Method and system for three-dimensional modeling of slope in hard mountainous area based on remote sensing technology

Abstract

The invention relates to the technical field of image data processing, in particular to a difficult mountainous area slope three-dimensional modeling method and system based on a remote sensing technology. The method comprises the following steps: obtaining an original remote sensing data set of a target hard mountain area, and carrying out data preprocessing to obtain a standardized remote sensing data set; performing target hard mountainous area surface feature extraction on the standardized remote sensing data set to obtain hard mountainous area surface feature data; constructing a three-dimensional terrain model by using the earth surface feature data of the hard mountain area to obtain preliminary three-dimensional terrain model data; and performing time sequence analysis on the standardized remote sensing data set, and performing slope dynamic change analysis on the initial three-dimensional terrain model data by using a time sequence analysis result to obtain slope displacement change data. Through three-dimensional modeling, dynamic change analysis, stability evaluation and landslide early warning, a set of complete disaster prevention and reduction system for the hard mountainous area is constructed, and the prediction and prevention capability for landslide disasters is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a three-dimensional modeling method and system for slopes in difficult mountainous areas based on remote sensing technology. Background Art

[0002] Currently, the technical methods available on the market for three-dimensional modeling of slopes in difficult mountainous areas still have significant flaws and are unable to fully address the comprehensive needs of complex terrain, dynamic changes, and disaster prediction. Existing remote sensing data processing technologies mostly focus on general analysis of large areas and lack specific optimization for the unique terrain conditions of difficult mountainous areas. Due to the complex terrain and landforms in difficult mountainous areas, existing modeling methods struggle to fully capture the local details of slope changes. As a result, the generated three-dimensional models have limitations in accuracy and applicability, making it difficult to meet the needs of dynamic slope change analysis. Monitoring of slope dynamics currently relies primarily on traditional field observations or time series analysis of single remote sensing images. This approach faces significant challenges in difficult mountainous areas due to the complex terrain and wide monitoring range. Existing risk assessment methods are mostly based on general indicators and insufficiently consider the unique geological conditions and environmental factors in difficult mountainous areas, resulting in a lack of targeted assessment results. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a three-dimensional modeling method and system for slopes in difficult mountainous areas based on remote sensing technology to solve at least one of the above technical problems.

[0004] To achieve the above objectives, a 3D modeling method for slopes in difficult mountainous areas based on remote sensing technology includes the following steps: Step S1: Obtain the original remote sensing dataset of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing dataset; Step S2: extracting the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain surface feature data of the difficult mountainous area; Step S3: constructing a three-dimensional terrain model using surface feature data of the difficult mountainous area to obtain preliminary three-dimensional terrain model data; Step S4: performing time series analysis on the standardized remote sensing data set, and using the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; Step S5: Performing environmental factor stability analysis on the slope displacement change data to obtain slope stability assessment data; Step S6: constructing a landslide prediction model based on the slope stability assessment data, and performing landslide warning based on the landslide prediction results to generate landslide warning information data; Step S7: Use the landslide warning information data to conduct risk assessment of the target difficult mountainous area, and formulate a prevention strategy based on the risk assessment results to obtain a landslide warning and prevention plan.

[0005] The present invention obtains raw remote sensing data from the target rugged mountainous area and preprocesses it to ensure data consistency and accuracy, providing a reliable foundation for subsequent analysis. Standardization removes noise and deviations, ensuring that the data fully reflects the actual conditions in the mountainous area. Surface feature extraction extracts surface morphological information with regional characteristics, providing detailed data for subsequent three-dimensional modeling, effectively capturing local detailed changes in slopes, and overcoming the limitations of existing modeling methods. A preliminary three-dimensional terrain model is constructed using surface feature data, providing precise spatial support for subsequent dynamic analysis. Time series analysis reveals surface change trends, providing historical data support for slope dynamic change analysis, ensuring the accuracy and timeliness of slope displacement change data, and providing a basis for stability assessment. Environmental factor stability analysis is performed on slope displacement change data to identify the impact of various environmental factors and generate accurate stability assessment data, providing a scientific basis for landslide prediction and prevention. Based on the stability assessment, a landslide prediction model is constructed to provide landslide warnings, providing decision support for early identification of potential risks and disaster prevention. Finally, risk assessment is conducted through landslide early warning information, prevention strategies are formulated, and a complete risk prevention and control system is formed to enhance the disaster prevention and mitigation capabilities in dangerous mountainous areas and reduce losses caused by landslide disasters.

[0006] Preferably, the present invention further provides a three-dimensional modeling system for slopes in difficult mountainous areas based on remote sensing technology, which is used to execute the above-mentioned three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology. The three-dimensional modeling system for slopes in difficult mountainous areas based on remote sensing technology comprises: The data preprocessing module is used to obtain the original remote sensing data set of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing data set; The surface feature extraction module is used to extract the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain the surface feature data of the difficult mountainous area; The 3D terrain modeling module is used to construct a 3D terrain model using surface feature data in difficult mountainous areas to obtain preliminary 3D terrain model data; The slope dynamic analysis module is used to perform time series analysis on standardized remote sensing data sets and use the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; Slope stability assessment module, used to perform environmental factor stability analysis on slope displacement change data to obtain slope stability assessment data; The landslide prediction and warning module is used to build a landslide prediction model based on slope stability assessment data, and to issue landslide warnings based on landslide prediction results, thereby generating landslide warning information data; The risk assessment and prevention strategy module is used to use landslide warning information data to conduct risk assessment in target difficult mountainous areas, and to formulate prevention strategies based on the risk assessment results to obtain landslide warning and prevention plans.

[0007] The present invention improves the quality and consistency of data by processing standardized remote sensing data sets, providing a reliable foundation for subsequent analysis. It accurately extracts surface features in difficult mountainous areas, helping to accurately identify and analyze the geographical and environmental characteristics of mountainous areas. By constructing a preliminary three-dimensional terrain model, it provides spatial support and detailed terrain information for slope dynamic changes and stability assessments. It uses time series analysis to identify slope displacement changes, providing key data for landslide monitoring and dynamic assessments. It conducts scientific assessments of slope stability, helps identify high-risk areas, and provides a basis for landslide prediction. Through the construction of a landslide prediction model, landslide warning information is generated in a timely manner, improving disaster prevention and response capabilities. Based on landslide warning information, a comprehensive risk assessment is conducted, an effective landslide prevention strategy is formulated, and regional risk management capabilities are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 A schematic flow chart of the steps of the three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology of the present invention; Figure 2 for Figure 1 Detailed step flow diagram of step S1; Figure 3 for Figure 1 Detailed step flow chart of step S2 in FIG. DETAILED DESCRIPTION

[0009] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0010] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0011] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0012] To achieve this, please refer to Figures 1 to 3 The present invention provides a three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology, the method comprising the following steps: Step S1: Obtain the original remote sensing dataset of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing dataset; Step S2: extracting the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain surface feature data of the difficult mountainous area; Step S3: constructing a three-dimensional terrain model using surface feature data of the difficult mountainous area to obtain preliminary three-dimensional terrain model data; Step S4: performing time series analysis on the standardized remote sensing data set, and using the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; Step S5: Performing environmental factor stability analysis on the slope displacement change data to obtain slope stability assessment data; Step S6: constructing a landslide prediction model based on the slope stability assessment data, and performing landslide warning based on the landslide prediction results to generate landslide warning information data; Step S7: Use the landslide warning information data to conduct risk assessment of the target difficult mountainous area, and formulate a prevention strategy based on the risk assessment results to obtain a landslide warning and prevention plan.

[0013] In the embodiment of the present invention, reference Figure 1 FIG. 1 is a flow chart showing the steps of a method for three-dimensional modeling of slopes in difficult mountainous areas based on remote sensing technology according to the present invention. In this example, the method for three-dimensional modeling of slopes in difficult mountainous areas based on remote sensing technology includes the following steps: Step S1: Obtain the original remote sensing dataset of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing dataset; This embodiment of the present invention first acquires a raw remote sensing dataset of the target rugged mountainous area. The remote sensing data comes from high-resolution satellite imagery or drone aerial photography. Data acquisition uses Landsat 8 satellite imagery, WorldView-3 imagery, or drone-mounted LiDAR equipment. The image data must fully cover the target rugged mountainous area and have high temporal resolution over a specified time span. After acquiring the remote sensing data, data preprocessing is performed. First, the image data undergoes geometric correction, using ground control points (GCPs) to calibrate geometric distortion and ensure spatial accuracy. Next, radiometric correction is performed to eliminate the effects of atmospheric and lighting factors on the remote sensing images. After radiometric correction, image fusion processing is performed. If the image data is multi-temporal, image registration methods are used to precisely align pixel-level features by comparing feature points, ensuring that the multi-temporal images have a consistent coordinate system. To enhance the effectiveness of subsequent analysis, multispectral image data is standardized, converting image data from different sensors into a unified data format. Image enhancement, such as stretching, is also performed to increase image contrast and enhance the prominence of feature points. The standardized remote sensing datasets are further processed using image processing software such as ENVI or ERDAS IMAGINE to ensure consistent data format for each temporal phase, thus ensuring data validity and reliability. Finally, the standardized remote sensing datasets are stored in a geographic information system (GIS) platform.

[0014] Step S2: extracting the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain surface feature data of the difficult mountainous area; This embodiment of the present invention first extracts surface features of the target rugged mountainous area from a standardized remote sensing dataset. This process uses multispectral imagery from the standardized remote sensing dataset, combined with remote sensing image processing techniques, to extract key surface features in the rugged mountainous area. Surface type classification is performed using vegetation indices (such as NDVI), water indexes (such as MNDWI), and terrain elevation data. The standardized remote sensing dataset is input into a remote sensing image processing tool such as ENVI or ERDAS IMAGINE. The images are classified using feature classification algorithms such as support vector machines (SVM), maximum likelihood classification (MLC), or decision trees to generate a spatial distribution map of various surface features. Using elevation data captured from the remote sensing data, a digital elevation model (DEM) is used to extract terrain information for the target rugged mountainous area, further refining the surface feature data. For complex terrain areas, LiDAR point cloud data is combined with the data to perform more detailed surface classification, particularly on slopes, accurately extracting different types of vegetation cover, exposed rock, soil, and vegetation growth. For water bodies, water indexes are combined to extract water area information, ensuring accurate identification of the spatial distribution of water bodies. By extracting and analyzing these surface feature data, we can eventually generate surface feature data for dangerous mountainous areas.

[0015] Step S3: constructing a three-dimensional terrain model using surface feature data of the difficult mountainous area to obtain preliminary three-dimensional terrain model data; This embodiment of the present invention utilizes surface feature data from difficult mountainous areas to construct a three-dimensional terrain model. This first requires fusing elevation data from standardized remote sensing datasets with surface feature data. For optical remote sensing images, stereo or high-resolution image pairs are used to generate a preliminary digital elevation model (DEM) through optical image matching. Disparity calculations are then used to extract surface elevation variation information, leveraging the geometric relationships between the images to obtain basic terrain elevation data. During LiDAR data processing, a high-precision three-dimensional point cloud model is generated using point cloud data. Filtering techniques are then used to remove non-ground point clouds above the ground, ensuring that the point cloud data primarily reflects ground elevation features. Furthermore, a triangulated irregular network (TIN) method is used to convert the point cloud data into a three-dimensional terrain model, ensuring accurate representation of elevation details. Combining the DEM and point cloud data, the elevation data is spatially interpolated using interpolation algorithms such as kriging or inverse distance weighted interpolation, refining the elevation values ​​and forming a complete three-dimensional surface model. Furthermore, slope and aspect analysis tools were used to simulate terrain slope changes and surface water flow directions, helping to reveal slope stability and potential landslide risk areas. Ultimately, the resulting preliminary 3D terrain model data not only contained detailed elevation information for the target difficult mountainous area, but also incorporated surface features and slope changes.

[0016] Step S4: performing time series analysis on the standardized remote sensing data set, and using the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; The embodiments of the present invention perform time series analysis on standardized remote sensing datasets. First, remote sensing image data from multiple time points must be collected to ensure temporal continuity and spatial consistency. Using this image data, image registration techniques are used to spatially align the images, eliminating geometric errors between images captured at different times and enabling accurate overlay. Next, based on the time series characteristics of remote sensing data, a difference image analysis method is applied to differentiate the remote sensing images at each moment and calculate the magnitude of change between different time periods. To obtain accurate change information, a method based on a change detection algorithm is used to compare and analyze images from different time points, extracting dynamic change information about the slope area. Using this change data, combined with terrain analysis tools, a dynamic change analysis is performed on the slope area initially constructed within the three-dimensional terrain model to obtain slope displacement change data. By extracting elevation change information from the slope area and combining it with spatial interpolation techniques such as kriging interpolation or inverse distance weighted methods, the displacement calculation is refined and a slope displacement change map is generated. The results of the time series analysis are used to comprehensively monitor the dynamic behavior of the slope through trend analysis of the time series data.

[0017] Step S5: Performing environmental factor stability analysis on the slope displacement change data to obtain slope stability assessment data; This embodiment of the present invention first performs an environmental factor stability analysis on slope displacement change data. By analyzing environmental factors that influence slope stability, slope stability is assessed. Meteorological data, particularly precipitation data, is collected for the target difficult mountainous area. Precipitation intensity and frequency are analyzed to assess its impact on slope stability. Vegetation indices (such as NDVI) are extracted from remote sensing imagery to analyze the contribution of vegetation cover to slope erosion resistance and stability. Next, combined with geological exploration data, the rock and soil structure and lithology of the target difficult mountainous area are analyzed to determine the stability of different rock strata. Spatial analysis tools, such as spatial overlay analysis and buffer zone analysis within a geographic information system (GIS), are used to comprehensively assess the impact of various environmental factors on slope stability. Specifically, spatial interpolation methods (such as inverse distance weighted methods) are used to predict the spatial distribution of soil type and lithology, and correlation analysis is performed with the slope displacement change data. For topographic factors, a three-dimensional terrain model and slope analysis tools are used to analyze the impact of different slope gradients on slope stability. In this process, multivariate statistical analysis methods, such as principal component analysis (PCA) or correlation analysis, are used to comprehensively analyze the relationship between various environmental factors and slope displacement changes to obtain slope stability assessment data.

[0018] Step S6: constructing a landslide prediction model based on the slope stability assessment data, and performing landslide warning based on the landslide prediction results to generate landslide warning information data; The present invention constructs a landslide prediction model based on slope stability assessment data. First, the slope stability assessment data and historical landslide event data are integrated as model input data. A multi-factor analysis method is used to establish a landslide risk factor model by analyzing the correlation between slope displacement changes, environmental factors, and historical landslide events. A classification algorithm is used to weight environmental factors influencing landslides and train a risk prediction model to establish the landslide prediction model. During this process, the slope stability assessment data is spatially overlaid with other environmental factor data using a GIS platform to identify high-risk areas for landslides. Model training uses known landslide events as training samples and, combined with standardized remote sensing datasets, further optimizes the model using regression analysis or neural network methods. Model validation cross-validates prediction results to ensure the model's applicability across different geographic regions and environmental conditions. Landslide warnings are issued based on the landslide prediction model and combined with real-time meteorological data. When the prediction model identifies a high-risk area and reaches the warning threshold, landslide warning information data is generated and displayed in real time on a map visualization platform, including the area's landslide risk level, warning range, and potential landslide time window.

[0019] Step S7: Use the landslide warning information data to conduct risk assessment of the target difficult mountainous area, and formulate a prevention strategy based on the risk assessment results to obtain a landslide warning and prevention plan.

[0020] The embodiment of the present invention utilizes landslide warning information data to conduct a risk assessment for a target difficult and dangerous mountainous area. First, the landslide warning information data is comprehensively analyzed with regional geographic information data, environmental factor data, and historical disaster data. Spatial analysis is performed using a GIS platform, and the high-risk areas marked in the landslide warning information data are combined to identify potential danger zones for landslides. Next, based on the landslide prediction results and the environmental factors of each region, a risk level classification is performed, dividing the target difficult and dangerous mountainous area into different risk areas, such as high-risk, medium-risk, and low-risk areas. To ensure the accuracy of the assessment, a multi-factor weighted analysis method is used, combining the spatial distribution of environmental factors, assigning different weights to each factor, and generating a landslide risk level map for each region through a weighted comprehensive assessment method. On this basis, a risk assessment report is further generated, providing a detailed description of the high-risk areas and proposing corresponding landslide prevention strategies. Specific strategies include, but are not limited to, strengthening vegetation coverage in high-risk areas, building drainage systems, improving soil stability, and monitoring precipitation changes. Furthermore, the assessment process must consider the distribution of infrastructure, population density, and historical landslide frequency in the target difficult mountainous area. This will lead to the development of an emergency response plan, including a landslide warning information release mechanism and evacuation route maps. Ultimately, a landslide warning and prevention plan will be generated.

[0021] Preferably, step S1 includes the following steps: Step S11: Acquire synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data of the target difficult mountainous area, and record the synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data as the original remote sensing data set; Step S12: performing data preprocessing on the original remote sensing data set to obtain a preprocessed remote sensing data set; Step S13: performing standardization processing on the pre-processed remote sensing data set to obtain a standardized remote sensing data set.

[0022] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in the embodiment of the present invention, step S1 includes the following steps: Step S11: Acquire synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data of the target difficult mountainous area, and record the synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data as the original remote sensing data set; The embodiment of the present invention first obtains multi-source remote sensing data from the target rugged mountainous area, including synthetic aperture radar (SAR) image data, laser radar (LiDAR) point cloud data, drone image data, and optical remote sensing image data. For SAR image data, the synthetic aperture radar's echo signal is used to obtain information on subtle changes in the ground surface, and an interferometric image is generated through interferometric processing to obtain terrain and displacement change data. LiDAR point cloud data uses laser scanning technology to obtain high-precision three-dimensional terrain information. Point cloud classification methods are used to classify the point cloud data into ground points and non-ground points, providing a basis for subsequent three-dimensional terrain modeling. Drone image data obtains high-definition images of the target area through low-altitude flight, and performs image matching and stereo matching to obtain high-precision three-dimensional surface information. Optical remote sensing image data provides visual information about the ground surface and obtains different surface features through reflectance spectra. When merging data, the spatial resolution of different data sources is first unified, and the images and point cloud data from different sources are aligned through image registration technology to ensure spatial consistency between different data sets. Subsequently, all data were spatially overlaid using a geographic information system (GIS) platform to ensure accurate fusion of different data sets and obtain a comprehensive original remote sensing data set.

[0023] Step S12: performing data preprocessing on the original remote sensing data set to obtain a preprocessed remote sensing data set; The embodiment of the present invention preprocesses the original remote sensing dataset, first denoising different types of data. For synthetic aperture radar (SAR) image data, Gaussian filtering or median filtering is used to remove noise from the image, improving image quality and enhancing the clarity of target information. For laser radar (LiDAR) point cloud data, a filtering operation is performed to remove void points, noise points, and non-ground points, and the RANSAC algorithm is used to extract ground points to ensure the accuracy of terrain information. Distortion and artifacts contained in drone image data are processed through image geometric correction, and then compared with a reference image using an image registration algorithm to correct perspective distortion and ensure the geometric accuracy of objects in the image. Optical remote sensing image data undergoes atmospheric correction to eliminate the influence of atmospheric effects, and the 6S radiation transfer model is used to estimate atmospheric transmittance, further improving data accuracy. Before data merging, image registration is performed using spatial alignment technology to ensure the consistency of spatial positions between different data sources. All types of data are uniformly projected using the same geographic coordinate system and adjusted to a uniform scale based on the spatial resolution of the data. In the data fusion process, a multi-resolution fusion method is used to perform weighted fusion on various types of data, retaining the characteristic information of each data and minimizing redundant information to obtain a preprocessed remote sensing dataset.

[0024] Step S13: performing standardization processing on the pre-processed remote sensing data set to obtain a standardized remote sensing data set.

[0025] Embodiments of the present invention standardize preprocessed remote sensing datasets to eliminate scale differences between different data sources and ensure data consistency. First, for synthetic aperture radar image data, data normalization methods are used to convert its amplitude values ​​into dimensionless values, typically using minimum-maximum normalization or Z-score normalization, to ensure uniform brightness values ​​across different regions within the image. Simultaneously, lidar point cloud data is normalized to unify the point cloud height values ​​to a common reference plane, often using orthographic projection to eliminate height differences between different lidar datasets. Unmanned aerial vehicle image data undergoes geometric transformation to unify the spatial resolution of the images. Resampling methods adjust the spatial resolution of the images to the finest resolution of the target area, aligning them with other data sources. During the standardization process for optical remote sensing image data, radiometric normalization is performed to convert the radiometric intensity of image data acquired at different times and by different sensors to a common reference standard to eliminate the impact of environmental factors on the images. All processed datasets are reprojected using a unified coordinate system to ensure spatial consistency and are normalized using the maximum and minimum values ​​to ensure that all data fall within the same numerical range. This improves the integration of different datasets and the accuracy of analysis, resulting in a standardized remote sensing dataset.

[0026] By acquiring and combining multiple remote sensing data, this method comprehensively covers the diverse terrain and environmental characteristics of difficult mountainous areas, providing a rich foundation for subsequent modeling. Data preprocessing helps eliminate noise and errors, improves data quality, and ensures the accuracy and reliability of subsequent analysis. Standardization ensures data uniformity, enabling comparison and analysis of data from different sources under the same standards, enhancing the operability of the modeling process.

[0027] Preferably, step S2 includes the following steps: Step S21: performing image partitioning processing of the target difficult mountainous area based on the standardized remote sensing data set, and performing preliminary surface coverage area extraction to obtain preliminary surface area division data of the mountainous area; Step S22: performing optical image spectrum feature analysis on the preliminary surface area division data of the mountainous area to obtain vegetation feature data of the mountainous area; Step S23: extracting the terrain characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain the mountainous area terrain factor characteristic data; Step S24: extracting water characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain water characteristics data of the mountainous area; Step S25: Analyze the surface roughness characteristics of the target difficult mountainous area using the standardized remote sensing data set to obtain the exposed surface characteristic data of the mountainous area; Step S26: extracting vegetation height and building height information in the target difficult mountainous area using the standardized remote sensing data set to obtain three-dimensional height feature data of the mountainous area; Step S27: performing a refined extraction of landslide-prone slope areas based on the mountainous terrain factor characteristic data and the mountainous water body characteristic data to obtain mountainous slope characteristic data; Step S28: fusing the mountain vegetation feature data, mountain water feature data, mountain exposed surface feature data, mountain three-dimensional height feature data, and mountain slope feature data to obtain surface feature data in difficult mountainous areas.

[0028] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 1 Detailed step flow diagram of step S2 in the embodiment of the present invention, step S2 includes the following steps: Step S21: performing image partitioning processing of the target difficult mountainous area based on the standardized remote sensing data set, and performing preliminary surface coverage area extraction to obtain preliminary surface area division data of the mountainous area; This embodiment of the present invention uses standardized remote sensing datasets to perform image zoning processing on target difficult mountainous areas. First, the remote sensing images are zoned using an image classification algorithm. A pixel-based cluster analysis method, using a K-means clustering algorithm or a support vector machine (SVM), is used to perform preliminary zoning of the target area, dividing the mountain image data into multiple surface cover regions. Next, the band information of the multispectral image is used to extract surface cover types. NDVI (Normalized Difference Vegetation Index) analysis is used to identify vegetation areas. NDWI (Water Body Index) analysis is used to extract water areas. The Soil Brightness Index (SAVI) is combined to further identify exposed surface areas. Finally, the spatial characteristics of the remote sensing images and an image segmentation algorithm are used to segment different surface types. Each surface cover area is labeled and classified to obtain preliminary surface zoning data for the mountainous area. Furthermore, GIS tools are used to accurately locate the spatial coordinates of each segmented region and mark them on a map to ensure accurate alignment of the segmented data with the geographic coordinate system.

[0029] Step S22: performing optical image spectrum feature analysis on the preliminary surface area division data of the mountainous area to obtain vegetation feature data of the mountainous area; This embodiment of the present invention performs optical image spectral feature analysis on preliminary surface area delineation data for mountainous areas. First, optical image data from a standardized remote sensing dataset is selected and spectral index analysis is used to extract the reflectance of each image band. Vegetation areas are analyzed using the Normalized Difference Vegetation Index (NDVI), which quantifies vegetation growth by the difference in reflectance between the red and near-infrared bands, thereby extracting vegetation characteristic data for the mountainous area. Next, water areas are analyzed using the Water Body Index (NDWI), which identifies water areas by the difference in reflectance between the green and near-infrared bands. Furthermore, the Soil Brightness Index (SAVI) is used to further analyze soil and exposed surface areas. The calculation process for all these indices is based on the image's spectral band data, combined with the image's spatial resolution and spectral characteristics, and the corresponding characteristic data is generated through rasterization. During the analysis process, surface cover types are extracted using the spectral reflectance information from the remote sensing image to obtain vegetation characteristic data for the mountainous area.

[0030] Step S23: extracting the terrain characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain the mountainous area terrain factor characteristic data; When extracting the terrain undulation characteristics of a target difficult mountainous area based on a standardized remote sensing dataset, the embodiment of the present invention first obtains the elevation data in the standardized remote sensing dataset and uses a digital elevation model (DEM) to analyze the terrain undulation of the target difficult mountainous area. A slope analysis method is used to describe the degree of inclination of the surface by calculating the slope value of each pixel point in the DEM, thereby extracting the undulation characteristics of the mountainous terrain. Next, the DEM is used to perform terrain undulation analysis, calculate key terrain features, and further identify the complexity and irregularity of the surface. In addition, the terrain wetness index (TWI) is used to analyze the terrain hydrological characteristics, and the slope and elevation data are combined to identify the water flow convergence areas and potential landslide-prone areas in the mountainous area. These analysis results are rasterized to convert the terrain factor values ​​of each pixel point into feature data, forming mountainous terrain factor feature data.

[0031] Step S24: extracting water characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain water characteristics data of the mountainous area; In this embodiment of the present invention, water features in a target rugged mountainous area are extracted based on a standardized remote sensing dataset. High-resolution optical imagery and synthetic aperture radar (SAR) imagery are first extracted from the dataset. These imagery data undergoes preprocessing to remove cloud and atmospheric influences to ensure image quality. Next, water indices (such as the Normalized Difference Water Index (NDWI)) are applied to the imagery. Areas with higher NDWI values ​​indicate areas of water. This process utilizes the spectral characteristics of remote sensing images and analyzes the reflectivity of different bands to accurately extract water features in the mountainous area. Furthermore, elevation information from LiDAR point cloud data can be used to identify changes in water depth and the influence of surrounding terrain. Water feature extraction includes not only water surface distribution but also water morphology, area, and the interface between the water area and the surrounding terrain. Finally, through rasterization, the extracted water feature data is mapped and integrated into the overall mountainous surface feature dataset to produce mountainous water feature data.

[0032] Step S25: Analyze the surface roughness characteristics of the target difficult mountainous area using the standardized remote sensing data set to obtain the exposed surface characteristic data of the mountainous area; In this embodiment of the present invention, surface roughness analysis in a target mountainous area based on a standardized remote sensing dataset begins by fusing LiDAR point cloud data with remote sensing imagery. LiDAR point cloud data provides precise surface elevation information, which is used to identify surface fluctuations, particularly the roughness of exposed surface areas. A surface roughness index (such as a coefficient of variation based on terrain elevation changes or a local roughness calculation method) is then used to analyze the target mountainous area's surface. This involves calculating the elevation difference for each pixel, thereby generating a roughness map. Filters of varying scales are used to smooth the data for different surface types, eliminating noise and highlighting the primary surface roughness features. This process also incorporates the reflectance characteristics of optical remote sensing images, analyzing changes in reflectivity across different wavelengths (such as red and near-infrared) to further confirm the presence of exposed surface areas. Roughness analysis can accurately identify exposed areas with high reflectivity and a lack of significant vegetation or water. Finally, the surface roughness characteristic data are rasterized and integrated into the entire surface characteristic dataset of the difficult mountainous area to obtain the exposed surface characteristic data of the mountainous area.

[0033] Step S26: extracting vegetation height and building height information in the target difficult mountainous area using the standardized remote sensing data set to obtain three-dimensional height feature data of the mountainous area; This embodiment of the present invention uses standardized remote sensing datasets to extract vegetation and building height information in challenging mountainous areas. The image resolution of the acquired optical remote sensing imagery is first optimized to ensure full detail. For vegetation height extraction, the Normalized Difference Vegetation Index (NDVI) is applied to initially partition the imagery to identify vegetation areas. LiDAR point cloud data is then used to extract height information for these areas. The vertical structure of the vegetation is analyzed based on point cloud reflectance intensity and elevation changes. Building height information is obtained by identifying and classifying artificial structures in the imagery and analyzing the LiDAR point cloud data. The height difference between the highest point of a building and a reference surface is then extracted to calculate the vertical height of the building. For complex mountainous environments, a local regression algorithm is used to process the elevation data to further improve the accuracy of vegetation and building height extraction. This process also uses multi-source data fusion to fuse the optical imagery and LiDAR data, ultimately generating three-dimensional height feature data for the mountainous area.

[0034] Step S27: performing a refined extraction of landslide-prone slope areas based on the mountainous terrain factor characteristic data and the mountainous water body characteristic data to obtain mountainous slope characteristic data; In this embodiment of the present invention, when fine-tuning landslide-prone slope areas based on terrain factor characteristic data and mountain water feature data, terrain factor data extracted from standardized remote sensing datasets is first used as the analysis basis for topographic analysis of the mountainous area using a digital elevation model (DEM). Combined with water feature data, data on water system distribution, waterlogged areas, and water flow direction are used to further identify potential areas of soil and water loss and their relationship to topographic changes. Next, a landslide susceptibility assessment model is established, combining terrain factor and water feature data for overlay analysis to identify slope areas prone to landslides. For complex mountainous environments, a weighted average method is used to fuse different feature data to ensure the accuracy of slope feature extraction. During this process, spatial analysis methods and region-based analysis algorithms, such as local regression analysis or spatial autocorrelation analysis, are used to further refine the boundaries of landslide areas and demarcate high-risk areas within the areas. Ultimately, characteristic data for mountain slopes is extracted.

[0035] Step S28: fusing the mountain vegetation feature data, mountain water feature data, mountain exposed surface feature data, mountain three-dimensional height feature data, and mountain slope feature data to obtain surface feature data in difficult mountainous areas.

[0036] When fusing mountain surface feature data, the present invention first extracts key feature information from mountain vegetation feature data, mountain water feature data, mountain exposed surface feature data, mountain three-dimensional height feature data, and mountain slope feature data. Spatial analysis methods are used to fuse these different feature data using a weighted average approach. Combined with multi-level decision analysis, the features from different data sources are weighted and superimposed based on their importance and spatial distribution. Furthermore, principal component analysis (PCA) is applied to reduce the dimensionality of these feature data to extract the key features reflecting the mountain surface structure. Next, geographic information system (GIS) tools are used to spatially register and refine these fused data to ensure spatial consistency between the data layers. After data fusion, comprehensive surface feature data for difficult mountainous areas is generated.

[0037] The present invention provides clear surface area division for subsequent analysis through image partitioning and preliminary surface cover area extraction, ensuring the efficiency and accuracy of data processing. Spectral feature analysis helps identify vegetation types and distribution in mountainous areas, improving the precision and accuracy of vegetation feature extraction. Terrain undulation feature extraction provides data support for the quantitative description of mountainous terrain features, enhancing the adaptability of the model to complex mountainous terrain. Water body feature extraction provides key data for the analysis of water resources and water body distribution in mountainous areas, assisting in disaster risk assessment. Surface roughness feature analysis helps identify exposed surface areas in mountainous areas, providing basic data for surface change monitoring. Vegetation height and building height extraction provide detailed data for the establishment of three-dimensional spatial features in mountainous areas, supporting refined modeling. Refined extraction of landslide-prone slope areas provides high-precision terrain information for landslide risk analysis and optimizes slope stability assessment. Data fusion processing combines multi-source feature data to provide comprehensive mountain surface features, providing a reliable foundation for comprehensive analysis and model construction.

[0038] Preferably, step S3 includes the following steps: Step S31: extracting surface elevation characteristics based on the surface characteristic data of the difficult mountainous area to obtain terrain elevation characteristic data; This embodiment of the present invention extracts surface elevation characteristics based on surface feature data in rugged mountainous areas. First, elevation data for the target area is acquired through remote sensing data, including elevation data acquired using LiDAR (Light Detection and Ranging) and surface information extracted from aerial or satellite imagery. For rugged mountainous areas, particular attention is paid to terrain features with drastic elevation variations. This acquired elevation data is processed using specialized Geographic Information System (GIS) tools to accurately match surface elevation values ​​with their corresponding spatial locations. Then, through differentiated surface feature analysis, the elevation data in the mountainous area is correlated with different types of surface features to extract elevation feature data within the area. During this process, surface heights are extracted using a Digital Elevation Model (DEM), generating elevation feature data for different terrain regions (such as ridges, valleys, and slopes). Next, a filtering algorithm is applied to the elevation data to denoise it, removing anomalies caused by weather or sensor errors to ensure the accuracy of the elevation data. Ultimately, terrain elevation feature data is obtained.

[0039] Step S32: performing three-dimensional gridding processing on the terrain elevation feature data and performing preliminary terrain grid construction to obtain three-dimensional terrain grid data; This embodiment of the present invention performs three-dimensional gridding on terrain elevation feature data. First, the elevation values ​​of all surface points are extracted from the terrain elevation feature data. This data includes the elevation value corresponding to each pixel extracted from remote sensing imagery and elevation data acquired through LiDAR (Light Detection and Ranging). This elevation data is then gridded using GIS tools to generate a regular three-dimensional grid model. The terrain elevation feature data is first divided into several fixed-size grid cells, each corresponding to a specific spatial range. Next, the original data points are mapped to grid nodes using interpolation methods (such as bilinear interpolation or kriging interpolation) to achieve a smooth transition and fill gaps between elevation data, ensuring uniform distribution of terrain data across the entire area. This process requires special processing for complex terrain, particularly in mountainous areas, especially in areas with steep slopes, to avoid loss of terrain information due to over-smoothing. This processing step generates three-dimensional terrain grid data.

[0040] Step S33: performing elevation adjustment on the three-dimensional terrain grid data based on the three-dimensional height characteristic data of the mountain area to obtain refined terrain grid data; This embodiment of the present invention performs elevation adjustment on three-dimensional terrain grid data. First, based on preliminary three-dimensional terrain grid data, three-dimensional elevation feature data for mountainous areas is aligned with the grid data to ensure terrain accuracy. Missing elevation values ​​in the grid are filled using an interpolation algorithm based on known elevation points (such as inverse distance weighted interpolation or kriging interpolation). This process specifically considers the unique undulations and steep slopes of mountainous areas and their impact on data accuracy. The adjustment method selects appropriate interpolation parameters based on the terrain type to reduce errors caused by terrain complexity. Next, accuracy is optimized based on the acquired elevation data, and the elevation values ​​of the grid nodes are corrected to better reflect the slope variations of the actual terrain. This process utilizes terrain gradient analysis and slope calculation methods, combined with field elevation data and remote sensing data, to further precisely adjust the elevation values ​​of specific areas. For example, in high mountainous or steep slope areas, fine-tuning is performed by analyzing the elevation change rate of the area to ensure that the adjusted terrain grid more closely matches the actual three-dimensional structure of the mountainous area. After the adjustment is completed, refined terrain grid data is generated.

[0041] Step S34: performing surface feature enhancement processing on the refined terrain grid data based on the mountainous area vegetation feature data and the mountainous area exposed surface feature data to obtain textured terrain grid data; This embodiment of the present invention enhances the surface features of refined terrain mesh data based on mountain vegetation and exposed surface data. This process first extracts mountain vegetation information and exposed surface features from remote sensing image data. This process is achieved through classification of remote sensing images. The Normalized Difference Vegetation Index (NDVI) algorithm is used to extract vegetation features and determine vegetation coverage in different mountainous areas. Reflectance data and other surface reflectance characteristics are then used to distinguish exposed and vegetated areas. Next, the refined terrain mesh data is processed, and surface feature enhancement is performed on both vegetated and exposed surface areas. For vegetated areas, texture mapping technology is used to fuse the extracted vegetation feature data with the terrain mesh data. This texturing process creates a more realistic representation of the vegetated areas. For exposed surface areas, surface reflectance characteristics are used to identify surface types, and an enhancement algorithm is used to enhance the details of the exposed areas. After surface feature enhancement, textured terrain mesh data is generated.

[0042] Step S35: using the mountain water body characteristic data to perform water body range correction on the textured terrain grid data to obtain water body corrected terrain grid data; This embodiment of the present invention corrects the water extent of textured terrain grid data based on water feature data in mountainous areas. First, the reflectance characteristics of water bodies within mountainous areas must be acquired through remote sensing imagery. A water index (such as MNDWI) algorithm is then used to process the remote sensing imagery and extract the water area. This water feature data is combined with existing textured terrain grid data to identify the location and extent of the water within the terrain. Next, spatial analysis techniques are used to correct the textured terrain grid data based on the spatial location and area of ​​the water body, precisely delineating the boundaries of the water body. This correction is accomplished by remapping the terrain grid data, refining the topographic features of the water area and separating it from other areas. The corrected water extent adjusts the elevation values ​​and texture information in the terrain grid based on the actual shape of the water body, resulting in a more accurate representation of the water's shape and position in the three-dimensional model. The resulting correction is the water-corrected terrain grid data.

[0043] Step S36: performing high-precision slope area refinement on the water body corrected terrain grid data to obtain high-precision slope terrain grid data; This embodiment of the present invention uses water-corrected terrain grid data to refine slope regions with high precision. First, slope features extracted from remote sensing imagery and terrain data are combined with the water-corrected terrain grid data to identify the topographic features of the slope region. Spatial analysis methods are used to refine the elevation of the slope region. Elevation measurement points are added within the slope region, and interpolation algorithms (such as Kriging or inverse distance weighted interpolation) are used to refine the slope elevation data, ensuring that every detail is accurately captured. For terrain variations on steep slopes, high-precision point cloud data (such as LiDAR data) is used to further optimize the elevation, improving the grid density and elevation accuracy of the slope region. The slope refinement process also adjusts the grid details based on different terrain features (such as rock, exposed soil, and vegetation cover), making the slope morphology of each region more realistic and detailed. Ultimately, high-precision slope terrain grid data is generated.

[0044] Step S37: Perform multi-perspective analysis on the high-precision slope terrain grid data and verify its accuracy to generate preliminary three-dimensional terrain model data.

[0045] This embodiment of the present invention uses multiple remote sensing data sources (such as high-resolution satellite imagery and aerial lidar data) to obtain terrain information of mountainous areas from different perspectives. This data is then registered to ensure spatial consistency and combined with high-precision slope terrain grid data for detailed analysis. The multi-perspective analysis method involves comparing terrain observation data from different directions to reveal hidden features of the slope area. Next, error analysis methods are used to verify accuracy based on a comparison of field measurement data with remote sensing data. Statistical methods are used to assess the accuracy of the three-dimensional terrain model. After accuracy verification, the model is combined with refined slope feature data to generate preliminary three-dimensional terrain model data that meets accuracy requirements.

[0046] The surface elevation feature extraction of the present invention provides accurate data support for terrain elevation analysis and enhances the recognition accuracy of terrain features. A preliminary terrain grid is constructed through three-dimensional gridding processing, laying the foundation for the subsequent refinement of the terrain model. Elevation adjustment improves the accuracy of the terrain grid, making the three-dimensional terrain model more consistent with the actual terrain features. Surface feature enhancement makes the texture of the terrain grid more realistic, reflecting the actual conditions of vegetation and exposed surfaces in mountainous areas. Water body range correction ensures the accuracy of water body areas in the terrain grid and enhances the authenticity of water body features. High-precision regional refinement of slopes provides more refined terrain data for landslide risk analysis and improves the reliability of terrain analysis. Multi-perspective analysis and accuracy verification optimize the three-dimensional terrain model to ensure the accuracy and practicality of the final model.

[0047] Preferably, step S4 includes the following steps: Step S41: extracting multi-temporal image data from the standardized remote sensing data set and sorting them in time sequence to obtain time series image data; The embodiment of the present invention first extracts multi-temporal image data from a standardized remote sensing dataset, extracts a dataset of a specific time point from a remote sensing image data source, and ensures that the time span of the data source covers the dynamic change period of the study area. During extraction, the image data is sorted according to the acquisition time in chronological order to ensure that the image data is arranged in chronological order and generate time series image data. Remote sensing image preprocessing tools are used to perform geometric correction and radiation correction on the original image data to remove noise from the image data and improve the quality of the image to ensure the accuracy of the data. In the data sorting link, the images are sorted in the order of acquisition time by calculating the timestamp of the image or using the acquisition time indicated in the metadata, and continuous time series image data are formed.

[0048] Step S42: performing change detection analysis on the time series image data and identifying the surface dynamic change areas to obtain surface change situation data; The embodiment of the present invention first performs change detection analysis on the time series image data. Using a change detection method based on spectral information, by comparing the image data of consecutive time points at the pixel level, the change information of the surface in different time periods is extracted. The change detection method determines the change of each pixel point in different periods by calculating the spectral difference of each pixel or an index based on the difference. In order to improve the analysis accuracy, the time series image data is firstly registered to ensure the spatial alignment of the image data at different time points. During the change detection analysis, a suitable threshold is selected to distinguish between the changing area and the stable area, and the difference index or change map of the image is used to clarify the dynamic changes of the area. Then, based on the results of the change detection, the dynamic change area of ​​the surface is identified, and according to the spatial distribution of the change area, the area where the surface has changed significantly is extracted and marked to form the surface change situation data.

[0049] Step S43: performing slope landform evolution regional analysis on the preliminary three-dimensional terrain model data based on the surface change situation data to obtain slope change hotspot data; The embodiment of the present invention first performs a regional analysis of slope landform evolution on preliminary three-dimensional terrain model data based on surface change situation data. The preliminary three-dimensional terrain model data is subjected to a refined analysis, and combined with the surface change situation data, the dynamic evolution characteristics of the slope area are identified. First, a three-dimensional terrain model is constructed using high-precision remote sensing data, and a preliminary three-dimensional terrain model of the slope is generated using a digital elevation model (DEM) and point cloud data. Next, by comparing the surface change situation data at different time points, the change trend of the slope area is analyzed, and areas with obvious landform evolution characteristics, especially the dynamic change areas of the slope, are identified. In this process, terrain analysis tools are used to extract the slope and aspect information of the slope area, and the stability and evolution laws of the slope are further analyzed. Using the extracted surface change situation data, these areas are calibrated to obtain slope change hotspot data.

[0050] Step S44: performing refined displacement monitoring on the slope change hotspot data based on the time series image data to obtain slope displacement monitoring data; The embodiment of the present invention first performs refined displacement monitoring on slope change hotspot data based on time series image data. The identified slope change hotspot area is further analyzed, and pixel-level displacement information in the area is extracted using high-precision remote sensing image data. During the operation, the time series image data is first registered to ensure that the spatial positions of image data in different time periods are accurately aligned to avoid image errors affecting the monitoring results. Then, the image data of different time periods are compared through the differential imaging method to identify the displacement occurring in the slope change hotspot area. Using the difference index or other image change detection algorithms, the spatial distribution and size information of the displacement are extracted by comparing the changes in image pixel values. In order to accurately monitor slope displacement, sub-pixel displacement detection technology is adopted, and high-precision calculation of image details is used to ensure that the detection results have higher spatial accuracy. Finally, based on the monitoring results, slope displacement monitoring data is generated.

[0051] Step S45: performing dynamic displacement characteristic analysis on the slope displacement monitoring data, and extracting the slope displacement rate and direction to obtain slope dynamic displacement characteristic data; The embodiment of the present invention first performs a dynamic displacement characteristic analysis on the slope displacement monitoring data. Using the slope displacement monitoring data extracted from the time series image data, the rate and direction of the slope displacement are identified through statistical and analytical methods. First, the displacement amount between different periods is calculated using the difference method using the acquired displacement data, and then the displacement rate of the slope is extracted. For the displacement direction, vector analysis technology is used, combined with the slope direction and slope information of the slope area, to determine the specific direction of the displacement. By performing a time series analysis on the displacement monitoring data, the changing trend of the slope displacement is identified, and the law of the dynamic change of the slope is further revealed. In order to ensure the accuracy of the analysis results, high-precision image registration technology is selected to accurately align the image data at different time points to avoid errors caused by image offset. The extraction of displacement rate and direction in this process can accurately reflect the stability and dynamic evolution trend of the slope in different time periods. Ultimately, the dynamic displacement characteristic data of the slope is obtained.

[0052] Step S46: performing multi-temporal surface deformation comparative analysis on the time series image data based on the slope dynamic displacement characteristic data to obtain surface deformation evolution data; The embodiment of the present invention performs a multi-phase surface deformation comparative analysis on time series image data based on the dynamic displacement characteristic data of the slope. First, the dynamic displacement characteristic data of the slope is combined with the time series image data to perform a comparative analysis on the deformation changes of the slope area. First, image registration processing is performed to ensure the precise alignment of the time series image data in spatial position. Then, multiple time nodes are selected to compare the image data of different time periods, identify the deformation trend of the slope area, and analyze the ups and downs of the slope surface. By comparing the differences between different time periods, the surface deformation information of the slope in each time period is obtained. In order to ensure the accuracy of the analysis, the difference index method or the pixel difference analysis method is used to calculate the image changes in different periods, and then the deformation data of the displacement area is extracted. In this process, the direction and speed of the slope deformation are analyzed in combination with terrain features such as slope and aspect, and the surface deformation evolution data is generated.

[0053] Step S47: Perform spatial interpolation processing on the surface deformation evolution data to obtain slope displacement change data.

[0054] The embodiment of the present invention first performs spatial interpolation processing on the surface deformation evolution data to obtain slope displacement change data. A suitable interpolation method, such as inverse distance weighted interpolation (IDW) or Kriging interpolation, is selected to process the surface deformation evolution data. The operation process first performs spatial analysis on the observation points in the surface deformation evolution data, and performs interpolation calculations based on their spatial positions and deformation amplitudes, thereby generating a displacement change map for the entire slope area. During the interpolation process, the data is weighted, taking into account the topographic characteristics of the slope, so that the interpolation result is more accurate. Then, the interpolated data is used to fill in the areas that have not been collected in the surface deformation evolution data to generate continuous slope displacement change data.

[0055] The multi-temporal image data extraction and sorting of the present invention provides a systematic image data foundation for subsequent time series analysis. Change detection analysis reveals the dynamic change areas of the surface, which helps to identify the change areas that affect the stability of the slope. The regional analysis of the slope landform evolution reveals the hot spots of slope changes, providing key area data for further monitoring and prediction. Refined displacement monitoring improves the accuracy of slope displacement detection and ensures detailed monitoring of the change area. Dynamic displacement characteristic analysis provides the rate and direction of slope displacement, providing key dynamic parameters for slope stability assessment. Surface deformation comparison analysis reveals surface changes in time series and enhances the understanding of slope evolution trends. Spatial interpolation processing provides refined slope displacement change data, providing high-precision input data for subsequent model construction and prediction.

[0056] Preferably, step S5 includes the following steps: Step S51: extracting key dynamic parameters from the slope displacement change data to obtain key slope dynamic parameter data; The embodiment of the present invention first extracts key dynamic parameters from the slope displacement change data. The existing slope displacement change data is analyzed to extract key dynamic parameters that reflect the slope change trend. During the operation, the slope displacement data is first analyzed in time series to calculate the displacement rate and displacement amplitude within each time period. Then, a filtering method is used to remove noise data to ensure that the extracted dynamic parameters accurately reflect the actual changes in the slope. For each time period, the displacement values ​​of different points in the slope area are calculated, and then the key displacement change areas are identified. The dynamic changes in these areas will serve as an important reference for slope stability assessment. Next, based on the displacement data, the dynamic phenomena of the slope are further analyzed, and representative dynamic parameters are extracted to form key slope dynamic parameter data.

[0057] Step S52: extracting meteorological factors, geological factors, and hydrological factors from the standardized remote sensing data set to obtain a mountainous area environmental factor data set; The embodiment of the present invention first extracts meteorological, geological, and hydrological factors from a standardized remote sensing dataset to construct a mountainous area environmental factor dataset. Meteorological factors are extracted by using temperature, precipitation, and humidity information from remote sensing images, supplemented by meteorological remote sensing products or ground-based meteorological station data, to generate a meteorological factor data layer for the corresponding region. Geological factor extraction relies on information reflecting geological features in remote sensing images. These features can be analyzed using high-resolution remote sensing images and digital elevation model (DEM) data to generate a geological factor layer. Hydrological factor extraction analyzes regional hydrological conditions based on information about water body distribution, flow direction, and flow rate in remote sensing images, combined with hydrological models, to generate a hydrological factor data layer. Through comprehensive analysis of these data layers, a complete mountainous area environmental factor dataset is constructed.

[0058] Step S53: performing correlation analysis on the key slope dynamic parameter data and the mountainous area environmental factor data set, and identifying the dominant slope dynamic change environmental factors to obtain the dominant environmental factor data; The embodiment of the present invention first performs a correlation analysis on the key slope dynamic parameter data and the mountain environmental factor data set. This analysis evaluates the strength of the relationship between each environmental factor (such as meteorological, geological and hydrological factors) and the slope dynamic parameters by calculating the Pearson correlation coefficient or the Spearman rank correlation coefficient. Through these correlation analyses, it is possible to determine which environmental factors have a significant impact on the dynamic changes of the slope and identify the environmental factors that dominate the dynamic changes of the slope. First, the dynamic parameters in the slope displacement data set are extracted, and then compared one by one with the meteorological, geological and hydrological factor data integrated by remote sensing data. Correlation analysis is performed using statistical analysis methods such as regression analysis to obtain the degree of influence of each factor on the slope displacement change. Based on the analysis results, the dominant environmental factors that are most closely related to the slope displacement change are screened out to form the dominant environmental factor data.

[0059] Step S54: performing weight distribution analysis on the dominant environmental factor data to obtain environmental factor weight distribution data; The embodiment of the present invention first performs a weight distribution analysis based on the dominant environmental factor data. This analysis uses the analytic hierarchy process (AHP) or grey correlation analysis method to quantify the influence of each dominant environmental factor as a weight value. First, an evaluation model that includes all dominant environmental factors is established, and the impact measurement standard of each factor is set. Then, the relative importance of each factor is evaluated through expert scoring or statistical methods, and the results are converted into numerical weights. To this end, a standardized processing method is adopted to adjust the influence of each factor to a unified standard scale. This process also needs to be combined with the actual geological environment characteristics, such as the differences in the impact of meteorological factors and hydrological factors on slope stability under different climatic conditions, to ensure the rationality and scientific nature of the analysis results. Finally, the environmental factor weight distribution data is obtained.

[0060] Step S55: Perform slope stability assessment based on the environmental factor weight distribution data and key slope dynamic parameter data to obtain slope stability assessment data.

[0061] The embodiment of the present invention first constructs a slope stability assessment model based on the combination of environmental factor weight distribution data and key slope dynamic parameter data. The model uses a weighted average method or a multiple linear regression method to quantify the weights of environmental factors and the corresponding slope dynamic parameters to obtain a stability score for each slope location. The weight data and the dynamic parameter data are aligned according to geographic coordinates to form a comprehensive assessment index for each monitoring point. Then, a GIS platform is used to perform spatial analysis on the slope topography and the distribution of various factors, and the slope stability level of each area is obtained by combining the dynamic displacement data of the slope and historical landslide records. In order to further improve the assessment accuracy, the dynamic change trend of the slope is modeled in combination with time series data, and early warning identification is performed on areas where landslides or collapses occur. Finally, the slope stability assessment data is output.

[0062] The extraction of key dynamic parameters of the present invention ensures that the core characteristics of slope displacement changes are accurately captured, providing accurate dynamic parameter data for subsequent analysis. The extraction of meteorological factors, geological factors and hydrological factors provides multi-dimensional data support for the comprehensive analysis of mountain environment characteristics. Correlation analysis and identification of dominant environmental factors help determine the environmental factors that have the greatest impact on the dynamic changes of slopes, optimizing the pertinence of subsequent analysis. The environmental factor weight distribution analysis reveals the relative importance of each environmental factor in the slope stability assessment, providing quantitative weight data for the assessment. Slope stability assessment based on environmental factor weights and dynamic parameter data provides more accurate analysis results of slope stability and supports effective risk prediction.

[0063] Preferably, step S6 includes the following steps: Step S61: Screening high-risk areas in the target difficult mountainous area based on the slope stability assessment data to obtain high-risk area distribution data; The embodiment of the present invention first uses a spatial analysis method to screen high-risk areas based on the stability scores of each area obtained from the slope stability assessment data. The screening process sets a stability threshold to mark those areas with stability scores lower than a predetermined value as high-risk areas. The stability scores of each monitoring point are spatially analyzed through the GIS platform to generate a stability assessment layer, and the threshold setting is applied to partition the layer to distinguish high-risk areas from low-risk areas. By combining with terrain data, misjudgments caused by the particularity of the terrain are further eliminated to ensure that the screened areas have a higher potential for landslides. Next, remote sensing image data is used to monitor the changing trends of high-risk areas, and combined with data from historical landslide events, the distribution of high-risk areas is confirmed, and the distribution data of high-risk areas is output.

[0064] Step S62: performing time series trend analysis on the high-risk area distribution data and identifying landslide evolution characteristics to obtain landslide trend data; The embodiment of the present invention first performs a time series analysis on each area based on the distribution data of high-risk areas, focusing on analyzing the stability change trend within the area. Remote sensing image data and historical slope deformation data are used to regularly monitor the time series changes in each high-risk area, and by comparing images at different time nodes, the displacement, cracks and deformation characteristics of the surface are evaluated. Trend analysis methods, such as linear regression or time series analysis, are applied to reveal the changing trend of landslide activities in high-risk areas. By extracting slope displacement data at multiple time points, the frequency, intensity and evolution characteristics of landslides are identified. First, multi-phase image data are extracted from remote sensing images, and after preprocessing, they are converted into a standardized form, and then the slope displacement data are obtained through change detection analysis. Then, the trend analysis method is used to identify the evolution pattern of the landslide and obtain landslide trend data.

[0065] Step S63: extracting landslide influencing factors based on the landslide trend data and key slope dynamic parameter data to obtain landslide influencing factor data; The embodiment of the present invention first combines landslide trend data with key slope dynamic parameter data to extract landslide influencing factors. Based on the aforementioned high-risk area screening results and time series trend analysis data, the landslide evolution trend in the area is identified, and combined with the dynamic displacement characteristics of the slope, the influencing factors of each area are quantified. The extraction of influencing factors is mainly based on the following aspects: meteorological factors (such as precipitation, temperature changes), geological factors (such as soil type, rock structure, shear strength) and hydrological factors (such as groundwater level, slope water flow changes). Multi-phase image data obtained by remote sensing technology is used to provide multi-dimensional spatial information for each high-risk area, so as to accurately analyze the relationship between these factors and landslide occurrence. By combining remote sensing images with ground measured data, the key factors affecting the occurrence of landslides are extracted and recorded in numerical form to obtain landslide influencing factor data.

[0066] Step S64: performing weight distribution analysis on the landslide influencing factor data to obtain influencing factor weight distribution data; The present invention performs a weighted analysis of landslide influencing factor data, primarily relying on multivariate analysis methods such as principal component analysis (PCA) or weighted regression analysis (WRA). The extracted landslide influencing factors are quantified and their impact on landslide occurrence is determined. First, based on the aforementioned landslide influencing factor extraction results, parameter data directly related to landslide occurrence is collected. This data is acquired through remote sensing imagery and supplemented with ground observation data. Next, each influencing factor category is standardized to make its quantitative values ​​comparable. Principal component analysis is then used to reduce the dimensionality of multiple factors and identify the most influential principal components. Next, weights are assigned to the identified principal components, determining the weight of each factor based on its contribution to the landslide impact. These factor weights are further calibrated through regression analysis to ensure data accuracy and consistency. Finally, the weight distribution data for the influencing factors is obtained.

[0067] Step S65: constructing a landslide prediction model using the influencing factor weight distribution data and the slope stability assessment data to generate landslide prediction result data; The embodiment of the present invention utilizes landslide influencing factor weight distribution data and slope stability assessment data to construct a landslide prediction model. First, the influencing factor data and slope stability assessment data are collected and organized. These data are derived from the processing results of the aforementioned steps and the slope stability analysis results. On this basis, a landslide prediction model is established using logistic regression analysis or a support vector machine (SVM) method. The weight of each landslide influencing factor is combined with the slope stability assessment results to perform model training. Through regression analysis of historical landslide data, the mathematical relationship between each factor and landslide occurrence is obtained, and the specific contribution of each factor is determined. After model training is completed, landslide prediction is performed using the established regression relationship or SVM model, and landslide prediction result data is output to predict the probability of landslide occurrence in the target area.

[0068] Step S66: Perform landslide warning based on the landslide prediction result data to generate landslide warning information data.

[0069] The embodiments of the present invention utilize landslide prediction data, combined with high-risk area distribution data and landslide trend data, to conduct a step-by-step analysis of the target area. By setting landslide warning thresholds within the target area, the probability of a landslide and the impact range are determined. First, the predicted landslide probability is extracted and compared with the set threshold to identify high-risk areas. Next, combining real-time remote sensing data, meteorological data, and geological monitoring data, a dynamic assessment of high-risk areas is performed to update landslide warning information data. In actual operation, a geographic information system (GIS) is used to perform spatial analysis of high-risk areas and landslide prediction results, generating a landslide warning area map that indicates the specific location, range, and time of impending landslides. By integrating real-time data streams, dynamic corrections and improvements are further performed to ensure the accuracy and timeliness of landslide warning information, ultimately generating specific landslide warning information data.

[0070] The high-risk area screening of the present invention ensures that key areas in dangerous mountainous areas are accurately identified, providing target area data support for subsequent risk prediction. Time series trend analysis reveals the dynamic change trend of high-risk areas, helps to identify potential landslide evolution characteristics, and lays a data foundation for landslide prediction. The extraction of landslide influencing factors combines landslide trend data with key slope dynamic parameters to accurately identify the main factors causing landslides, providing key inputs for the prediction model. The weight distribution analysis of landslide influencing factors reveals the relative importance of each factor in the occurrence of landslides, which helps to further refine the risk assessment. The construction of a landslide prediction model based on weight distribution data and stability assessment realizes the scientific prediction of the probability of landslide occurrence and enhances the accuracy of the prediction results. By analyzing the prediction results, the landslide warning provides timely and effective landslide warning information to support emergency response and disaster prevention and mitigation decision-making.

[0071] Preferably, step S66 includes the following steps: Step S661: Designing a landslide early warning model based on the landslide prediction result data to obtain landslide early warning model data; This embodiment of the present invention designs a landslide early warning model based on landslide prediction data. First, based on the aforementioned landslide prediction dataset, environmental factors and dynamic slope parameters relevant to landslide occurrence are selected to establish the initial input data for the early warning model. Leveraging known historical landslide data, as well as geological and meteorological factors, a regression analysis method or classification algorithm is used to establish the prediction model. By comparing historical landslide records across different regions, the spatial distribution patterns of landslide occurrence are determined. Subsequently, based on existing remote sensing datasets, spatial statistical analysis is used to analyze high-risk areas within the region. Furthermore, multi-source remote sensing data is combined to extract correlations between slope stability and landslide occurrence, thereby refining the early warning model. The model design utilizes a regional approach to address the characteristics of different mountainous areas. Remote sensing imagery, DEM (digital elevation model), and other topographic factors are used to construct a landslide early warning model framework. This framework uses a weighted algorithm to assign weights to each factor based on the spatial distribution of landslides, historical landslide data, and environmental factors, ultimately generating the landslide early warning model data.

[0072] Step S662: performing real-time monitoring data fusion analysis on the landslide early warning model data to obtain real-time landslide early warning data; The present invention utilizes real-time monitoring data collected through remote sensing technology, combined with data from landslide early warning models, to conduct a comprehensive analysis of this real-time data. Real-time monitoring data is acquired through sensors and drones, ensuring its real-time and accuracy. This monitoring data is wirelessly transmitted to a central data processing system and compared with historical landslide data. Subsequently, a data fusion algorithm is used to combine this real-time monitoring data with the early warning model data for a multi-dimensional fusion analysis. Kalman filtering or weighted averaging methods are used to denoise and weight the data to improve its reliability. This real-time calculation generates real-time landslide early warning data.

[0073] Step S663: issuing landslide warnings based on the real-time landslide warning data, and performing level classification to obtain landslide warning level data; This embodiment of the present invention first utilizes a geographic information system (GIS) to issue warning information based on real-time landslide warning data. By processing and analyzing real-time data, the specific location of the slope landslide and the warning timeliness are determined. The warning issuance system then sets different warning levels based on the real-time landslide warning data, classifying the risk level according to the severity of the landslide. Furthermore, by analyzing multi-level landslide information, this multidimensional data is input into a multi-level warning information analysis system. This system uses a weighted synthesis method to output real-time landslide warning level data based on the weights of various factors. The landslide warning level is determined by combining various monitoring indicators, such as slope displacement rate, precipitation intensity, and historical landslide data, to assess the urgency of the landslide and generate the landslide warning level data.

[0074] Step S664: integrating the landslide prediction result data, the real-time landslide warning data and the landslide warning level data to obtain landslide warning information data.

[0075] This embodiment of the present invention first fuses landslide prediction result data with real-time landslide warning data, merging the two types of data using a weighted average method to ensure consistency and reliability of data from different sources during the fusion process. During the data fusion process, the landslide prediction results are combined with terrain changes, precipitation, and soil moisture factors in the real-time monitoring data, combined with the three-dimensional modeling information of the slope, to generate a comprehensive landslide warning information dataset. Next, the landslide warning information data is spatially analyzed using a GIS platform to ensure that the geographic information of the warning data accurately corresponds to the corresponding geographic location. Furthermore, the real-time landslide warning data, prediction result data, and landslide warning level data are integrated to create a dataset that includes spatial location, risk assessment, warning level, and real-time monitoring status. This dataset is managed through a data warehouse to ensure timely data updates and efficient access. Ultimately, the integrated data is used to obtain landslide warning information data.

[0076] The landslide early warning model designed in this invention provides a theoretical framework for landslide prediction, ensuring the effectiveness and accuracy of the early warning system. The fusion analysis of real-time monitoring data and the landslide early warning model provides immediate assessment of landslide risk, enhancing the timeliness of the early warning system. Multi-level analysis of landslide early warning information provides a clear warning classification for different risk levels, facilitating the development of targeted emergency response measures. By integrating landslide prediction results, real-time warning data, and warning level information, comprehensive landslide early warning information is generated, providing comprehensive support for decision-making.

[0077] Preferably, step S7 includes the following steps: Step S71: Analyze the impact range of each landslide risk level using the landslide warning information data to obtain landslide impact range data; Based on landslide early warning information data, the present invention utilizes high-resolution remote sensing imagery and digital elevation model (DEM) data to analyze the impact range of landslide risk levels. First, remote sensing imagery is used to analyze the topography and landforms of the slope to identify potential landslide areas. The landslide prediction results are then combined with real-time monitoring data, and a landslide model is used to calculate the landslide's propagation range. Based on this, the system analyzes the slope's inclination angle, slope variation, geological characteristics, and precipitation factors to classify areas with different risk levels and generate landslide impact range data. A geographic information system (GIS) is used for spatial analysis to accurately match risk levels with regional boundaries, ensuring that the landslide impact range data accurately reflects the location and extent of different risk areas. This analysis yields landslide impact range data.

[0078] Step S72: performing an environmental sensitivity assessment of the high-risk area based on the landslide impact range data to obtain environmental sensitivity assessment data; Based on landslide impact data, this embodiment of the present invention selects high-risk areas for environmental sensitivity assessment. The core of this environmental sensitivity assessment is to analyze multiple environmental factors within the region, including geology, vegetation, hydrology, and land use, to assess their potential impact on the environment in the event of a landslide. First, the geological characteristics of the slope are analyzed. Remote sensing technology is used to extract different rock and soil types and structural characteristics. By analyzing their physical properties and stability, areas susceptible to landslides are identified. Furthermore, remote sensing imagery is used to obtain vegetation cover information to assess the impact of vegetation on slope stability, specifically the protective effects of vegetation water absorption capacity and root system stability on landslides in conditions of high precipitation. Next, hydrological data is used to analyze the impact of precipitation and groundwater levels on landslide risk. This data is spatially analyzed using a geographic information system (GIS). The environmental sensitivity assessment data is combined with landslide impact data to determine the environmental sensitivity level of each area. Finally, through environmental sensitivity analysis of high-risk areas, specific environmental sensitivity assessment data is obtained.

[0079] Step S73: performing risk identification and assessment on the environmental sensitivity assessment data to obtain comprehensive landslide risk assessment data; Based on environmental sensitivity assessment data, the present invention employs a multi-factor analysis method to assess the comprehensive risk of a region. First, the various factors in the environmental sensitivity assessment data are quantified. Three-dimensional slope models acquired through remote sensing technology provide detailed topographic data for analysis. Combined with digital elevation models (DEMs) and geological survey data, the stability of different regions is analyzed. Next, a landslide probability model is used to calculate the risk value for each high-risk area. This model considers multiple factors, including precipitation intensity, soil moisture, and slope gradient, to calculate the likelihood and impact of a landslide under different conditions in each region. Based on this, a weighted comprehensive approach is used to weight each risk factor according to its impact on the landslide, resulting in comprehensive landslide risk assessment data. Spatial analysis is performed using a GIS platform to ensure that the assessment data is accurately mapped to the geographic location of each region, creating a spatial distribution map showing the risk level of different regions. Ultimately, comprehensive landslide risk assessment data is obtained.

[0080] Step S74: formulating landslide prevention strategy priorities based on the landslide comprehensive risk assessment data to obtain landslide prevention strategy priority data; The embodiment of the present invention uses a decision analysis method to prioritize landslide prevention strategies based on comprehensive landslide risk assessment data. First, the comprehensive risk assessment data of each region is sorted, and the prevention needs of each region are determined based on risk level, geographical location, and infrastructure influencing factors. The three-dimensional slope model obtained by remote sensing technology provides support for the terrain characteristics of different regions. Combined with the GIS platform, the landslide risk of each region is spatially divided to ensure that high-risk areas have a higher priority. Then, based on the comprehensive risk assessment results, detailed prevention strategies are formulated for high-risk areas. Each prevention strategy includes monitoring system construction, geological reinforcement, and drainage system optimization, and these strategies are ranked according to the urgency of implementation, feasibility, and regional resource conditions. In this process, by comparing with historical landslide data, the impact of different prevention strategies on the probability of landslide occurrence is evaluated, and finally the landslide prevention strategy priority data is obtained.

[0081] Step S75: performing regional adaptability analysis on the landslide prevention strategy priority data to obtain landslide prevention strategy planning data; The embodiment of the present invention conducts a regional adaptability analysis of landslide prevention strategies based on landslide prevention strategy priority data. During this process, remote sensing image data and a geographic information system (GIS) are first used to conduct a detailed analysis of the geographical characteristics of each landslide area to determine the actual situation in each area. Combined with the established three-dimensional slope model, a matching analysis is conducted between different landslide risk levels and infrastructure in the region, with special attention paid to the vulnerability and carrying capacity of infrastructure in high-risk areas. For these high-risk areas, spatial analysis tools are used to conduct a regional adaptability assessment of landslide prevention strategies, including the layout of monitoring facilities, geological reinforcement measures, and the feasibility of implementing drainage engineering plans. All prevention strategies will give priority to technical means with strong adaptability in areas with complex terrain and inconvenient transportation. By integrating environmental data and prevention strategy priority data, targeted landslide prevention strategy planning data is formed.

[0082] Step S76: Perform dynamic feasibility analysis on the landslide prevention strategy planning data to obtain dynamic landslide warning adjustment data; This embodiment of the present invention conducts dynamic feasibility analysis of landslide prevention strategy planning based on landslide prevention strategy planning data. First, using high-resolution satellite imagery and terrain data acquired through remote sensing technology, the latest conditions in landslide-affected areas are monitored in real time. Combined with historical landslide data, the impact of current environmental changes on landslide prevention strategies is assessed. Furthermore, combined with three-dimensional slope modeling data, parameters in high-risk areas are dynamically evaluated to ensure that prevention strategies can adapt to varying climatic conditions and geological variations. Next, by comparing this data with existing infrastructure data in a geographic information system (GIS) platform, the feasibility and effectiveness of various landslide prevention measures are evaluated. Based on the risk characteristics of different regions, a multi-dimensional risk assessment method is employed, taking into account factors such as precipitation and soil changes, to dynamically adjust the specific implementation of prevention strategies. For example, in mountainous areas with inaccessible transportation, more efficient geological reinforcement plans are prioritized; whereas in areas with more convenient transportation, the focus is on optimizing monitoring and early warning facilities. Ultimately, through continuous updating and dynamic adjustments, dynamic landslide early warning adjustment data tailored to different regions and environmental changes is generated.

[0083] Step S77: Dynamically optimize the landslide prevention strategy planning data using the dynamic landslide early warning adjustment data to obtain a landslide early warning and prevention plan.

[0084] This embodiment of the present invention dynamically optimizes landslide prevention strategy planning data based on dynamic landslide early warning adjustment data, combined with the latest satellite imagery, topographic survey data, and meteorological data acquired through remote sensing technology. This process utilizes 3D modeling to integrate information about the slopes, geological structures, and vegetation cover in the landslide-affected area in three-dimensional space, creating a highly accurate 3D model. Then, based on real-time monitoring data and dynamic environmental change data, protective measures are adjusted for different areas within the 3D slope model. For example, in high-risk areas, prevention plans are optimized by adjusting soil reinforcement, slope vegetation restoration, and drainage measures. During this optimization process, the potential risk of landslides is reassessed through Geographic Information System (GIS) and remote sensing image analysis, further combined with meteorological changes and precipitation forecasts. The early warning system is then calibrated to ensure real-time responsiveness to changing environmental conditions. Based on this data, a comprehensive landslide early warning and prevention plan is ultimately generated, incorporating specific protective measures tailored to different geological and meteorological conditions and regional characteristics.

[0085] The landslide risk level impact range analysis of the present invention provides the spatial distribution of areas with different risk levels and clarifies the specific impact areas of landslide risks. The environmental sensitivity assessment of high-risk areas provides priority environmental factors for landslide prevention measures, supporting more accurate risk management and control. The comprehensive risk analysis provides a comprehensive landslide risk assessment by combining multiple factors, helping to formulate more scientific response strategies. The priority formulation of landslide prevention strategies ensures timely response to high-risk areas and improves resource allocation efficiency. The regional adaptability analysis ensures the applicability of landslide prevention strategies under different regional conditions and enhances the effectiveness of prevention measures. The dynamic feasibility analysis provides a basis for the adjustment of landslide prevention strategies, ensuring that the strategies are always effective in a constantly changing environment. The dynamic optimization of landslide early warning and prevention plans improves the timeliness and accuracy of landslide response strategies and ensures real-time prevention capabilities.

[0086] Preferably, the present invention further provides a three-dimensional modeling system for slopes in difficult mountainous areas based on remote sensing technology, which is used to execute the above-mentioned three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology. The three-dimensional modeling system for slopes in difficult mountainous areas based on remote sensing technology comprises: The data preprocessing module is used to obtain the original remote sensing data set of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing data set; The surface feature extraction module is used to extract the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain the surface feature data of the difficult mountainous area; The 3D terrain modeling module is used to construct a 3D terrain model using surface feature data in difficult mountainous areas to obtain preliminary 3D terrain model data; The slope dynamic analysis module is used to perform time series analysis on standardized remote sensing data sets and use the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; Slope stability assessment module, used to perform environmental factor stability analysis on slope displacement change data to obtain slope stability assessment data; The landslide prediction and warning module is used to build a landslide prediction model based on slope stability assessment data, and to issue landslide warnings based on landslide prediction results, thereby generating landslide warning information data; The risk assessment and prevention strategy module is used to use landslide warning information data to conduct risk assessment in target difficult mountainous areas, and to formulate prevention strategies based on the risk assessment results to obtain landslide warning and prevention plans.

Claims

1. A three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology, characterized in that: The following steps are involved: Step S1: Obtain the original remote sensing dataset of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing dataset; Step S2: extracting the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain surface feature data of the difficult mountainous area; Step S3: constructing a three-dimensional terrain model using surface feature data of the difficult mountainous area to obtain preliminary three-dimensional terrain model data; Step S4: performing time series analysis on the standardized remote sensing data set, and using the time series analysis results to perform slope dynamic change analysis on the preliminary three-dimensional terrain model data to obtain slope displacement change data; Step S5: Performing environmental factor stability analysis on the slope displacement change data to obtain slope stability assessment data; Step S6: constructing a landslide prediction model based on the slope stability assessment data, and performing landslide warning based on the landslide prediction results to generate landslide warning information data; Step S7: Use the landslide warning information data to conduct a risk assessment of the target difficult mountainous area, and formulate a prevention strategy based on the risk assessment results to obtain a landslide warning and prevention plan.

2. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Acquire synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data of the target difficult mountainous area, and record the synthetic aperture radar image data, lidar point cloud data, drone image data, and optical remote sensing image data as the original remote sensing data set; Step S12: performing data preprocessing on the original remote sensing data set to obtain a preprocessed remote sensing data set; Step S13: performing standardization processing on the pre-processed remote sensing data set to obtain a standardized remote sensing data set.

3. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 2 is characterized in that: Step S2 includes the following steps: Step S21: performing image partitioning processing of the target difficult mountainous area based on the standardized remote sensing data set, and performing preliminary surface coverage area extraction to obtain preliminary surface area division data of the mountainous area; Step S22: performing optical image spectrum feature analysis on the preliminary surface area division data of the mountainous area to obtain vegetation feature data of the mountainous area; Step S23: extracting the terrain characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain the mountainous area terrain factor characteristic data; Step S24: extracting water characteristics of the target difficult mountainous area based on the standardized remote sensing data set to obtain water characteristics data of the mountainous area; Step S25: Analyze the surface roughness characteristics of the target difficult mountainous area using the standardized remote sensing data set to obtain the exposed surface characteristic data of the mountainous area; Step S26: extracting vegetation height and building height information in the target difficult mountainous area using the standardized remote sensing data set to obtain three-dimensional height feature data of the mountainous area; Step S27: performing a refined extraction of landslide-prone slope areas based on the mountainous terrain factor characteristic data and the mountainous water body characteristic data to obtain mountainous slope characteristic data; Step S28: fusing the mountain vegetation feature data, mountain water feature data, mountain exposed surface feature data, mountain three-dimensional height feature data, and mountain slope feature data to obtain surface feature data in difficult mountainous areas.

4. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 3 is characterized in that: Step S3 includes the following steps: Step S31: extracting surface elevation characteristics based on the surface characteristic data of the difficult mountainous area to obtain terrain elevation characteristic data; Step S32: performing three-dimensional gridding processing on the terrain elevation feature data and performing preliminary terrain grid construction to obtain three-dimensional terrain grid data; Step S33: performing elevation adjustment on the three-dimensional terrain grid data based on the three-dimensional height characteristic data of the mountain area to obtain refined terrain grid data; Step S34: performing surface feature enhancement processing on the refined terrain grid data based on the mountainous area vegetation feature data and the mountainous area exposed surface feature data to obtain textured terrain grid data; Step S35: using the mountain water body characteristic data to perform water body range correction on the textured terrain grid data to obtain water body corrected terrain grid data; Step S36: performing high-precision slope area refinement on the water body corrected terrain grid data to obtain high-precision slope terrain grid data; Step S37: Perform multi-perspective analysis on the high-precision slope terrain grid data and verify its accuracy to generate preliminary three-dimensional terrain model data.

5. The three-dimensional modeling method of slopes in difficult mountainous areas based on remote sensing technology according to claim 4 is characterized in that: Step S4 includes the following steps: Step S41: extracting multi-temporal image data from the standardized remote sensing data set and sorting them in time sequence to obtain time series image data; Step S42: performing change detection analysis on the time series image data and identifying the surface dynamic change areas to obtain surface change situation data; Step S43: performing slope landform evolution regional analysis on the preliminary three-dimensional terrain model data based on the surface change situation data to obtain slope change hotspot data; Step S44: performing refined displacement monitoring on the slope change hotspot data based on the time series image data to obtain slope displacement monitoring data; Step S45: performing dynamic displacement characteristic analysis on the slope displacement monitoring data, and extracting the slope displacement rate and direction to obtain slope dynamic displacement characteristic data; Step S46: performing multi-temporal surface deformation comparative analysis on the time series image data based on the slope dynamic displacement characteristic data to obtain surface deformation evolution data; Step S47: Perform spatial interpolation processing on the surface deformation evolution data to obtain slope displacement change data.

6. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 5 is characterized in that: Step S5 includes the following steps: Step S51: extracting key dynamic parameters from the slope displacement change data to obtain key slope dynamic parameter data; Step S52: extracting meteorological factors, geological factors, and hydrological factors from the standardized remote sensing data set to obtain a mountainous area environmental factor data set; Step S53: performing correlation analysis on the key slope dynamic parameter data and the mountainous area environmental factor data set, and identifying the dominant slope dynamic change environmental factors to obtain the dominant environmental factor data; Step S54: performing weight distribution analysis on the dominant environmental factor data to obtain environmental factor weight distribution data; Step S55: Perform slope stability assessment based on the environmental factor weight distribution data and key slope dynamic parameter data to obtain slope stability assessment data.

7. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 6 is characterized in that: Step S6 includes the following steps: Step S61: Screening high-risk areas in the target difficult mountainous area based on the slope stability assessment data to obtain high-risk area distribution data; Step S62: performing time series trend analysis on the high-risk area distribution data and identifying landslide evolution characteristics to obtain landslide trend data; Step S63: extracting landslide influencing factors based on the landslide trend data and key slope dynamic parameter data to obtain landslide influencing factor data; Step S64: performing weight distribution analysis on the landslide influencing factor data to obtain influencing factor weight distribution data; Step S65: constructing a landslide prediction model using the influencing factor weight distribution data and the slope stability assessment data to generate landslide prediction result data; Step S66: Perform landslide warning based on the landslide prediction result data to generate landslide warning information data.

8. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 7 is characterized in that: Step S66 includes the following steps: Step S661: Designing a landslide early warning model based on the landslide prediction result data to obtain landslide early warning model data; Step S662: performing real-time monitoring data fusion analysis on the landslide early warning model data to obtain real-time landslide early warning data; Step S663: issuing landslide warnings based on the real-time landslide warning data, and performing level classification to obtain landslide warning level data; Step S664: integrating the landslide prediction result data, the real-time landslide warning data and the landslide warning level data to obtain landslide warning information data.

9. The three-dimensional modeling method for slopes in difficult mountainous areas based on remote sensing technology according to claim 8 is characterized in that: Step S7 includes the following steps: Step S71: Analyze the impact range of each landslide risk level using the landslide warning information data to obtain landslide impact range data; Step S72: performing an environmental sensitivity assessment of the high-risk area based on the landslide impact range data to obtain environmental sensitivity assessment data; Step S73: performing risk identification and assessment on the environmental sensitivity assessment data to obtain comprehensive landslide risk assessment data; Step S74: formulating landslide prevention strategy priorities based on the landslide comprehensive risk assessment data to obtain landslide prevention strategy priority data; Step S75: performing regional adaptability analysis on the landslide prevention strategy priority data to obtain landslide prevention strategy planning data; Step S76: Perform dynamic feasibility analysis on the landslide prevention strategy planning data to obtain dynamic landslide warning adjustment data; Step S77: Dynamically optimize the landslide prevention strategy planning data using the dynamic landslide early warning adjustment data to obtain a landslide early warning and prevention plan.

10. A three-dimensional modeling system for slopes in difficult mountainous areas based on remote sensing technology, characterized in that: Used to execute the three-dimensional modeling method of slopes in difficult mountainous areas based on remote sensing technology as claimed in claim 1, the three-dimensional modeling system of slopes in difficult mountainous areas based on remote sensing technology comprises: The data preprocessing module is used to obtain the original remote sensing data set of the target difficult mountainous area and perform data preprocessing to obtain a standardized remote sensing data set; The surface feature extraction module is used to extract the surface features of the target difficult mountainous area from the standardized remote sensing data set to obtain the surface feature data of the difficult mountainous area; The 3D terrain modeling module is used to construct a 3D terrain model using surface feature data in difficult mountainous areas to obtain preliminary 3D terrain model data; The slope dynamic analysis module is used to perform time series analysis on standardized remote sensing data sets and use the time series analysis results to perform slope dynamic change analysis on preliminary three-dimensional terrain model data to obtain slope displacement change data; Slope stability assessment module, used to perform environmental factor stability analysis on slope displacement change data to obtain slope stability assessment data; The landslide prediction and warning module is used to build a landslide prediction model based on slope stability assessment data, and to issue landslide warnings based on landslide prediction results, thereby generating landslide warning information data; The risk assessment and prevention strategy module is used to use landslide warning information data to conduct risk assessment in target difficult mountainous areas, and to formulate prevention strategies based on the risk assessment results to obtain landslide warning and prevention plans.

Images