number of neighboring points.

6. The method for reconstructing reservoir capacity curves based on multi-source data fusion according to claim 1, characterized in that, The specific steps for obtaining the graphical representation result of the reconstructed storage capacity curve are as follows: S511: Based on the spatial position coordinates and corresponding elevation values ​​of each grid cell in the standard coordinate isoelastic grid layer, and combined with the set water level line numbering partition information, extract the boundary point positions of all grid cells under each water level line number, and connect the boundary points in sequence according to the two-dimensional grid spatial order to construct an initial point path set and generate a two-dimensional spatial point path line sequence. S512: Based on the two-dimensional spatial dot matrix trajectory line sequence, sort the numbers in ascending order according to the water level line numbering, identify the closed correspondence between adjacent trajectory lines, perform sequential aggregation operation to integrate each segment trajectory, obtain the complete trajectory boundary structure under all numbered segments, and establish a two-dimensional boundary trajectory set. S513: Based on all closed boundary structures in the set of two-dimensional boundary trajectories, extract the spatial position sequence and sorting number of each trajectory node, reconstruct and map adjacent boundary sequences and establish a linear connection index, combine all closed trajectories, transform them into two-dimensional graphic representation entities according to the trajectory structure, and output the reconstructed graphic representation result of the reservoir capacity curve.

7. A system for reconstructing reservoir capacity curves based on multi-source data fusion, characterized in that, The system is used to implement the reservoir capacity curve reconstruction method based on multi-source data fusion as described in any one of claims 1-6, including: The remote sensing boundary extraction module is used to perform S1: acquire UAV remote sensing images and ground control point layout layers, identify boundary pixel regions, perform automatic vectorization of water body edge boundary lines, establish closed polygon regions of water body boundary outlines and perform area calculations, and generate water body boundary surface layer datasets. The area difference analysis module is used to perform S2: extract the area value from the water body boundary surface layer dataset, match the corresponding water surface area value below the water level recorded in the original reservoir capacity curve, calculate the area value difference to obtain the difference index, identify the water level line number area where the difference index exceeds the preset area comparison benchmark difference limit, and output the area mismatch boundary identification information. The elevation data fusion module is used to perform S3: based on the water level line number area in the area mismatch boundary identification information, it matches the reservoir surface elevation point cloud sequence and underwater elevation scatter data, performs unified geographic projection transformation, extracts the distance between adjacent elevation points for filtering, and constructs a set of surface and underwater fused elevation data. The iso-grid standardization module is used to perform S4: extract the above-water and underwater integrated elevation data set, construct an equally spaced regular grid, generate contour line closed graphics for grid cell boundary points, perform interpolation correction on deviation points, and output a standard coordinate iso-elevation grid layer; The curve reconstruction modeling module is used to execute S5: based on the standard coordinate isoelastic grid layer, according to the correspondence between the two-dimensional grid and the water level, establish a sequence of spatial node trajectory lines, integrate them in order of water level, and output the result of the reservoir capacity curve reconstruction graphic representation.

Images