A method for realizing three-dimensional visualization of landslide catastrophe process

Abstract

The application discloses a kind of to realize landslide catastrophe process three-dimensional visualization method, belong to unmanned aerial vehicle measurement technical field, comprising the following steps: S1, data acquisition and pre-processing;S2, fill the process dynamic simulation;S3, multi-source data fusion modeling;S4, three-dimensional visualization system display.The application uses the above-mentioned one to realize landslide catastrophe process three-dimensional visualization method, improves data consistency, mutual verification reduces error, effectively eliminates the error introduced due to different data sources, provides reliable basis for subsequent analysis, improves dynamic simulation precision, provides guarantee for landslide fill the process dynamic simulation and multi-source data fusion.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) measurement technology, and in particular to a method for realizing three-dimensional visualization of landslide disaster processes. Background Technology

[0002] Traditional landslide monitoring methods primarily rely on ground surveys or single remote sensing images. These methods are not only inefficient and costly, but also struggle to acquire high-precision 3D topographic data. This limitation restricts the effective monitoring and analysis of landslide dynamics. Most current 3D landslide models are built upon static digital elevation model (DEM) data, lacking effective simulation of the dynamic evolution of landslides, such as depression filling and changes in the sliding path. This static display method cannot comprehensively reflect the development process of landslides and their potential risks. Accurate spatial fusion of geological borehole data and surface DEM data is a major challenge. Due to the different sources and processing techniques of these two data types, their spatial fusion accuracy is low, making it difficult to accurately construct underground stratigraphic structures, which is crucial for a deeper understanding of landslide mechanisms. Existing technologies often suffer from inefficiency when performing large-scale 3D topographic rendering, especially exhibiting significant lag in dynamic process visualization. This not only affects the user experience but also reduces the effectiveness and reliability of disaster analysis. Summary of the Invention

[0003] The purpose of this invention is to provide a method for realizing three-dimensional visualization of landslide disaster processes, improving data consistency, cross-verifying to reduce errors, effectively eliminating errors introduced by different data sources, providing a reliable basis for subsequent analysis, improving the accuracy of dynamic simulation, and providing a guarantee for dynamic simulation of landslide filling processes and multi-source data fusion.

[0004] To achieve the above objectives, the present invention provides a method for realizing three-dimensional visualization of landslide disaster processes, comprising the following steps:

[0005] S1. Data acquisition and preprocessing;

[0006] S2. Dynamic simulation of the depression-filling process;

[0007] S3, Multi-source data fusion modeling;

[0008] S4, 3D visualization system display.

[0009] Preferably, S1 specifically includes the following steps:

[0010] S1.1 Data Acquisition: Using UAV oblique photogrammetry technology, multi-view images of the ground surface at different time points before and after the landslide were acquired, and a DEM digital elevation model was constructed. The images were then processed by software to generate high-resolution DEM data.

[0011] S1.2 Data preprocessing: Spatiotemporal registration of DEM data before and after the landslide is performed to eliminate differences in coordinate systems, resolution, and geometric deformation.

[0012] Preferably, S1.2 specifically includes the following steps:

[0013] S1.2.1 Data preparation: Input the DEM data before the landslide (DEM1), the DEM data after the landslide (DEM2), and the orthophoto range of the landslide body; ensure that DEM1 and DEM2 have the same geographic coordinate system, projection method, and resolution; process the spatial offset caused by different UAV image acquisition times; calculate and save the difference DEM raster data of DEMΔ = DEM2 - DEM1.

[0014] S1.2.2 Unify the coordinate system of DEM data before and after the landslide: Check the coordinate system of DEM data before and after the landslide and analyze whether they have the same coordinate system. If the coordinate systems are inconsistent, the coordinate systems need to be transformed.

[0015] S1.2.3 Matching the resolution of DEM data before and after the landslide: Check the resolution of DEM data before and after the landslide and analyze whether they have the same resolution. If the resolutions are different, the data with the lower resolution needs to be resampled.

[0016] S1.2.4 Resampling and Matching: Use tools to adjust the resolution of the DEM data;

[0017] S1.2.5 Correction and registration: mainly includes registration based on control points and correction based on feature points;

[0018] S1.2.6 Result Verification: Using the orthogonal range of the landslide body, the landslide area is excluded, and the difference is calculated and the root mean square error (RMSE) is verified against the surrounding terrain to determine the matching accuracy.

[0019] Preferably, S1.2.6 specifically includes the following operations:

[0020] The formula for calculating the difference is as follows:

[0021] ΔZ(x,y)=Z DEM2 (x,y)-Z DEM1 (x,y);

[0022] Among them, Z DEM1 (x,y) and Z DEM2 (x,y) represent the elevation values ​​DEM1 and DEM2 at grid point (x,y), respectively.

[0023] Visual verification and statistical verification;

[0024] Visual verification: Load the original DEM1, DEM2 and the registered DEM2 using software, and observe whether there are any obvious misalignments;

[0025] Statistical validation: Calculate the overall root mean square error (RMSE) of DEM1 and DEM2:

[0026]

[0027] Where N is the number of sample points, i.e., the total number of pixels involved in the calculation, ΔZ(x i ,y i ) represents the elevation difference at the i-th pixel, i.e., the difference in elevation between the DEM and the i-th pixel before and after the landslide at position x. i y i The difference in elevation.

[0028] Preferably, S2 specifically includes the following steps:

[0029] S2.1: Data preparation: Input the DEM data before the landslide (DEM1) and the DEM data after the landslide (DEM2), ensuring that DEM1 and DEM2 have completed spatiotemporal registration. Spatiotemporal registration includes coordinate system, resolution, and correction registration. Calculate the pixel-by-pixel difference of the DEM data before and after the landslide, and generate a dynamic transition sequence of the depression filling process.

[0030] S2.2, Difference weight control;

[0031] S2.3, Dynamic transition sequence generation;

[0032] S2.4 Implement timing animation playback in the CesiumJS engine.

[0033] Preferably, S2.2 specifically includes the following operations:

[0034] S2.2.1 Difference Calculation: Perform pixel-by-pixel difference calculation on DEM1 and DEM2 within the orthographic range of the landslide body:

[0035] ΔZ(x,y)=Z DEM2 (x,y)-Z DEM1 (x,y);

[0036] Positive values ​​indicate areas filled by landslides due to topographic uplift, while negative values ​​indicate areas of landslides due to topographic subsidence.

[0037] S2.2.2, Dynamic Transition Formula:

[0038] The dynamic changes in the depression-filling process are controlled using the time scaling parameter t = [0,1].

[0039] Z t (x,y)=Z DEM1(x,y)-t·ΔZ(x,y);

[0040] When t = 0, the result is the initial state DEM1; when t = 1, the result is the target state DEM2; when t = (0,1), the result is the intermediate state of the depression filling process.

[0041] S2.2.3, Difference Weight Smoothing Process: Combine the inverse distance weighted difference method to optimize the intermediate state and make the filling process smoother;

[0042] S2.3 specifically includes the following operations:

[0043] S2.3.1 Time Segmentation: Divide the filling process into several equal time intervals, such as 10 frames or 20 frames, with each frame corresponding to a time ratio t;

[0044] S2.3.2 Generating intermediate DEM data: For each time scale t, calculate the corresponding difference result Z according to the formula. t (x,y), generate a series of intermediate DEM data;

[0045] S2.3.3, Generate terrain tiles for each time period: Generate corresponding terrain tile files for each intermediate DEM to optimize the rendering speed of large-scale terrain on the web.

[0046] S2.4 specifically includes the following operations:

[0047] S2.4.1, Data Loading;

[0048] S2.4.2, Animation playback;

[0049] S2.4.3 Visualization effects.

[0050] Preferably, S3 specifically includes the following operations:

[0051] S3.1 Data preparation: Input geological borehole data and surface DEM data. The geological borehole data includes borehole location coordinates, depth and properties of each layer of rock and soil. Output an integrated "surface-subsurface" model with DEM topographic model superimposed on the underground geological model.

[0052] S3.2. Use the radial basis function (RBF) interpolation method to process irregularly distributed data points;

[0053] S3.3 Construct an underground stratigraphic structure model;

[0054] S3.4, 3D fusion modeling.

[0055] Preferably, S3.3 specifically includes the following operations:

[0056] S3.3.1 Data preprocessing: Clean the geological borehole data to remove outliers and erroneous data;

[0057] S3.3.2 Convert the data to RBF interpolation format;

[0058] S3.3.3, RBF interpolation implementation steps;

[0059] S3.3.4, Construct the difference model;

[0060] S3.3.5, Generate a three-dimensional stratigraphic model;

[0061] S3.4 specifically includes the following operations:

[0062] S3.4.1 Data Fusion: The generated underground stratigraphic model is fused with the surface DEM data;

[0063] S3.4.2 Constructing an integrated model: Use CesiumJS to load the surface DEM and subsurface strata model to create a comprehensive "surface-subsurface" integrated model.

[0064] Preferably, S4 specifically includes the following steps:

[0065] S4.1, Interaction of stratigraphic profiles;

[0066] S4.2, Strata blasting interaction;

[0067] S4.3, landslide-related stacking.

[0068] Therefore, the present invention adopts the above-mentioned method for realizing three-dimensional visualization of landslide disaster process, which improves data consistency, cross-verifies and reduces errors, effectively eliminates errors introduced by different data sources, provides a reliable basis for subsequent analysis, improves the accuracy of dynamic simulation, and provides a guarantee for dynamic simulation of landslide filling process and multi-source data fusion.

[0069] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0070] Figure 1 This invention provides a method for realizing three-dimensional visualization of landslide disaster processes, showing a comparison of the effects before and after a landslide. Figure 1 Image (a) in the image is the one taken before the landslide. Figure 1 (b) in the image is the post-landslide image;

[0071] Figure 2 This is an example image of terrain data display for a method of realizing three-dimensional visualization of landslide disaster process according to the present invention;

[0072] Figure 3This is a schematic diagram of landslide hazard annotation, illustrating a method for realizing three-dimensional visualization of landslide disaster processes according to the present invention.

[0073] Figure 4 This is a schematic diagram of a three-dimensional fusion modeling process of geological borehole data for realizing a three-dimensional visualization method of landslide disaster process according to the present invention. Figure 4 (a) in the figure is a three-dimensional display of borehole elevation and borehole lithology elevation. Figure 4 (b) in the figure is a three-dimensional representation of the continuous surface topography calculated by interpolation based on the borehole elevation of a. Figure 4 (c) in the diagram represents a three-dimensional mesh constructed by dividing a unit spatial region based on (b). Figure 4 (d) in the diagram is a three-dimensional schematic diagram showing the spatial geological conditions based on (c).

[0074] Figure 5 This invention provides a three-dimensional geological model diagram for realizing a three-dimensional visualization method of landslide disaster processes;

[0075] Figure 6 This is a schematic diagram of the terrain structure for a method to realize three-dimensional visualization of landslide disaster process according to the present invention;

[0076] Figure 7 This is an interactive diagram of strata blasting, illustrating a method for realizing three-dimensional visualization of landslide disaster processes according to the present invention.

[0077] Figure 8 This is a schematic diagram of the groundwater content of a method for realizing three-dimensional visualization of landslide disaster process according to the present invention;

[0078] Figure 9 This is a schematic diagram illustrating the three-dimensional fusion of ground stress and displacement vector data in a method for realizing three-dimensional visualization of landslide disaster processes according to the present invention. Figure 9 (a) in the diagram is the generalized shear strain contour plot. Figure 9 (b) in the diagram is the displacement contour map. Detailed Implementation

[0079] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0080] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0081] Example 1

[0082] This invention provides a method for realizing three-dimensional visualization of landslide disaster processes, comprising the following steps:

[0083] S1. Data Acquisition and Preprocessing; specifically including the following steps:

[0084] S1.1 Data Acquisition: Using UAV oblique photogrammetry technology, multi-view images of the ground surface at different time points before and after the landslide were acquired, and a DEM digital elevation model was constructed. The images were then processed by software to generate high-resolution DEM data.

[0085] S1.2 Data Preprocessing: Spatiotemporal registration of DEM data before and after the landslide is performed to eliminate differences in coordinate systems, resolutions, and geometric deformations. This includes the following steps:

[0086] S1.2.1 Data preparation: Input the DEM data before the landslide (DEM1), the DEM data after the landslide (DEM2), and the orthophoto range of the landslide body; ensure that DEM1 and DEM2 have the same geographic coordinate system, projection method, and resolution; process the spatial offset caused by different UAV image acquisition times; calculate and save the difference DEM raster data of DEMΔ = DEM2 - DEM1.

[0087] S1.2.2 Unify the coordinate system of DEM data before and after the landslide: Check the coordinate system of DEM data before and after the landslide and analyze whether they have the same coordinate system. Using ArcMap or ArcGIS Pro in ArcGIS, load the two DEM data and check the coordinate system information of DEM1 and DEM2. If the coordinate systems are inconsistent, the coordinate systems need to be converted.

[0088] For data that needs to be transformed, a projection transformation can be performed. Taking ArcGIS as an example, input DEM2 as the raster data to be processed into ArcGIS's ProjectRaster tool, set the output coordinate system to the coordinate system of DEM1, run the tool, generate a new DEM2 file, and convert DEM2 to the same coordinate system as DEM1.

[0089] S1.2.3 Matching the resolution of DEM data before and after the landslide: Check the resolution of DEM data before and after the landslide. Check the resolution information (usually in meters or degrees) in the DEM layer properties and analyze whether they have the same resolution. If the resolution is different, the data with the lower resolution needs to be resampled.

[0090] S1.2.4 Resampling and Matching: Adjust the resolution of the DEM data using ArcGIS's Resample tool;

[0091] Input DEM2 as the raster data to be processed, set the output cell size to the target resolution (consistent with DEM1), set the resampling method (Bilinear or Cubic is recommended to preserve terrain details), run the tool, and generate a new DEM2 file.

[0092] S1.2.5 Correction and registration: mainly includes registration based on control points and correction based on feature points;

[0093] (1) Control point-based matching

[0094] Manually perform geometric correction on DEM2 using ArcGIS's Georeferencing tool. Load DEM1 and DEM2 into the same map window; enable the Georeferencing toolbar and select DEM2 as the layer to be corrected; add ground control points (GCPs) to DEM2 and align them to their corresponding positions in DEM1. Control points can be selected based on obvious terrain features (such as mountain peaks, valleys, etc.), and at least four control points are required to ensure accuracy; set the transformation method (Polynomial Transformation is recommended); after completing the correction, save the registration results.

[0095] (2) Feature point-based matching

[0096] If manual registration is inefficient, you can use ArcGIS Image Analysis tools or third-party plugins (such as ENVI) for automatic registration. Automatic registration is usually based on feature point matching algorithms (such as SIFT or SURF).

[0097] S1.2.6 Result Verification: Using the orthogonal range of the landslide body, excluding the landslide area, the difference is calculated and the root mean square error (RMSE) is verified against the surrounding terrain to determine the matching accuracy. This specifically includes the following operations:

[0098] The formula for calculating the difference is as follows:

[0099] ΔZ(x,y)=Z DEM2 (x,y)-Z DEM1 (x,y);

[0100] Among them, Z DEM1 (x,y) and Z DEM2 (x,y) represent the elevation values ​​of DEM1 and DEM2 at grid point (x,y), respectively; using ArcToolbox, a difference raster DEMΔ is generated, the orthophoto range of the landslide body is overlaid, and the landslide body area is removed.

[0101] Visual verification and statistical verification;

[0102] Visual verification: Using ArcGIS software, load the original DEM1, DEM2 and the registered DEM2 to observe whether there are obvious misalignments; use the Transparency function to overlay the two DEMs to visually check the registration effect.

[0103] Statistical validation: Calculate the overall root mean square error (RMSE) of DEM1 and DEM2:

[0104]

[0105] Where N is the number of sample points, i.e., the total number of pixels involved in the calculation, ΔZ(x i ,y i ) represents the elevation difference at the i-th pixel, i.e., the difference in elevation between the DEM and the i-th pixel before and after the landslide at position x. i y i The difference in elevation.

[0106] If the RMSE is less than the preset threshold (e.g., 0.3 meters), the registration is successful. Comparison of images before and after the landslide: Figure 1 As shown, where Figure 1 Image (a) in the image is the one taken before the landslide. Figure 1 (b) in the image is the image after the landslide.

[0107] S2. Dynamic simulation of the depression-filling process; specifically including the following steps:

[0108] S2.1: Data preparation: Input the DEM data before the landslide (DEM1) and the DEM data after the landslide (DEM2), ensuring that DEM1 and DEM2 have completed spatiotemporal registration. Spatiotemporal registration includes coordinate system, resolution, and correction registration. Calculate the pixel-by-pixel difference of the DEM data before and after the landslide, and generate a dynamic transition sequence of the filling process.

[0109] S2.2, Difference Weight Control; specifically includes the following operations:

[0110] S2.2.1 Difference Calculation: Perform pixel-by-pixel difference calculation on DEM1 and DEM2 within the orthographic range of the landslide body:

[0111] ΔZ(x,y)=Z DEM2 (x,y)-Z DEM1 (x,y);

[0112] A positive difference indicates a depression-filled area where the terrain has risen, while a negative difference indicates a landslide area where the terrain has subsided.

[0113] S2.2.2, Dynamic Transition Formula:

[0114] The dynamic changes in the depression-filling process are controlled using the time scaling parameter t = [0,1].

[0115] Z t (x,y)=Z DEM1 (x,y)-t·ΔZ(x,y);

[0116] When t = 0, the result is the initial state DEM1; when t = 1, the result is the target state DEM2; when t = (0,1), the result is the intermediate state of the depression filling process.

[0117] S2.2.3, Difference Weight Smoothing Process: Combining the inverse distance weighted difference method to optimize the intermediate state makes the filling process smoother; as shown below:

[0118]

[0119] Where, d i Z represents the distance between the current location and the sampling point, where p is the power exponent, and Z is the distance between the current location and the sampling point. t (x,y) represents the interpolation result of the target position, w i Z represents the weight of the i-th sampling point. i This represents the actual measured value of the i-th sampling point.

[0120] S2.3, Dynamic Transition Sequence Generation; specifically includes the following operations:

[0121] S2.3.1 Time Segmentation: Divide the depression filling process into several equal time intervals, such as 10 frames or 20 frames. Each frame corresponds to a time ratio t, where t = [0.0, 0.1, 0.2, ..., 1.0].

[0122] S2.3.2 Generating intermediate DEM data: For each time scale t, calculate the corresponding difference result Z according to the formula. t (x,y), generate a series of intermediate DEM data, and save the intermediate DEM data of each frame as an independent file;

[0123] S2.3.3. Generate terrain tiles for each time period: For each intermediate DEM, generate corresponding terrain tile files (.terrain) to reduce network data transmission and optimize the web rendering speed of large-scale terrain;

[0124] Terrain is a terrain data format specifically designed for Cesium, based on Quantized Mesh technology. This format significantly reduces the storage space of terrain data and improves loading and rendering performance by compressing and quantizing the data. Cesium uses this format to process large terrain datasets, especially terrain tiles that require efficient streaming. An example of terrain data display is shown below. Figure 2 As shown.

[0125] S2.4 Implementing time-series animation playback in the CesiumJS engine; specifically including the following operations:

[0126] S2.4.1 Data Loading: The intermediate DEM data sequence generated is loaded using the CesiumJS engine.

[0127] S2.4.2 Animation Playback: The Clock module of CesiumJS is used to implement time-series animation playback. Based on the set total event length and playback speed, the current time-series DEM data is loaded frame by frame, and previous time-series data is removed. This creates a time-series terrain change animation. The built-in clock component of CesiumJS is used to control the pause and playback of the animation.

[0128] S2.4.3, Visualization effects;

[0129] Frame-by-frame display: The intermediate DEM data of each frame is rendered using CesiumJS to create a continuous dynamic transition effect.

[0130] Color mapping: Using different colors to represent changes in terrain height enhances visual appeal.

[0131] Overlaying satellite imagery: The filling process can be overlaid with satellite imagery to further enhance realism. Landslide hazard markings are shown below. Figure 3As shown.

[0132] S3. Multi-source data fusion modeling: Combining geological borehole data, the radial basis function (RBF) interpolation method is used to fit the underground stratigraphic structure of the landslide area. The stratigraphic model is then fused with the surface DEM in three dimensions to construct an integrated "surface-underground" model of the landslide body, such as... Figure 4 As shown, where, Figure 4 (a) in the figure is a three-dimensional display of borehole elevation and borehole lithology elevation. Figure 4 (b) in the figure is a three-dimensional representation of the continuous surface topography calculated by interpolation based on the borehole elevation of a. Figure 4 (c) in the diagram represents a three-dimensional mesh constructed by dividing a unit spatial region based on (b). Figure 4 (d) in the diagram is a three-dimensional schematic diagram showing the spatial geological conditions based on (c).

[0133] Specifically, the following operations are included:

[0134] S3.1 Data Preparation: Input geological borehole data and surface DEM data. The geological borehole data includes borehole location coordinates, depth and properties of each layer of rock and soil (such as thickness, type, etc.). Output an integrated "surface-subsurface" model that overlays the DEM terrain model with the subsurface geological model.

[0135] S3.2. Use the radial basis function (RBF) interpolation method to process irregularly distributed data points;

[0136] Radial basis function (RBF) interpolation is a meshless interpolation method commonly used for interpolating data in multidimensional spaces. Its basic idea is to predict the values ​​of unknown points using known values ​​and locations. For simulating the underground geological structure of landslide areas, RBF can effectively handle irregularly distributed data points.

[0137] The basic form of RBF interpolation is:

[0138]

[0139] Where f(x) is the predicted value; α i These are the weighting coefficients; φ() is the radial basis function; p(x) is the polynomial part, usually a first or second-order polynomial, used to ensure the smoothness of the interpolation result; N is the number of known sample points; x i x and y represent the positions of the known sample point and the predicted point, respectively. Commonly used radial basis functions include the Gaussian function and the multiple quadratic surface function.

[0140] S3.3 Constructing an underground stratigraphic structure model; specifically including the following operations:

[0141] S3.3.1 Data Preprocessing: Clean the geological borehole data to remove outliers and erroneous data; convert the borehole data into a form suitable for RBF interpolation, that is, determine the coordinates, depth and corresponding stratigraphic information of each borehole.

[0142] S3.3.2 Convert the data to RBF interpolation format;

[0143] Construct 3D point cloud data, convert the stratum information at the depth of each borehole into 3D spatial points (X,Y,Z), and attach corresponding attribute values.

[0144] Sandstone layer at a depth of 5m in borehole BK-001: 3D coordinates (100.0, 200.0, -5.0), attribute values: lithology type = 2 (sandstone), permeability = 0.1cm / s.

[0145] Borehole BK-002 at a depth of 3m in silty clay layer: 3D coordinates (150.0, 250.0, -3.0), attribute values: lithology type = 3, permeability = 0.005cm / s.

[0146] Data formatting: Organize the data into an input format suitable for RBF interpolation: Input matrix: contains the three-dimensional coordinates (X,Y,Z) of all points; Output vector: attribute values ​​of the corresponding points (such as lithology type, permeability).

[0147] S3.3.3, RBF interpolation implementation steps;

[0148] Commonly used RBF types include:

[0149] Gaussian function:

[0150] Multiple quadratic functions:

[0151] Thin plate spline: φ(r)=r 2 ln(r).

[0152] S3.3.4, Construct the difference model;

[0153] Implement RBF interpolation using Python's scipy.interpolate.Rbf.

[0154] S3.3.5, Generate a three-dimensional stratigraphic model;

[0155] like Figure 5 As shown, the prediction results are converted into a 3D model (such as VTK, GeoJSON, or a format supported by Cesium), and the VTU model is converted into glTF format for easy overlay display of CesiumJS and DEM on the web.

[0156] S3.4, 3D fusion modeling; specifically including the following operations:

[0157] S3.4.1 Data Fusion: The generated underground stratigraphic model is fused with the surface DEM data; a direct overlay method is used, with the surface DEM as the top layer and the underground stratigraphic model arranged below it.

[0158] S3.4.2 Constructing an integrated model: Use CesiumJS to load the surface DEM and subsurface strata model to create a comprehensive "surface-subsurface" integrated model.

[0159] In the visualization environment, the viewing angle can be dynamically adjusted to observe different levels of the geological structure, analyze the overall characteristics of the landslide body, and the topographic structure diagram is shown below. Figure 6 As shown.

[0160] S4. 3D Visualization System Demonstration: In the 3D visualization system for landslide disaster processes, based on the original dynamic simulation of the filling process, data from the geological model is further overlaid to achieve a more comprehensive and in-depth spatial geographic analysis. Specifically, this includes the following steps:

[0161] S4.1, Interaction of stratigraphic profiles;

[0162] Stratigraphic Profile: By adding a stratigraphic profile function to the 3D visualization system, users can select specific locations for vertical cutting to view the stratigraphic structure at different depths. This helps in understanding the complexity of the geological structure within the landslide area and its impact on landslide activity.

[0163] Interactive exploration: Allows users to freely choose the cutting location and direction, and observe changes in the strata in real time. This interactive approach makes the analysis process more intuitive and flexible.

[0164] S4.2, Strata blasting interaction;

[0165] Exploded Strata View: Provides an "exploded" view mode, allowing users to more clearly see the distribution of different underground layers. An interactive diagram of the exploded strata view is shown below. Figure 7 As shown.

[0166] S4.3, landslide-related stacking;

[0167] Overlay of related data: By combining geological attribute data (such as lithology, water content, etc.) and the analysis results of the landslide body's above-ground and underground features, detailed underground information is provided to users. This not only improves the model's practicality but also enhances users' overall understanding of the landslide body's above-ground and underground characteristics. A schematic diagram of underground water content is shown below. Figure 8 , Figure 9 As shown, where, Figure 9(a) in the diagram is the generalized shear strain contour plot. Figure 9 (b) in the diagram is the displacement contour map.

[0168] Therefore, the present invention adopts the above-mentioned method for realizing three-dimensional visualization of landslide disaster process, which improves data consistency, cross-verifies and reduces errors, effectively eliminates errors introduced by different data sources, provides a reliable basis for subsequent analysis, improves the accuracy of dynamic simulation, and provides a guarantee for dynamic simulation of landslide filling process and multi-source data fusion.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for realizing three-dimensional visualization of landslide disaster processes, characterized in that: Includes the following steps: S1. Data acquisition and preprocessing; Specifically, the following steps are included: S1.1 Data Acquisition: Using UAV oblique photogrammetry technology, multi-view images of the ground surface at different time points before and after the landslide were acquired, and a DEM digital elevation model was constructed. The images were then processed by software to generate high-resolution DEM data. S1.2 Data preprocessing: Spatiotemporal registration of DEM data before and after the landslide is performed to eliminate differences in coordinate system, resolution, and geometric deformation. S2. Dynamic simulation of the depression-filling process; Specifically, the following steps are included: S2.1: Data preparation: Input the DEM data before the landslide (DEM1) and the DEM data after the landslide (DEM2), ensuring that DEM1 and DEM2 have completed spatiotemporal registration. Spatiotemporal registration includes coordinate system, resolution, and correction registration. Calculate the pixel-by-pixel difference of the DEM data before and after the landslide, and generate a dynamic transition sequence of the depression filling process. S2.2, Difference weight control; Specifically, the following operations are included: S2.2.1 Difference Calculation: Perform pixel-by-pixel difference calculation on DEM1 and DEM2 within the orthographic range of the landslide body: ； Positive values ​​indicate areas filled by landslides due to topographic uplift, while negative values ​​indicate areas of landslides due to topographic subsidence. S2.2.2, Dynamic Transition Formula: Use time ratio parameters Controlling the dynamic changes during the depression-filling process: ； when When, the result is the initial state DEM1; when When, the result is the target state DEM2, when At that time, the result is an intermediate state in the filling process; S2.2.3, Difference Weight Smoothing Process: Combine the inverse distance weighted difference method to optimize the intermediate state and make the filling process smoother; S2.3, Dynamic transition sequence generation; Specifically, the following operations are included: S2.3.1 Time Segmentation: Divide the filling process into several equal time intervals, such as 10 frames or 20 frames, with each frame corresponding to a time proportion. ; S2.3.2, Generate intermediate DEM data: for each time scale Calculate the corresponding difference result according to the formula. This generates a series of intermediate DEM data; S2.3.3, Generate terrain tiles for each time period: Generate corresponding terrain tile files for each intermediate DEM to optimize the rendering speed of large-scale terrain on the web. S2.4 Implementing time-series animation playback in the CesiumJS engine; S2.4 specifically includes the following operations: S2.4.1, Data Loading; S2.4.2, Animation playback; S2.4.3 Visualization effects; S3 Multi-source data fusion modeling; Specifically, the following operations are included: S3.1 Data preparation: Input geological borehole data and surface DEM data. The geological borehole data includes borehole location coordinates, depth and rock and soil properties of each layer. Output an integrated "surface-subsurface" model with DEM topographic model superimposed on subsurface geological model. S3.

2. Use the radial basis function (RBF) interpolation method to process irregularly distributed data points; S3.3 Construct an underground stratigraphic structure model; S3.4, 3D fusion modeling; S4, 3D visualization system display.

2. The method for realizing three-dimensional visualization of landslide disaster process according to claim 1, characterized in that: S1.2 specifically includes the following steps: S1.2.1 Data Preparation: Input the pre-landslide DEM data (DEM1), the post-landslide DEM data (DEM2), and the orthophoto range of the landslide body; ensure that DEM1 and DEM2 have the same geographic coordinate system, projection method, and resolution; handle spatial offsets caused by different UAV image acquisition times; calculate and save. The difference in DEM raster data; S1.2.2 Unify the coordinate system of DEM data before and after the landslide: Check the coordinate system of DEM data before and after the landslide and analyze whether they have the same coordinate system. If the coordinate systems are inconsistent, the coordinate systems need to be transformed. S1.2.3 Matching the resolution of DEM data before and after the landslide: Check the resolution of DEM data before and after the landslide and analyze whether they have the same resolution. If the resolutions are different, the data with the lower resolution needs to be resampled. S1.2.4 Resampling and Matching: Use tools to adjust the resolution of the DEM data; S1.2.5 Correction and registration: mainly includes registration based on control points and correction based on feature points; S1.2.6 Result Verification: Using the orthogonal range of the landslide body, the landslide area is excluded, and the difference is calculated and the root mean square error (RMSE) is verified against the surrounding terrain to determine the matching accuracy.

3. The method for realizing three-dimensional visualization of landslide disaster process according to claim 2, characterized in that: S1.2.6 specifically includes the following operations: The formula for calculating the difference is as follows: ； in, and These represent the grid points of DEM1 and DEM2, respectively. elevation value at the location ; Visual verification and statistical verification; Visual verification: Load the original DEM1, DEM2 and the registered DEM2 using software, and observe whether there are any obvious misalignments; Statistical validation: Calculate the overall root mean square error of DEM1 and DEM2. : ； in, This represents the number of sample points, i.e., the total number of pixels involved in the calculation. For the first The elevation difference at each pixel, i.e., the difference in position of the DEM before and after the landslide. , The difference in elevation.

4. The method for realizing three-dimensional visualization of landslide disaster process according to claim 3, characterized in that: S3.3 specifically includes the following operations: S3.3.1 Data preprocessing: Clean the geological borehole data to remove outliers and erroneous data; S3.3.2 Convert the data to RBF interpolation format; S3.3.3, RBF interpolation implementation steps; S3.3.4, Construct the difference model; S3.3.5, Generate a three-dimensional stratigraphic model; S3.4 specifically includes the following operations: S3.4.1 Data Fusion: The generated underground stratigraphic model is fused with the surface DEM data; S3.4.2 Build an integrated model: Use CesiumJS to load the surface DEM and underground strata model to create a comprehensive "surface-underground" integrated model.

5. The method for realizing three-dimensional visualization of landslide disaster process according to claim 4, characterized in that: S4 specifically includes the following steps: S4.1, Interaction of stratigraphic profiles; S4.2, Strata blasting interaction; S4.3, landslide-related stacking.

Images