An optimization method for coastal near-shore water depth inversion based on fusion of multi-source remote sensing data and nautical charts

Abstract

The present application relates to the coastal nearshore water depth inversion technology, disclose a kind of based on the fusion of multiple remote sensing data and chart coastal nearshore water depth inversion optimization method.Acquisition satellite laser depth sounding data, multispectral remote sensing image, water transparency remote sensing data, electronic chart water depth data and coastal bottom data, form uniform grid data;Determine initial optical inversion water depth, laser reference water depth and chart constrained water depth;According to transparency level, chart consistency index, laser reliability index and coastal bottom data determine fusion weight, obtain fusion water depth result.The present application can improve the reliability of determining weight per grid, improve the local continuity and reliability of fusion water depth result.

Description

Technical Field

[0001] This invention relates to the field of coastal zone remote sensing monitoring technology, and in particular to a method for optimizing the inversion of nearshore water depth based on the fusion of multi-source remote sensing data and nautical charts. Background Technology

[0002] Coastal nearshore water depth data is fundamental for coastal engineering design, channel maintenance, island and reef topographic surveys, and nearshore marine management. While shipborne bathymetry can obtain highly accurate depth points or survey lines, in areas such as shoals, reefs, tidal channels, and nearshore fracture zones, field measurements are easily limited by water depth, tide levels, and navigation conditions. Multispectral remote sensing imagery, with its wide coverage, has been used for shallow water depth inversion. Chinese invention patent application CN105651263A describes a multi-source remote sensing fusion inversion method for shallow water depth. This method preprocesses multi-source remote sensing data and performs depth segment identification and fusion inversion based on single-source depth inversion results. This document demonstrates that multi-source remote sensing data fusion has become a feasible technical approach for nearshore shallow water depth mapping.

[0003] In marine areas lacking densely populated depth control points, spaceborne laser bathymetry data can provide reference depths for multispectral remote sensing depth inversion. Chinese invention patent application CN121437583A describes a depth inversion method based on SDW-LPT multi-temporal fusion remote sensing imagery. This method projects ICESat-2 photon points onto the Sentinel-2 image coordinate system and obtains a standardized reference depth through noise photon removal, refraction correction, and tidal level correction. Then, it uses multi-temporal remote sensing imagery for depth inversion. Chinese invention patent application CN120521568A describes a method for ICESat-2 turbid water bathymetry correction and depth inversion based on water optical properties. This method combines water optical properties to compensate for ICESat-2 turbid water bathymetry and combines it with remote sensing imagery for depth inversion. The above literature indicates that spaceborne laser bathymetry data, multispectral remote sensing imagery, and water optical information can jointly participate in nearshore shallow water depth inversion.

[0004] Electronic nautical chart depth data and coastal sediment data can also be used for shallow water depth remote sensing inversion. Chinese invention patent application CN116295285A describes a region-adaptive shallow water depth remote sensing inversion method. This method collects remote sensing imagery, measured depth data, nautical chart data, tidal data, and sediment type data, and delineates the region and optimizes the model based on depth range, sediment type, and water quality. While this method can improve the usability of depth inversion results by utilizing data from different sources, in nearshore areas where water transparency varies significantly, the fusion weights between optically inverted depth, spaceborne laser reference depth, and electronically chart-constrained depth are still difficult to reliably determine grid-by-grid, easily affecting the continuity and local reliability of the fused depth results. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the inversion of nearshore water depth based on the fusion of multi-source remote sensing data and nautical charts. This method can solve the problem that the fusion weights of optical inversion water depth, spaceborne laser reference water depth, and electronic nautical chart constrained water depth are difficult to determine reliably on a grid-by-grid basis, thereby improving the local continuity and reliability of the fused water depth results.

[0006] To achieve the above objectives, the present invention provides the following solution: A method for optimizing nearshore water depth retrieval based on the fusion of multi-source remote sensing data and nautical charts includes: Acquire spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data of the target coastal nearshore area. Then, perform spatial benchmark unification, water depth benchmark unification, and grid cell matching on the spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data to obtain unified grid data. The initial optical inversion water depth of each grid cell is determined based on multispectral remote sensing images in the unified grid data; the laser reference water depth of each grid cell is determined based on spaceborne laser bathymetry data in the unified grid data; and the chart-constrained water depth of each grid cell is determined based on electronic chart water depth data in the unified grid data. The transparency level of each grid cell is determined based on the water transparency remote sensing data in the unified grid data, and the initial optical inversion water depth is attenuated and corrected according to the transparency level to obtain the corrected optical inversion water depth of each grid cell. The sea level of each grid cell is determined based on the difference between the corrected optical inversion depth and the chart-constrained depth. Figure 1 Consistency indicators, and determine the laser reliability index of each grid cell based on the distribution density of spaceborne laser depth sounding data in the neighborhood of each grid cell; According to transparency level, sea Figure 1 Consistency index, laser reliability index, and coastal seabed data in unified grid data are used to determine the fusion weights of corrected optical inversion depth, laser reference depth, and chart-constrained depth in each grid cell; According to the fusion weight, the corrected optical inversion water depth, laser reference water depth and chart-constrained water depth in each grid cell are fused to obtain the fused water depth result of the nearshore area of ​​the target coastal zone.

[0007] Preferably, spatial benchmark unification, water depth benchmark unification, and grid cell matching are performed on spaceborne laser bathymetry data, multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart water depth data, and coastal sediment data to obtain unified grid data, including: Grid units are divided based on the pixel coverage of multispectral remote sensing images; The spaceborne laser bathymetry data, water transparency remote sensing data, electronic nautical chart water depth data, and coastal sediment data were matched to their respective grid cells. Convert spaceborne laser bathymetry data and electronic nautical chart depth data to the same depth datum; Within the same grid cell, establish the correspondence between multispectral remote sensing images, water transparency remote sensing data, spaceborne laser bathymetry data, electronic nautical chart water depth data, and coastal sediment data to obtain unified grid data.

[0008] Preferably, determining the initial optical inversion water depth of each grid cell based on multispectral remote sensing imagery in unified grid data includes: Extracting water spectral features from multispectral remote sensing images of each grid cell from unified grid data; Using at least one of the spaceborne laser bathymetry data and electronic nautical chart water depth data from the unified grid data as water depth control data, an inversion relationship between water body spectral characteristics and water depth is established. The initial optical inversion depth of each grid cell is determined based on the inversion relationship.

[0009] Preferably, determining the laser reference water depth corresponding to each grid cell based on spaceborne laser bathymetry data in the unified grid data includes: Determine the spaceborne laser depth sounding points within the neighborhood of each grid cell; The distance weight of the spaceborne laser sounding point is determined based on the distance between the spaceborne laser sounding point and the corresponding grid cell. The water depth values ​​of the spaceborne laser depth sounding points are weighted according to distance weights to obtain the laser reference water depth corresponding to each grid cell.

[0010] Preferably, the transparency level of each grid cell is determined based on the water transparency remote sensing data in the unified grid data, and the initial optically retrieved water depth is attenuated and corrected according to the transparency level to obtain the corrected optically retrieved water depth of each grid cell, including: The water transparency value of each grid cell is determined based on remote sensing data of water transparency. The transparency level of each grid cell is determined based on the transparency range to which the water transparency value belongs; The attenuation correction coefficient for each grid cell is determined based on the transparency level. The lower the water transparency represented by the transparency level, the greater the correction magnitude of the attenuation correction coefficient for the initial optical inversion water depth. The initial optical inversion depth is corrected using the attenuation correction coefficient, and the corrected optical inversion depth of each grid cell is obtained.

[0011] Preferably, the sea depth of each grid cell is determined based on the difference between the corrected optical inversion depth and the chart-constrained depth. Figure 1 Consistency indicators, including: Calculate the depth difference between the corrected optical inversion depth and the chart-constrained depth in the same grid cell; The degree of water depth deviation between the corrected optical inversion water depth and the chart-constrained water depth in the same grid cell is determined based on the water depth difference. The sea level of each grid cell is determined based on the degree of water depth deviation. Figure 1 Consistency indicators, and the greater the deviation in water depth, the more important it is for the sea. Figure 1 The lower the consistency of the consistency indicators.

[0012] Preferably, the laser reliability index of each grid cell is determined based on the distribution density of spaceborne laser depth sounding data in the neighborhood of each grid cell, including: Count the number of spaceborne laser sounding points in the neighborhood of each grid cell; The distribution density is determined based on the number of satellite-borne laser sounding points and the neighborhood area of ​​each grid cell. The laser reliability index of each grid cell is determined based on the distribution density, and the higher the distribution density, the higher the reliability represented by the laser reliability index.

[0013] Preferably, based on transparency level, sea Figure 1 Using consistency indices, laser reliability indices, and coastal sediment data from the unified grid data, the fusion weights for corrected optical inversion depth, laser reference depth, and chart-constrained depth in each grid cell are determined, including: The weight of correcting optical inversion water depth is determined based on the transparency level, and the lower the water transparency represented by the transparency level, the smaller the weight of correcting optical inversion water depth. The weight of the laser reference water depth is determined based on the laser reliability index, and the higher the reliability represented by the laser reliability index, the greater the weight of the laser reference water depth. According to the sea Figure 1 The consistency index adjusts the weights of optically inverted depth and chart-constrained depth; in the sea Figure 1 When the consistency of the consistency index is reduced and the laser reference water depth has an effective water depth value, reduce the weight of the water depth that deviates significantly from the laser reference water depth in the corrected optical inversion water depth and the chart-constrained water depth. The weights for optically inverted water depths are corrected based on coastal sediment data from the unified grid data. The weights of the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth are normalized to obtain the fusion weights.

[0014] Preferably, the weights for optically inverted water depth are corrected based on coastal zone sediment data in the unified grid data, including: The substrate type of each grid cell was determined based on the coastal zone substrate data in the unified grid data; The degree of influence of the substrate on the reflection of multispectral remote sensing images in each grid cell is determined based on the substrate type; The weight of the optical inversion depth is adjusted according to the degree of reflection influence, and the greater the degree of reflection influence, the smaller the weight of the optical inversion depth.

[0015] Preferably, the corrected optical inversion depth, laser reference depth, and chart-constrained depth in each grid cell are fused according to the fusion weights to obtain the fused depth result for the nearshore area of ​​the target coastal zone, including: In each grid cell, determine whether the corrected optical inversion depth, laser reference depth, and chart-constrained depth have valid depth values; The weights corresponding to water depths that do not have valid water depth values ​​are reset to zero, and the weights corresponding to water depths that have valid water depth values ​​are normalized. Based on the normalized weights, the corrected optical inversion water depth, laser reference water depth, and chart-constrained water depth with effective water depth values ​​are weighted and fused to obtain the grid fused water depth of each grid cell. The merged water depth of the target coastal zone nearshore area is determined based on the merged water depth of each grid cell.

[0016] The present invention discloses the following beneficial effects: This invention unifies the spatial reference, water depth reference, and grid cell matching of spaceborne laser bathymetry data, multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart water depth data, and coastal sediment data. This ensures that data from different sources have corresponding relationships within the same grid cell, preventing water depth values, transparency levels, sediment data, and chart-constrained water depths from belonging to different spatial locations and affecting subsequent fusion weight determination. An initial optical inversion water depth is generated from multispectral remote sensing imagery, and attenuation correction is applied after determining the transparency level based on water transparency remote sensing data. This allows the optical inversion water depth to be corrected as water transparency changes, reducing the direct impact of optical inversion biases in low-transparency water bodies on the fusion result. The fusion weight is determined by correcting the difference between the optical inversion water depth and the chart-constrained water depth. Figure 1 The consistency index, combined with the distribution density of spaceborne laser bathymetry data in the neighborhood of the grid cell, determines the laser reliability index. This ensures that the fusion weight of each grid cell is no longer determined solely by a fixed data source, but is simultaneously influenced by transparency level, sea surface temperature, and other factors. Figure 1Consistency, laser bathymetry support, and coastal sediment data constraints; therefore, this invention addresses the problem in the prior art where the fusion weights between optically inverted water depth, spaceborne laser reference water depth, and electronic chart-constrained water depth are difficult to reliably determine grid-by-grid. It obtains fusion weights that match the reliability of each grid cell's data and improves the local continuity and reliability of the fused water depth results without increasing the complexity of the model structure. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method provided in an embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of the formation of unified grid data provided in an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram illustrating the determination of fusion weights in an embodiment of the present invention.

[0021] Figure 4 This is a typical grid cell fusion weight variation diagram provided in an embodiment of the present invention.

[0022] Figure 5 This is a schematic diagram of the coastal nearshore water depth inversion optimization system provided in an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] Figure 1 The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown in the figure, this embodiment provides a method for optimizing the inversion of nearshore water depth in the coastal zone based on the fusion of multi-source remote sensing data and nautical charts, including: Step 100: Acquire spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data of the target coastal nearshore area. Perform spatial benchmark unification, water depth benchmark unification, and grid cell matching on the spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data to obtain unified grid data. Step 200: Determine the initial optical inversion water depth of each grid cell based on the multispectral remote sensing image in the unified grid data, determine the laser reference water depth corresponding to each grid cell based on the spaceborne laser bathymetry data in the unified grid data, and determine the chart-constrained water depth of each grid cell based on the electronic chart water depth data in the unified grid data. Step 300: Determine the transparency level of each grid cell based on the water transparency remote sensing data in the unified grid data, and perform attenuation correction on the initial optical inversion water depth according to the transparency level to obtain the corrected optical inversion water depth of each grid cell. Step 400: Determine the sea depth of each grid cell based on the difference between the corrected optical inversion depth and the chart-constrained depth. Figure 1 Consistency indicators, and determine the laser reliability index of each grid cell based on the distribution density of spaceborne laser depth sounding data in the neighborhood of each grid cell; Step 500: Based on transparency level, sea Figure 1 Consistency index, laser reliability index, and coastal seabed data in unified grid data are used to determine the fusion weights of corrected optical inversion depth, laser reference depth, and chart-constrained depth in each grid cell; Step 600: According to the fusion weight, the corrected optical inversion water depth, laser reference water depth and chart constrained water depth in each grid cell are fused to obtain the fused water depth result of the nearshore area of ​​the target coastal zone.

[0026] Specifically, in step 100, the spatial extent of the target coastal nearshore area is first determined, and the water coverage area of ​​the target coastal nearshore area is used as the subsequent grid cell division range. Multispectral remote sensing images are separated into land and water pixels, and the spatial extent corresponding to these water pixels is used to divide the grid cells. Spaceborne laser bathymetry data is matched to the corresponding grid cell or its neighborhood according to the planar position of the bathymetry points. Water transparency remote sensing data is matched to the corresponding grid cell according to the pixel position. Electronic nautical chart depth data is matched to the corresponding grid cell according to the spatial position of depth points, isobaths, or gridded nautical chart depth values. Coastal sediment data is matched to the corresponding grid cell according to the distribution range of sediment type. When spaceborne laser bathymetry data and electronic nautical chart depth data have different depth references, the spaceborne laser bathymetry data and electronic nautical chart depth data are converted to the same depth reference; when multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data and coastal sediment data have different spatial references, the above data are converted to the same spatial reference.

[0027] After completing the unification of spatial reference, water depth reference, and grid cell matching, each grid cell records at least the grid cell location, the water spectral information corresponding to the multispectral remote sensing image, the water transparency value corresponding to the water transparency remote sensing data, the nautical chart water depth value corresponding to the electronic nautical chart water depth data, and the seabed type corresponding to the coastal zone seabed data. When there is a spaceborne laser bathymetry point in the grid cell or its neighborhood, the location and water depth value of the spaceborne laser bathymetry point are also recorded, thus obtaining unified grid data.

[0028] Figure 2 The formation relationship of the unified grid data in this embodiment is illustrated. See also... Figure 2 The left side represents the input data section, including spaceborne laser bathymetry data, multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data. The middle section represents the unified data processing section, which includes spatial benchmark unification, depth benchmark unification, and grid cell matching. The spaceborne laser bathymetry data, multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data are processed through spatial benchmark unification, depth benchmark unification, and grid cell matching to form unified grid data. The right side shows a schematic diagram of the unified grid data. The unified grid data is organized in a grid format. A target grid cell is shown in the figure, marked with a red box. The black dots within the red box represent spaceborne laser bathymetry points matched to the target grid cell or its neighboring area. The target grid cell records the grid cell location, water spectral information, water transparency value, nautical chart depth value, sediment type, and the location and depth value of the laser bathymetry point. Figure 2This is used to illustrate that after the five types of input data are processed in a unified manner, they enter the same grid cell and establish a corresponding relationship within the same grid cell, forming the unified grid data required for subsequent water depth inversion and fusion calculation.

[0029] In this embodiment, a control grid cell set is divided from the unified grid data. and verification grid cell set Control grid cell set Used to calculate inversion parameters, transparency correction parameters, and sea surface parameters. Figure 1 Consistency scale parameters, laser reliability scale parameters, and substrate correction parameters. Verification of the grid cell set. It is not involved in parameter calculations, but is used to evaluate the error changes of water depth results at different stages.

[0030] Control grid cell set This includes grid cells with valid spaceborne laser bathymetry data, and grid cells with valid electronic chart depth data after spatial location matching. Valid spaceborne laser bathymetry data refers to spaceborne laser bathymetry points that have been matched to the corresponding grid cell or within the neighborhood of the corresponding grid cell and have depth values. Valid electronic chart depth data refers to electronic chart depth points, contour depth values, or gridded chart depth values ​​that can be matched to the corresponding grid cell or within the neighborhood of the corresponding grid cell and have depth values.

[0031] For control grid units The water depth control data used for parameter calculation is denoted as Control grid unit When valid spaceborne laser depth sounding data is available, Retrieve the water depth value corresponding to the valid spaceborne laser bathymetry data. Control grid unit. When valid spaceborne laser bathymetry data is unavailable but valid electronic chart depth data is available, Take the water depth values ​​corresponding to the valid electronic nautical chart water depth data. After the water depth control data is determined in the above order, the subsequent inversion parameters and fusion weights are obtained from the sample statistics in the unified grid data.

[0032] In step 200, the grid cells are... The reflectance in the blue light band, green light band, and red light band are denoted as: Set positive numbers , The minimum effective resolution, less than the water reflectivity, is used to keep logarithmic operations feasible. Grid cell Water spectral feature vector Represented as: Set up control grid cells The matrix is ​​composed of the water spectral eigenvectors of all control grid cells. The corresponding water depth control data are combined into a vector. Initial optical inversion depth Determined according to the following formula: In the formula, , , and These are the inversion parameters. Calculate according to the following formula: In the formula, for The transpose of the matrix, It is the identity matrix. is a non-negative positive coefficient. When reversible, Set to 0; when the number of control grid cells is small or the correlation between spectral features is strong, Take values ​​greater than 0. Inversion parameters. Set of control grid cells After calculation, the water spectral characteristics of each grid cell are substituted into the inversion relation to obtain the initial optical inversion water depth. .

[0033] In step 300, the grid cells are... The water transparency value is denoted as Grid cells are determined based on the transparency range to which the water transparency value belongs. The transparency level is denoted as . Regarding transparency levels From the control grid cell set Select the same transparency level. The control grid cells form a transparency control subset. Transparency level Corresponding attenuation correction factor Determined according to the following formula: In the formula, This represents the median. It is a positive number. Used to prevent The ratio is abnormal when close to zero. Grid cell Belongs to the transparency level At the same time, correct the optical inversion depth. Determined according to the following formula: Transparency level Corresponding corrected error Determined according to the following formula: In the formula, for The number of control grid cells in the [database]. The minimum value of the corrected error corresponding to each transparency level is denoted as [the value]. Transparency level Corresponding transparency weighting factor Determined according to the following formula: In the formula, This is a non-negative stable term. When the corrected error for each transparency level is greater than zero, A value of 0 is acceptable. Transparency level. The larger the corresponding corrected error, The smaller the value, the lower the weight of the corrected optical inversion depth in the subsequent fusion.

[0034] Table 1. Calculation Table for Transparency Level Parameters

[0035] In Table 1, The depth is 0.42m. The transparency weighting factor corresponding to the medium transparency level is 0.42 divided by 0.57, which is approximately 0.74. The transparency weighting factor corresponding to the low transparency level is 0.42 divided by 0.82, which is approximately 0.51. The correction error is larger at the low transparency level, and the transparency weighting factor corresponding to the corrected optical inversion depth is reduced.

[0036] In step 400, the grid cells are... The chart-bound water depth is recorded as Grid unit It also has the function of correcting optical inversion water depth. Water depth constrained by nautical charts At that time, the difference in water depth Determined according to the following formula: From the control grid cell set Control grid cells that simultaneously provide corrected optical inversion depth and chart-constrained depth are selected to form a chart control subset. Control subset based on nautical charts Determine the scale parameters of nautical chart differences : ocean Figure 1 Consistency indicators Determined according to the following formula: In the formula, The value range is from 0 to 1. The closer it is to 1, the closer the corrected optical inversion depth is to the chart-constrained depth; The closer a value is to 0, the greater the difference between the corrected optical inversion depth and the chart-constrained depth. (Chart difference scale parameter) From the chart control subset The statistical results of the difference, Hai Figure 1 Consistency indicators can vary with the differences in water depth within the grid cells.

[0037] In step 400, the grid cells are... The number of spaceborne laser sounding points in the neighborhood is denoted as , grid cells The area of ​​the neighborhood is denoted as Spaceborne laser depth sounding data in grid cells Distribution density within the neighborhood Determined according to the following formula: From the control grid cell set A control grid cell with a spaceborne laser depth sounding point within its neighborhood is selected to form a laser control subset. According to the laser control subset Determine the reference distribution density : Laser reliability index Determined according to the following formula: In the formula, The value ranges from 0 to 1. Grid cell The denser the satellite-borne laser depth sounding points in the neighborhood, the better. The larger, The closer to 1, the better. Grid cell When there are no spaceborne laser sounding points in the vicinity , .

[0038] Exemplary, within the same grid cell, the sea Figure 1 The consistency index reflects the degree of closeness between the corrected optical inversion depth and the chart-constrained depth, while the laser reliability index reflects the degree of support for the laser reference depth at bathymetry points in the spatial neighborhood. Figure 1The consistency index and laser reliability index are respectively included in the subsequent fusion weight calculation. When the difference between the corrected optical inversion depth and the chart-constrained depth is small, the reference role of the chart-constrained depth in the fusion is retained; when the spaceborne laser bathymetry points are densely distributed in the neighborhood, the reference role of the laser reference depth in the fusion is improved. The above two indices participate in the weight determination from the perspectives of depth consistency and space sampling support, respectively.

[0039] In step 500, the substrate type is denoted as From the control grid cell set The selected substrate type is The control grid units form a substrate control subset. Substrate type Corresponding optical inversion error Determined according to the following formula: In the formula, for The number of control grid cells in the matrix. The minimum value of the optical inversion error corresponding to each substrate type is denoted as... Substrate type Corresponding substrate correction factor Determined according to the following formula: In the formula, This is a non-negative stable term. When the optical inversion error for each substrate type is greater than zero, It can be 0. Substrate type The larger the corresponding optical inversion error, The smaller the value, the lower the weight of the corrected optical inversion depth for that substrate type.

[0040] Table 2 Calculation Table of Substrate Type Parameters

[0041] In Table 2, The depth is 0.62m. The correction factor for muddy bottom is 0.62 divided by 0.78, approximately 0.79. The correction factor for reef bottom is 0.62 divided by 0.95, approximately 0.65. The correction factor for mixed bottom is 0.62 divided by 0.86, approximately 0.72. After the optical inversion errors under different bottom types are included in the bottom correction factor, the influence of bottom type on the reflectance of multispectral remote sensing images can be reflected in the fusion weight of correcting optical inversion depth.

[0042] In step 500, when determining the fusion weights, the unnormalized weights corresponding to each water depth are first determined. For grid cells... Let the transparency level be 1. Substrate type is Correction of weights before normalization of optical inversion depth. Laser reference water depth weights before normalization Weights of water depth before normalization under nautical chart constraints Determine them according to the following formulas respectively: The weights before normalization were derived from statistics on transparency level errors, seabed type errors, laser bathymetry point distribution density, and the difference between the corrected optical inversion depth and the chart-constrained depth. The fusion weight of each grid cell was determined based on the data status of the corresponding grid cell.

[0043] Figure 3 The relationship for determining the fusion weights in this embodiment is illustrated. See also... Figure 3 The diagram shows four input branches on the left. The first input branch is the transparency level, which determines the transparency weighting factor, and this factor is further used to determine the weighting for correcting optical inversion depth. The second input branch is coastal sediment data, which determines the sediment correction factor, and this factor is further used to correct the weighting for correcting optical inversion depth. The third input branch is the laser reliability index, which determines the laser reference depth weighting. The fourth input branch is the sea... Figure 1 Consistency indicators, sea Figure 1 Consistency indices are used to determine the weighting of chart-constrained depths. These indices consist of transparency grade, transparency weighting factor, corrected optical inversion depth weighting, coastal sediment data, sediment correction factor, corrected optical inversion depth weighting correction, laser reliability index, laser reference depth weighting, and sea surface sediment data. Figure 1 Consistency indicators and chart-constrained depth weights are jointly entered into the pre-normalization weight determination stage. After the pre-normalization weights are determined, deviation adjustment, validity labeling judgment, and normalization processing are performed sequentially to obtain the fusion weights. The right side of the figure shows the output content corresponding to the fusion weights, including the fusion weights for corrected optical inversion depth, laser reference depth, and chart-constrained depth. The arrows in the figure indicate the transfer relationship between the input quantities, intermediate processing quantities, and output quantities, illustrating that the fusion weights are determined by transparency level, sea level, and other factors. Figure 1 Consistency index, laser reliability index and coastal sediment data were used to determine the fusion weights corresponding to the three water depths, which were then adjusted for deviation, judged for validity and normalized.

[0044] Exemplary, in grid cells In the normalized calculation, the unnormalized weights of the corrected optical inversion depth are jointly determined by the transparency weight factor and the seabed correction factor. The transparency weight factor reflects the statistical error of the corrected optical inversion depth under the corresponding transparency level, while the seabed correction factor reflects the statistical error of the corrected optical inversion depth under the corresponding seabed type. The unnormalized weights of the laser reference depth are determined by the laser reliability index, and the unnormalized weights of the chart-constrained depth are determined by the seabed reference index. Figure 1 Consistency indicators are determined. After the initial weight determination is completed, the unnormalized weights of the corrected optical inversion depth and the chart-constrained depth are adjusted according to the degree of deviation of each relative to the laser reference depth. Deviation adjustment only applies to the sea... Figure 1 Consistency index lower than sea Figure 1 Execution is performed when the median of the consistency index is reached and the laser reference water depth has an effective water depth value.

[0045] When the sea Figure 1 Consistency indicators Below the control grid cell set China Overseas Figure 1 Median of consistency index And laser reference water depth When an effective water depth value is available, the laser reference water depth is used to adjust the deviation between the corrected optical inversion water depth and the chart-constrained water depth. First, the laser difference scale parameters are determined. : In the formula, This is a set of control grid cells with an effective laser reference water depth. The weights before normalization of the corrected optical inversion water depth after deviation adjustment. Weights of water depth before normalization under nautical chart constraints Determined according to the following formula: when Not less than Or laser reference water depth When there is no effective water depth value, , The greater the deviation between the corrected optical inversion depth or the chart-constrained depth and the laser reference depth, the greater the reduction in the weight before normalization.

[0046] The validity markers for corrected optical inversion depth, laser reference depth, and chart-constrained depth are respectively denoted as: , and When a corresponding water depth has a valid water depth value, the validity flag is set to 1; when a corresponding water depth does not have a valid water depth value, the validity flag is set to 0. Normalize the denominator. Determined according to the following formula: when When the fusion weights are used, they are determined according to the following formula: In the formula, , and These represent the fusion weights for corrected optical inversion depth, laser reference depth, and chart-constrained depth, respectively. Data sources without valid depth values ​​are marked as valid (0), and their corresponding fusion weights are 0. Non-water cell grid units in the nearshore area of ​​the target coastal zone are not included in depth fusion calculations.

[0047] In step 600, the grid cells The depth of fusion Determined according to the following formula: When a certain water depth does not have a valid water depth value, the corresponding fusion weight is 0, and the corresponding water depth does not participate in the fusion calculation. When spaceborne laser bathymetry data is distributed along the trajectory, electronic chart water depth data is partially missing, or optical inversion water depth is partially invalid, data sources with valid water depth values ​​can still participate in the water depth fusion of the corresponding grid cell.

[0048] Exemplary, when grid unit When none of the three water depths have a valid water depth value, the grid cell is not included in the fused water depth calculation. When only one water depth in grid cell i has a valid water depth value, that valid water depth value is used as the grid cell value. The depth of water fusion. When grid units When two water depths in grid cell i have valid water depth values, only the pre-normalized weights corresponding to the two valid water depths are normalized before weighted fusion. When all three water depths in grid cell i have valid water depth values, the pre-normalized weights corresponding to the three water depths are normalized before weighted fusion.

[0049] This embodiment selects five grid cells G1 to G5 in the nearshore area of ​​the target coastal zone to illustrate the calculation process of steps 300 to 600. G1 is located in a sandy shoal area with high transparency, G2 is located in a muddy nearshore area with medium transparency, G3 is located in a reef edge area with low transparency, the area corresponding to G4 lacks effective spaceborne laser bathymetry points, and the area corresponding to G5 lacks effective electronic chart depth data. Table 3-1 lists the basic depth data of G1 to G5, and Table 3-2 lists the index data, fusion weights, and fused depths of G1 to G5. The weights in Table 3-2 are the weights after deviation adjustment and normalization.

[0050] Table 3-1 Grid Cell Foundation Water Depth Data

[0051] Table 3-2 Grid Unit Indicators and Fusion Results

[0052] In G1, the transparency level is high transparency, and the substrate type is sandy substrate. According to Tables 1 and 2, It is 1.00. It is 1.00. It is 1.00. G1's It is 0.76. The value was 0.91, indicating no deviation adjustment was triggered. After normalization, It is 0.37. It is 0.29. The value is 0.34. The fusion water depth is... The calculated value is approximately 5.35m. When the three water depths are similar, the fusion weight remains relatively balanced.

[0053] In G2, the transparency level is medium transparency, and the substrate type is muddy. According to Tables 1 and 2, It is 0.74. It is 0.79. for It is approximately 0.58. G2's It is 0.68. It is 0.87. After normalization, It is 0.27. It is 0.32. The value is 0.41. The fusion water depth is... The calculated depth is approximately 6.76 m. The sediment correction factor for muddy sediment is lower than that for sandy sediment, and the weight of the optical inversion depth correction is correspondingly reduced.

[0054] In G3, the transparency level is low, and the substrate type is reef substrate. According to Tables 1 and 2, It is 0.51. It is 0.65. for It is approximately 0.33. G3's It is 0.55, lower than the sea level. Figure 1 Median of consistency index Furthermore, the laser reference water depth has an effective water depth value. It is 5.90m. It is 5.40m. The deviation is 6.20m. The deviation of the chart-constrained water depth from the laser reference water depth is greater than the deviation of the corrected optically inverted water depth from the laser reference water depth. After deviation adjustment and normalization are completed, It is 0.15. It is 0.78. The value is 0.07. The fusion water depth is... The calculated depth is approximately 6.10m. When low transparency levels, reef bottom, and chart discrepancies coexist, the fusion weights are tilted towards the laser reference depth.

[0055] In G4, the laser reference water depth has no effective water depth value. The denominator is 0. The unnormalized weights of the laser reference depth are not included. After correcting for optical inversion depth and chart-constrained depth, the values ​​are renormalized. It is 0.39. The value is 0.61. The fusion water depth is... The calculated depth is approximately 8.09m. The missing laser reference water depth in this grid cell is not included in the fusion calculation.

[0056] In G5, the chart-constrained depth is a value with no effective depth. The value is 0. After correcting the optical inversion depth and laser reference depth and renormalizing, It is 0.39. The value is 0.61. The fusion water depth is... The calculated depth is approximately 4.77m. The missing chart-constrained depths within this grid cell are not included in the fusion calculation.

[0057] Tables 3-1 and 3-2 show that the same fusion rule can adapt to different transparency levels, different seabed types, and different data missing states. When transparency is high and the influence of the seabed is small, the corrected optical inversion depth retains a high weight. When transparency is low and the influence of the seabed is large, the weight of the corrected optical inversion depth decreases. When the difference between the corrected optical inversion depth and the chart-constrained depth is large and the laser reference depth is effective, the weight of depths that deviate significantly from the laser reference depth decreases.

[0058] Figure 4 The diagram illustrates the fusion weight changes for a typical grid cell in this embodiment. See also... Figure 4The horizontal axis represents the grid cells, which are G1, G2, G3, G4, and G5 in sequence; the vertical axis represents the fusion weights, ranging from 0 to 1. Each grid cell in the figure corresponds to three sets of bar data: the diagonally filled bars represent the weights for corrected optical inversion depth, the dotted filled bars represent the weights for laser reference depth, and the gray filled bars represent the weights for chart-constrained depth. The weights for corrected optical inversion depth (G1) are 0.37, laser reference depth (Laser Reference Depth) is 0.29, and chart-constrained depth (Chart Constrained Depth) is 0.34; for G2, the weights are 0.27, 0.32, and 0.41; for G3, the weights are 0.15, 0.78, and 0.07; for G4, the weights are 0.39, 0.00, and 0.61; and for G5, the weights are 0.39, 0.61, and 0.00. The corresponding values ​​above each bar in the figure illustrate the differences in the fusion weights of the corrected optical inversion depth, laser reference depth, and chart-constrained depth for different grid cells under varying conditions of transparency level, seabed type, effectiveness of laser reference depth, and effectiveness of chart-constrained depth. Specifically, the laser reference depth has a higher weight in G3, zero weight in G4, and zero weight in G5, reflecting the changes in fusion weights under different data conditions.

[0059] To illustrate the variation in water depth error in this embodiment, the control grid cell set is used. Select a set of verification grid cells in addition to the above. Verify the set of grid cells. Not included in inversion parameters, attenuation correction coefficient, sea level Figure 1 Calculation of consistency scale parameters, laser confidence scale parameters, and substrate correction factors. Verification of the grid cell set. The water depth control data was used as the verification water depth to calculate the errors of the initial optical inversion water depth, the corrected optical inversion water depth, and the fused water depth results. The root mean square error was used. Determined according to the following formula: In the formula, This indicates the water depth result to be evaluated. Indicates verification grid cell Verify water depth, To verify the number of grid cells, the verification results are shown in Table 4.

[0060] Table 4. Statistical Table of Grid Cell Errors

[0061] As shown in Table 4, within the same set of verification grid cells... In the process, after transparency level is used for attenuation correction, the corrected optical inversion depth compared to the initial optical inversion depth is... The depth was reduced from 1.08m to 0.84m; further utilization of the sea... Figure 1 After determining the fusion weights using consistency indicators, laser reliability indicators, and coastal sediment data, the fused water depth results are... The error decreased to 0.57m. The above-mentioned error changes indicate that the transparency level and sea level... Figure 1 After determining the weights for consistency indicators, laser reliability indicators, and coastal sediment data in grid-by-grid fusion, the local reliability of the fused water depth results is improved.

[0062] like Figure 5 As shown, this embodiment also provides a coastal nearshore water depth inversion and optimization system based on the fusion of multi-source remote sensing data and nautical charts. This system is used to execute the aforementioned coastal nearshore water depth inversion and optimization method based on the fusion of multi-source remote sensing data and nautical charts. The system includes a data acquisition and grid construction module, a three-type water depth determination module, a transparency correction module, an index determination module, a fusion weight determination module, and a fused water depth generation module. Each module can be deployed on the same computing device, or in a computing environment consisting of a server, a remote sensing data processing terminal, and a water depth results management terminal.

[0063] The data acquisition and grid construction module is used to acquire spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data of the target coastal nearshore area. It also performs spatial benchmark unification, water depth benchmark unification, and grid cell matching on the spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data to obtain unified grid data.

[0064] Specifically, the data acquisition and grid construction module first determines the spatial extent of the nearshore area of ​​the target coastal zone and divides it into grid cells based on the water pixel coverage in the multispectral remote sensing image. Spaceborne laser bathymetry data is matched to the corresponding grid cell or its neighborhood according to the planar position of the bathymetry points. Water transparency remote sensing data is matched to the corresponding grid cell according to the pixel position. Electronic nautical chart depth data is matched to the corresponding grid cell according to the spatial position of depth points, isobaths, or gridded nautical chart depth values. Coastal sediment data is matched to the corresponding grid cell according to the distribution range of sediment type. When spaceborne laser bathymetry data and electronic nautical chart depth data have different depth references, the data acquisition and grid construction module converts them to the same depth reference. When multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data have different spatial references, the data acquisition and grid construction module converts these data to the same spatial reference.

[0065] In the unified grid data, each grid cell records at least the grid cell location, the water spectral information corresponding to the multispectral remote sensing image, the water transparency value corresponding to the water transparency remote sensing data, the nautical chart depth value corresponding to the electronic nautical chart depth data, and the seabed type corresponding to the coastal sediment data. When a spaceborne laser bathymetry point exists within the grid cell or its neighborhood, the unified grid data also records the location and depth value of the spaceborne laser bathymetry point. The data acquisition and grid construction module sends the unified grid data to three types of depth determination modules, transparency correction modules, index determination modules, and fusion weight determination modules.

[0066] The three types of water depth determination modules are used to determine the initial optical inversion water depth of each grid cell based on multispectral remote sensing images in the unified grid data, to determine the laser reference water depth corresponding to each grid cell based on spaceborne laser bathymetry data in the unified grid data, and to determine the chart-constrained water depth of each grid cell based on electronic chart water depth data in the unified grid data.

[0067] The three-category depth determination module extracts the water spectral features of each grid cell from multispectral remote sensing images in the unified grid data. It then uses depth values ​​provided by at least one of the spaceborne laser bathymetry data and electronic chart depth data from the unified grid data as depth control data to establish an inversion relationship between water spectral features and depth. The module substitutes the water spectral features of each grid cell into the inversion relationship to obtain the initial optical inversion depth for each grid cell. For the laser reference depth, the module identifies spaceborne laser bathymetry points within the neighborhood of each grid cell, determines distance weights based on the distance between the spaceborne laser bathymetry points and the corresponding grid cells, and weights the depth values ​​of the spaceborne laser bathymetry points according to these distance weights to obtain the laser reference depth for each grid cell. For the chart-constrained depth, the module determines the chart-constrained depth for each grid cell based on electronic chart depth points, isobath depth values, or gridded chart depth values ​​matched to the grid cell or its neighborhood. The three-level water depth determination module sends the initial optical inversion water depth, laser reference water depth, and chart-constrained water depth to the transparency correction module, index determination module, fusion weight determination module, and fusion water depth generation module.

[0068] The transparency correction module is used to determine the transparency level of each grid cell based on the water transparency remote sensing data in the unified grid data, and to perform attenuation correction on the initial optical inversion water depth according to the transparency level, so as to obtain the corrected optical inversion water depth of each grid cell.

[0069] The transparency correction module reads the water transparency value of each grid cell and determines the transparency level based on the transparency range to which the water transparency value belongs. The module selects control grid cells with the same transparency level from the control grid cell set, calculates the ratio between the water depth control data and the initial optically retrieved water depth, and obtains the attenuation correction coefficient for the corresponding transparency level. The module uses the attenuation correction coefficient to correct the initial optically retrieved water depth, obtaining the corrected optically retrieved water depth. The module can also obtain a transparency weighting factor based on the error between the corrected optically retrieved water depth and the water depth control data at the same transparency level, and send the transparency weighting factor to the fusion weight determination module.

[0070] The index determination module is used to determine the sea level of each grid cell based on the difference between the corrected optical inversion depth and the chart-constrained depth. Figure 1 Consistency indicators were established, and the laser reliability index of each grid cell was determined based on the distribution density of spaceborne laser depth sounding data in the neighborhood of each grid cell.

[0071] The index determination module calculates the depth difference between the corrected optical inversion depth and the chart-constrained depth within the same grid cell, and determines the depth based on the depth difference and the chart difference scale parameter. Figure 1 Consistency index. The greater the difference in water depth, the more consistent the sea... Figure 1 The lower the consistency of the consistency index characterization, the higher the reliability of the laser reliability index characterization. The index determination module also counts the number of satellite-borne laser bathymetry points in the neighborhood of each grid cell and determines the distribution density based on the number of satellite-borne laser bathymetry points and the neighborhood area of ​​the grid cell; the index determination module determines the laser reliability index based on the distribution density and the reference distribution density. The denser the satellite-borne laser bathymetry points in the neighborhood of the grid cell, the higher the reliability of the laser reliability index characterization. The index determination module will... Figure 1 The consistency index and laser reliability index are sent to the fusion weight determination module.

[0072] The fusion weight determination module is used to determine the weights based on transparency level, sea level, etc. Figure 1 By using consistency index, laser reliability index, and coastal sediment data from the unified grid data, the fusion weights of corrected optical inversion depth, laser reference depth, and chart-constrained depth in each grid cell are determined.

[0073] Specifically, the fusion weight determination module determines the transparency weight factor for correcting the optically inverted water depth based on the transparency level, determines the seabed correction factor based on coastal sediment data, and determines the pre-normalized weights for correcting the optically inverted water depth based on the transparency weight factor and the seabed correction factor. The fusion weight determination module determines the pre-normalized weights for the laser reference water depth based on the laser reliability index, and determines the pre-normalized weights for the laser reference water depth based on the sea... Figure 1 Consistency indicators determine the unnormalized weights of chart-constrained depths. When the sea... Figure 1 Consistency index lower than sea Figure 1 When the median of the consistency index is reached and the laser reference water depth has an effective water depth value, the fusion weight determination module adjusts the pre-normalization weights of the corrected optical inversion water depth and the chart-constrained water depth according to the degree of deviation of the corrected optical inversion water depth and the chart-constrained water depth from the laser reference water depth.

[0074] The fusion weight determination module is also used to determine whether the corrected optical inversion depth, laser reference depth, and chart-constrained depth have valid depth values. When a corresponding depth has a valid depth value, the fusion weight determination module sets the corresponding validity flag to 1; when a corresponding depth does not have a valid depth value, the fusion weight determination module sets the corresponding validity flag to 0. Based on the validity flag, the fusion weight determination module normalizes the unnormalized weights of the corrected optical inversion depth, laser reference depth, and chart-constrained depth to obtain the fusion weights for the corrected optical inversion depth, laser reference depth, and chart-constrained depth. The fusion weight corresponding to a data source without a valid depth value is 0.

[0075] The fused water depth generation module is used to fuse the corrected optical inversion water depth, laser reference water depth, and chart-constrained water depth in each grid cell according to the fusion weight, so as to obtain the fused water depth result of the nearshore area of ​​the target coastal zone.

[0076] Specifically, the fused depth generation module reads the corrected optical inversion depth, laser reference depth, chart-constrained depth, and corresponding fusion weights from each grid cell. When all three depth types in a grid cell have valid depth values, the fused depth generation module performs weighted fusion according to the fusion weights corresponding to the three depth types. When only two depth types in a grid cell have valid depth values, the fused depth generation module performs weighted fusion according to the fusion weights corresponding to the two valid depth types. When only one depth type in a grid cell has a valid depth value, the fused depth generation module uses that valid depth value as the fused depth for that grid cell. When none of the three depth types in a grid cell have valid depth values, the fused depth generation module does not perform depth fusion calculations for that grid cell. The fused depths of all grid cells together form the fused depth result for the nearshore area of ​​the target coastal zone.

[0077] In this embodiment, the data acquisition and grid construction modules form unified grid data, ensuring that spaceborne laser bathymetry data, multispectral remote sensing imagery, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data have a corresponding relationship within the same grid cell. The three-category depth determination modules obtain the initial optical inversion depth, laser reference depth, and chart-constrained depth. The transparency correction module performs attenuation correction on the initial optical inversion depth based on the transparency level. The index determination module generates the sea... Figure 1 Consistency indicators and laser reliability indicators. The fusion weight determination module will incorporate transparency level, sea level, and other relevant factors. Figure 1 Consistency indicators, laser reliability indicators, and coastal sediment data are converted into grid-by-grid fusion weights. The fused water depth generation module obtains the fused water depth result based on the fusion weights. The data transfer relationships between the modules correspond to steps 100 to 600 and can be used to implement all the processing procedures in the above method embodiments.

[0078] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0079] Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. Furthermore, those skilled in the art will recognize that, based on the ideas of this invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for optimizing the inversion of nearshore water depth in the coastal zone based on the fusion of multi-source remote sensing data and nautical charts, characterized in that, include: Acquire spaceborne laser bathymetry data, multispectral remote sensing images, water transparency remote sensing data, electronic nautical chart depth data, and coastal sediment data of the target coastal nearshore area. Then, perform spatial benchmark unification, water depth benchmark unification, and grid cell matching on the spaceborne laser bathymetry data, the multispectral remote sensing images, the water transparency remote sensing data, the electronic nautical chart depth data, and the coastal sediment data to obtain unified grid data. The initial optical inversion water depth of each grid cell is determined based on the multispectral remote sensing image in the unified grid data; the laser reference water depth corresponding to each grid cell is determined based on the spaceborne laser bathymetry data in the unified grid data; and the chart-constrained water depth of each grid cell is determined based on the electronic chart water depth data in the unified grid data. The transparency level of each grid cell is determined based on the water transparency remote sensing data in the unified grid data, and the initial optical inversion water depth is attenuated and corrected based on the transparency level to obtain the corrected optical inversion water depth of each grid cell. The chart consistency index of each grid cell is determined based on the difference between the corrected optical inversion depth and the chart-constrained depth, and the laser reliability index of each grid cell is determined based on the distribution density of the spaceborne laser bathymetry data in the neighborhood of each grid cell. Based on the transparency level, the chart consistency index, the laser reliability index, and the coastal sediment data in the unified grid data, determine the fusion weights of the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth in each grid cell; According to the fusion weight, the corrected optical inversion water depth, the laser reference water depth, and the chart-constrained water depth in each grid cell are fused to obtain the fused water depth result of the nearshore area of ​​the target coastal zone.

2. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion as described in claim 1, characterized in that, Spatial benchmark unification, water depth benchmark unification, and grid cell matching are performed on the spaceborne laser bathymetry data, the multispectral remote sensing imagery, the water transparency remote sensing data, the electronic nautical chart depth data, and the coastal sediment data to obtain unified grid data, including: The grid cells are divided based on the pixel coverage area of ​​the multispectral remote sensing image; The spaceborne laser bathymetry data, the water transparency remote sensing data, the electronic nautical chart water depth data, and the coastal sediment data are respectively matched to the corresponding grid cells; The spaceborne laser bathymetry data and the electronic nautical chart depth data are converted to the same depth reference. Within the same grid cell, establish the correspondence between the multispectral remote sensing image, the water transparency remote sensing data, the spaceborne laser bathymetry data, the electronic nautical chart water depth data, and the coastal sediment data to obtain the unified grid data.

3. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion as described in claim 1, characterized in that, Determining the initial optical inversion water depth of each grid cell based on the multispectral remote sensing image in the unified grid data includes: Extract water spectral features from the multispectral remote sensing images in the unified grid data; Using at least one of the spaceborne laser bathymetry data and the electronic nautical chart water depth data in the unified grid data as water depth control data, an inversion relationship between the spectral characteristics of the water body and the water depth is established. The initial optical inversion depth of each grid cell is determined based on the inversion relationship.

4. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion as described in claim 1, characterized in that, Determining the laser reference water depth corresponding to each grid cell based on the spaceborne laser bathymetry data in the unified grid data includes: Determine the spaceborne laser depth sounding points within the neighborhood of each grid cell; The distance weight of the spaceborne laser sounding point is determined based on the distance between the spaceborne laser sounding point and the corresponding grid cell. The water depth values ​​of the spaceborne laser depth sounding points are weighted according to the distance weight to obtain the laser reference water depth corresponding to each grid cell.

5. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion as described in claim 1, characterized in that, Based on the water transparency remote sensing data in the unified grid data, the transparency level of each grid cell is determined, and the initial optically retrieved water depth is attenuated and corrected according to the transparency level to obtain the corrected optically retrieved water depth of each grid cell, including: The water transparency value of each grid cell is determined based on the remote sensing data of water transparency. The transparency level of each grid cell is determined based on the transparency range to which the water transparency value belongs; The attenuation correction coefficient of each grid cell is determined based on the transparency level, and the lower the water transparency represented by the transparency level, the greater the correction magnitude of the attenuation correction coefficient to the initial optical inversion water depth. The initial optical inversion depth is corrected using the attenuation correction coefficient to obtain the corrected optical inversion depth for each grid cell.

6. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion according to claim 1, characterized in that, The chart consistency index for each grid cell is determined based on the difference between the corrected optical inversion depth and the chart-constrained depth, including: Calculate the depth difference between the corrected optical inversion depth and the chart-constrained depth in the same grid cell; The degree of water depth deviation between the corrected optical inversion water depth and the chart-constrained water depth in the same grid cell is determined based on the water depth difference. The chart consistency index of each grid cell is determined based on the degree of water depth deviation, and the greater the degree of water depth deviation, the lower the consistency represented by the chart consistency index.

7. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion according to claim 1, characterized in that, The laser reliability index for each grid cell is determined based on the distribution density of the spaceborne laser depth sounding data within the neighborhood of each grid cell, including: Count the number of spaceborne laser sounding points in the neighborhood of each grid cell; The distribution density is determined based on the number of spaceborne laser depth sounding points and the neighborhood area of ​​each grid cell. The laser reliability index of each grid cell is determined based on the distribution density, and the higher the distribution density, the higher the reliability represented by the laser reliability index.

8. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion according to claim 1, characterized in that, Based on the transparency level, the chart consistency index, the laser reliability index, and the coastal seabed data in the unified grid data, the fusion weights of the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth in each grid cell are determined, including: The weight of the corrected optical inversion water depth is determined based on the transparency level, and the lower the water transparency represented by the transparency level, the smaller the weight of the corrected optical inversion water depth. The weight of the laser reference water depth is determined based on the laser reliability index, and the higher the reliability represented by the laser reliability index, the greater the weight of the laser reference water depth. The weights of the corrected optical inversion depth and the chart-constrained depth are adjusted according to the chart consistency index. When the consistency represented by the chart consistency index decreases and the laser reference depth has an effective depth value, the weights of the corrected optical inversion depth and the chart-constrained depth that deviate significantly from the laser reference depth are reduced. The weights of the corrected optical inversion depth are adjusted based on the coastal sediment data in the unified grid data; The weights of the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth are normalized to obtain the fusion weights.

9. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion according to claim 8, characterized in that, The weights for correcting the optically inverted water depth are adjusted based on the coastal sediment data in the unified grid data, including: The substrate type of each grid cell is determined based on the coastal zone substrate data in the unified grid data; The degree of influence of the substrate on the reflection of the multispectral remote sensing image in each grid cell is determined based on the substrate type. The weight of the corrected optical inversion depth is adjusted according to the degree of reflection influence, and the greater the degree of reflection influence, the smaller the weight of the corrected optical inversion depth.

10. The coastal nearshore water depth inversion and optimization method based on multi-source remote sensing data and nautical chart fusion according to claim 1, characterized in that, According to the fusion weights, the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth in each grid cell are fused to obtain the fused depth result for the nearshore area of ​​the target coastal zone, including: In each grid cell, it is determined whether the corrected optical inversion depth, the laser reference depth, and the chart-constrained depth have valid depth values. The weights corresponding to water depths that do not have valid water depth values ​​are reset to zero, and the weights corresponding to water depths that have valid water depth values ​​are normalized. According to the normalized weights, the corrected optical inversion water depth, the laser reference water depth, and the chart-constrained water depth with effective water depth values ​​are weighted and fused to obtain the grid fused water depth of each grid cell; The merged water depth of the target coastal nearshore area is formed based on the merged water depth of each grid cell.

Images