Disaster area high-precision surveying and mapping method based on point cloud data

By using point cloud data processing technology, the problem of high-precision 3D data acquisition and processing in complex environments has been solved, enabling high-precision mapping of disaster areas and timely early warning of disaster hazards, providing important support for disaster prevention and mitigation.

CN120953298APending Publication Date: 2025-11-14GUANGDONG UNIV OF TECH
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202511056111.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-30
Publication Date
2025-11-14

AI Technical Summary

Technical Problem

Existing technologies struggle to acquire high-precision three-dimensional spatial information and process massive amounts of data in complex environments, leading to errors in disaster hazard identification and deformation trend analysis, especially in areas with insufficient data quality and monitoring coverage due to vegetation cover or complex terrain.

Method used

A point cloud-based approach is adopted to acquire three-dimensional spatial data through multi-source sensor fusion, perform point cloud segmentation and spatial interpolation to construct a terrain model, perform gridding and noise filtering, extract key deformation features, and analyze deformation trends through time series comparison.

Benefits of technology

It enables precise monitoring of terrain deformation and timely early warning of potential disasters in complex environments, providing high-precision disaster prevention and control support.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120953298A_ABST
    Figure CN120953298A_ABST
Patent Text Reader

Abstract

The invention discloses a disaster area high-precision surveying and mapping method based on point cloud data. The disaster area high-precision surveying and mapping method comprises the steps of obtaining a spatial data set of a disaster area; performing hierarchical processing on the spatial data set of the disaster area, extracting core point cloud data related to the terrain, and constructing a terrain point cloud set; according to the constructed terrain point cloud set, performing data completion on the missing region to obtain a three-dimensional terrain model; segmenting the three-dimensional terrain model to obtain segmented terrain data units; performing de-noising processing on each terrain data unit to obtain a de-noised terrain data unit set; performing data fluctuation detection on the denoised topographic data unit set, and performing local feature enhancement processing on the topographic data units with detected data fluctuation exceeding a preset threshold to obtain a key deformation feature set; and according to the key deformation feature set, carrying out matching analysis on the current feature and historical data to obtain a deformation trend distribution diagram. According to the invention, accurate monitoring of terrain deformation is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of disaster area mapping technology, and in particular relates to a high-precision mapping method for disaster areas based on point cloud data. Background Technology

[0002] Geological disaster monitoring and mapping plays an irreplaceable role in safeguarding people's lives and property and maintaining social stability, and is one of the core components of disaster prevention and mitigation. With the increasing frequency of natural disasters, accurate and timely monitoring and assessment of disaster areas has become an urgent need, directly impacting early warning effectiveness and emergency response capabilities. However, current mainstream methods have revealed many shortcomings in practical applications, particularly in dealing with complex environments and dynamic changes, failing to meet the demands for both high precision and high efficiency. Existing technologies are often limited by the singularity of data acquisition methods and the lag in processing capabilities, leading to significant errors in disaster hazard identification and deformation trend analysis, especially in areas with vegetation cover or complex terrain, where data quality and monitoring coverage often fall short of expectations. Against this backdrop, the main challenges in this field focus on how to effectively acquire high-precision three-dimensional spatial information and how to process massive amounts of data to extract key features. First, disaster areas are typically characterized by complex terrain and variable environments; traditional data acquisition methods struggle to penetrate vegetation or adapt to steep terrain, resulting in insufficiently accurate and comprehensive surface information. This limitation in data acquisition further affects the accuracy of subsequent analysis, often leading to omissions or misjudgments in the early identification of disaster hazards. Secondly, due to the massive volume of collected data and its high noise levels, efficiently filtering out irrelevant information while preserving feature details becomes a key bottleneck affecting the construction of monitoring models and the accuracy of deformation analysis. These two factors are interrelated; the former determines the basic quality of the data, while the latter directly impacts the data's ability to be transformed into practical results. Therefore, how to achieve rapid acquisition of high-precision 3D data in complex environments and extract key deformation features of disaster areas through efficient data processing methods has become a critical problem that urgently needs to be solved. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes a high-precision mapping method for disaster areas based on point cloud data. This method effectively solves the problems of terrain and vegetation obstruction in complex environments, enables precise monitoring of terrain deformation and timely early warning of potential disaster hazards, and provides important technical support for geological disaster prevention and control.

[0004] To achieve the above objectives, this invention provides a high-precision mapping method for disaster areas based on point cloud data, comprising:

[0005] Obtain spatial datasets of the disaster area;

[0006] The spatial dataset of the disaster area is processed in layers to extract core point cloud data related to the terrain and construct a terrain point cloud set.

[0007] Based on the constructed terrain point cloud set, data is filled in for missing areas to obtain a three-dimensional terrain model;

[0008] The three-dimensional terrain model is segmented to obtain segmented terrain data units;

[0009] Each terrain data unit is denoised to obtain a set of denoised terrain data units;

[0010] Data fluctuation detection is performed on the denoised terrain data unit set. For terrain data units where the detected data fluctuation exceeds the preset threshold, local feature enhancement processing is performed to obtain the key deformation feature set.

[0011] Based on the set of key deformation features, the current features are matched and analyzed with historical data to obtain a deformation trend distribution map;

[0012] The deformation trend of the deformation trend distribution map is compared with a preset safety threshold, and areas exceeding the preset threshold are marked with priority to determine the disaster area distribution information.

[0013] Technical Effects of this Invention: This invention discloses a high-precision mapping method for disaster areas based on point cloud data. It acquires three-dimensional spatial data through multi-source sensor fusion, constructs a complete terrain model using point cloud segmentation and spatial interpolation techniques, and then performs gridding and noise filtering. For units with data fluctuations exceeding a threshold, local feature enhancement is performed to extract potential deformation features, and deformation trends are analyzed through time-series comparison. Finally, areas exceeding safety thresholds are marked based on the deformation trend distribution map, generating key areas of concern for potential disasters. This invention effectively solves the problem of terrain and vegetation obstruction in complex environments, achieving accurate monitoring of terrain deformation and timely early warning of potential disasters, providing important technical support for geological disaster prevention and control. Attached Figure Description

[0014] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0015] Figure 1 This is a flowchart illustrating the high-precision mapping method for disaster areas based on point cloud data, as described in an embodiment of the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0018] like Figure 1 As shown, this embodiment provides a high-precision mapping method for disaster areas based on point cloud data, including:

[0019] Obtain spatial datasets of the disaster area;

[0020] The spatial dataset of the disaster area is processed in layers to extract core point cloud data related to the terrain and construct a terrain point cloud set.

[0021] Based on the constructed terrain point cloud set, data is filled in for missing areas to obtain a three-dimensional terrain model;

[0022] The three-dimensional terrain model is segmented to obtain segmented terrain data units;

[0023] Each terrain data unit is denoised to obtain a set of denoised terrain data units;

[0024] Data fluctuation detection is performed on the denoised terrain data unit set. For terrain data units where the detected data fluctuation exceeds the preset threshold, local feature enhancement processing is performed to obtain the key deformation feature set.

[0025] Based on the set of key deformation features, the current features are matched and analyzed with historical data to obtain a deformation trend distribution map;

[0026] The deformation trend of the deformation trend distribution map is compared with a preset safety threshold, and areas exceeding the preset threshold are marked with priority to determine the disaster area distribution information.

[0027] Furthermore, obtaining spatial datasets of the disaster area includes:

[0028] Collect surface signal data in complex environments. For areas obscured by terrain and vegetation, obtain three-dimensional spatial information containing raw signals from at least one sensor. Use a preset threshold to filter the signal intensity to obtain a preliminary set of three-dimensional spatial signals.

[0029] Based on the preliminary three-dimensional spatial signal set, the multi-source signals are integrated and processed according to the signal differences under different terrain conditions. If the signal strength is lower than the preset threshold, the missing area is supplemented with data to determine the fused three-dimensional spatial data set.

[0030] By extracting terrain features from the fused 3D spatial data set, and denoising the data set to address signal interference in vegetation-covered areas, denoised 3D spatial data units are obtained, and it is determined whether the denoised 3D spatial data units meet the preset surface information integrity standards.

[0031] Based on the denoised 3D spatial data units, spatial reconstruction is performed on the denoised 3D spatial data units to meet the surface information coverage requirements in complex environments. If there are blank areas in the reconstructed data grid, data filling is performed to obtain the spatial dataset of the disaster area.

[0032] Specifically, in terrain feature extraction and denoising, median filtering can be used to process the fused data set and remove abnormal noise points to address signal interference in areas obscured by vegetation. For example, if abnormally high values ​​in lidar data are caused by leaf reflection in a certain area, median filtering can smooth these anomalies, resulting in 3D spatial data units that more closely resemble the actual surface morphology. Subsequently, it is determined whether the data unit meets integrity standards, such as requiring a data coverage rate of over 90%. If it does not meet the standard, further processing is required. This step effectively improves data quality and ensures the accuracy of subsequent reconstruction. Specifically, in the spatial reconstruction stage, to meet the surface information coverage requirements in complex environments, a grid partitioning tool can be used to divide the data units into 10m × 10m grid cells. If data is missing in a certain grid, such as due to signal gaps caused by mountain shadows, data from surrounding grids is used to fill the gaps, generating a smooth 3D spatial dataset. The advantage of this method is that it can construct a continuous surface model, meeting the needs of high-precision terrain analysis.

[0033] Furthermore, constructing the terrain point cloud set includes:

[0034] Based on the spatial dataset of the disaster area, the data is processed in layers. Taking into account the different characteristics of surface signals and vegetation interference, the point cloud part related to terrain features is separated from at least one data layer to obtain a preliminary separated point cloud subset.

[0035] For the initially separated point cloud subsets, vegetation interference and other non-topographic signals are filtered out. If the proportion of interference signals exceeds a preset threshold, the interference area is corrected to determine the filtered point cloud grouping.

[0036] Based on the filtered point cloud groups, the spatial distribution is analyzed. To meet the requirement of continuity of terrain features, point cloud units related to the core data are extracted from at least one group to obtain refined terrain point cloud units.

[0037] For the refined terrain point cloud units, the data is spatially reorganized. If data gaps are found in the grid, the gaps are filled to determine the terrain point cloud set.

[0038] Specifically, when performing spatial distribution analysis based on filtered point cloud groups, clustering tools can group points based on distance and density. For example, in a mountainous scene where terrain features continuous slopes, a distance threshold can be set; adjacent points less than 0.5 meters apart can be grouped together. Point cloud units related to core terrain data can then be extracted to form refined terrain point cloud units. This approach highlights the main terrain features, providing a clear data foundation for subsequent processing. For these refined terrain point cloud units, a grid mapping tool can be used for spatial reorganization, mapping the data to a regular grid. Assuming a grid size of 5 meters by 5 meters, if a grid contains fewer data points than expected (e.g., fewer than 10 points), it is considered a missing data area. In this case, interpolation tools are used to supplement the missing data using valid data points from surrounding grids, estimating the value of the missing area and ultimately forming a complete terrain point cloud set. This method ensures the spatial continuity of the data, providing reliable support for subsequent terrain analysis.

[0039] Furthermore, obtaining a three-dimensional terrain model includes:

[0040] Based on the terrain point cloud set, a comprehensive inspection of the terrain data is carried out. For at least one missing area, the relevant spatial distribution range is obtained from the information gaps to determine the priority range of area restoration.

[0041] For missing areas within the priority range, data is used to fill the information gaps. Surrounding point cloud data is obtained from the gaps caused by vegetation occlusion and terrain occlusion to obtain a preliminary completed subset of point clouds.

[0042] If the initially completed point cloud subset has uneven data distribution in some areas, the spatially reorganized data is adjusted to obtain a continuous distribution that conforms to the terrain features, and the adjusted point cloud grouping is determined.

[0043] Based on the adjusted point cloud grouping, the terrain data is finally integrated. To ensure data integrity, relevant data units are extracted from the adjusted point cloud grouping to obtain a three-dimensional terrain model.

[0044] Furthermore, acquiring the segmented terrain data units includes:

[0045] Based on the 3D terrain model, the overall terrain is initially segmented. For at least one large area, the size of the segmented grid cells is determined according to a preset threshold, and the preliminary terrain data cells are obtained from them.

[0046] For the initially divided terrain data units, the regional boundaries of each unit are identified. If uneven distribution of regional boundaries is detected, the units are re-divided to obtain terrain segmentation units with clear boundaries.

[0047] Based on the clearly defined terrain segmentation units, the data distribution within each unit is detected. If the data distribution uniformity within a unit is lower than a preset threshold, the data within the unit is redistributed to determine a uniformly distributed set of block data.

[0048] For a uniformly distributed set of segmented data, data integration processing is performed on all segmented units to extract unit-scale data that meets the processing scale requirements, thus obtaining segmented terrain data units.

[0049] Specifically, for initially segmented terrain data units, boundary detection tools are used to identify the regional boundaries of each unit. The principle of boundary detection is to analyze the distribution of data points at the unit's edges to determine if the boundaries are clear. If, in a unit, the detection reveals sparse data point distribution at the boundaries, or even breaks in some boundaries, the unit can be re-divided using boundary adjustment tools. Specific adjustments can involve expanding or shrinking the boundary range, such as extending the boundary outward by 2 meters to include more data points, ultimately resulting in clearly defined terrain segmentation units. This adjustment helps ensure the independence and integrity of each unit. When processing terrain segmentation units with clear boundaries, data distribution analysis tools are used to detect the uniformity of data within the unit. The principle is to statistically analyze the density and distribution characteristics of data points within the unit. If the data distribution uniformity of a unit is lower than a preset threshold, for example, data points are mainly concentrated in the unit center with almost no data in the edge areas, the data can be redistributed using data balancing tools.

[0050] Furthermore, the denoised terrain data unit set includes:

[0051] Based on the segmented terrain data units, the data in each unit is initially decomposed. For the data of at least one region, the part that may contain noise signals is separated out, and the initially separated data set is obtained from it.

[0052] For the initially separated dataset, noise signals within each unit are identified. If the detected noise signal exceeds a preset threshold, the dataset is cleaned to obtain purified terrain data fragments.

[0053] Based on the purified terrain data fragments, the regional distribution of each unit is compared. If uneven regional distribution is found, the boundaries of the fragments are adjusted to determine a reasonable combination of terrain data.

[0054] For a reasonably distributed combination of terrain data, all block units are aggregated and processed, and the combination is finally calibrated to obtain a set of denoised terrain data units.

[0055] Specifically, when processing purified terrain data fragments, the data integrity check tool is used to compare the regional distribution of each unit. Its principle is to analyze the distribution density of data points within the unit to determine if there is uneven distribution. If a unit's data points are mainly concentrated in the central area, with sparse data at the edges, a data integration tool can be used to adjust the fragment boundaries, for example, extending the boundaries outward by 1 meter to cover more data points, ultimately determining a reasonably distributed terrain data combination. This adjustment method helps ensure the integrity and consistency of each unit's data. For reasonably distributed terrain data combinations, the unit integration tool is used to aggregate all segmented units into a single dataset. Its principle is to merge unit data according to preset standards, while simultaneously performing final calibration using a filtering accuracy verification tool. If, in a merged dataset, some units do not meet the accuracy requirements (e.g., data point density is less than 5 points per square meter), a calibration tool can be used to supplement these units, ultimately forming a denoised terrain data unit set that meets the requirements. This calibration process ensures the overall data quality and usability, providing solid support for subsequent terrain modeling or analysis.

[0056] Furthermore, obtaining the key deformation feature set includes:

[0057] Based on the denoised terrain data unit set, each unit is scanned one by one. If the detected data fluctuation exceeds the preset threshold, the terrain data unit is marked to obtain the marked unit group to be processed.

[0058] For the labeled unit group to be processed, the local features within the terrain data unit are enhanced to extract hidden deformation-related information and identify feature segments containing potential deformation.

[0059] Based on the feature segments containing potential deformations, the deformation information within the segments is classified and compared. If the significance of the deformation information is higher than a preset threshold, it is classified as a key deformation feature, and the classified key deformation feature group is obtained.

[0060] For the classified key deformation feature groups, the key deformation features are summarized and organized, and feature sets related to terrain data are extracted from them to obtain the key deformation feature set.

[0061] Specifically, for the labeled unit groups to be processed, the feature enhancement tool is used to delve deeper into the local feature information within the unit. Its core lies in extracting hidden deformation-related information through enhancement processing. For example, in a mountainous unit, if the data in a local area shows weak irregularities, the feature enhancement tool will increase the signal strength in this area to more clearly present potential deformation features. This processing method helps analysts more accurately identify subtle changes that may affect terrain stability. For feature segments containing potential deformation, the feature filtering tool is used to classify and compare deformation information. Its principle is to compare the significance of deformation information with a preset threshold, for example, a threshold set at 8%. If the significance of a certain segment reaches 10%, it will be classified as a key deformation feature. This classification and comparison method can effectively filter out information crucial to terrain analysis, ensuring that subsequent processing focuses on the core issues.

[0062] Furthermore, obtaining the deformation trend distribution map includes:

[0063] Based on the key deformation feature set, relevant time series segments are extracted from the pre-established historical data archive. A preliminary comparison is made between the segments and the current features to obtain a feature difference set after the preliminary comparison.

[0064] For the feature difference set, the deformation trend within the difference set is analyzed. If the direction and magnitude of change within the difference set exceed a preset threshold, it is marked as a significant change feature, and a significant change feature group is determined.

[0065] Based on the significant change feature groups, the feature groups are correlated with spatial distribution information to obtain deformation trend distribution data and determine the change characteristics of the data in different regions.

[0066] For deformation trend distribution data, the direction and magnitude of change are visualized to obtain a deformation trend distribution map.

[0067] Specifically, for feature difference sets, feature matching tools are used to deeply analyze the direction and magnitude of deformation trend changes. The core of this approach is to compare the data within the difference set with a preset threshold to determine whether it needs to be marked as a significant change feature. For example, if the preset threshold is a change rate of 1.5%, and the change rate difference in the above difference set is 2%, and the change direction shows a continuous intensification, then this feature will be marked as a significant change feature. This method helps to quickly identify key change points that may affect terrain stability. For groups of significant change features, data mapping tools are used to associate the feature groups with spatial distribution information to obtain deformation trend distribution data. The principle is to analyze the performance characteristics of deformation features in different regions through spatial mapping. For example, in a mountainous area, if the significant change feature group shows a higher change rate on the northern slope and greater stability in the south, the mapping tool will generate corresponding distribution data, clearly presenting the change characteristics of the northern region. This association provides a spatial dimension reference for subsequent analysis.

[0068] This invention discloses a high-precision mapping method for disaster areas based on point cloud data. It acquires three-dimensional spatial data through multi-source sensor fusion, constructs a complete terrain model using point cloud segmentation and spatial interpolation techniques, and then performs gridding and noise filtering. For units with data fluctuations exceeding a threshold, local feature enhancement is performed to extract potential deformation features, and deformation trends are analyzed through time-series comparison. Finally, areas exceeding safety thresholds are marked based on the deformation trend distribution map, generating key areas of concern for potential disasters. This invention effectively solves the problem of terrain and vegetation obstruction in complex environments, achieving accurate monitoring of terrain deformation and timely early warning of potential disasters, providing important technical support for geological disaster prevention and control.

[0069] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A high-precision mapping method for disaster areas based on point cloud data, characterized in that, include: Obtain spatial datasets of the disaster area; The spatial dataset of the disaster area is processed in layers to extract core point cloud data related to the terrain and construct a terrain point cloud set. Based on the constructed terrain point cloud set, data is filled in for missing areas to obtain a three-dimensional terrain model; The three-dimensional terrain model is segmented to obtain segmented terrain data units; Each terrain data unit is denoised to obtain a set of denoised terrain data units; Data fluctuation detection is performed on the denoised terrain data unit set. For terrain data units where the detected data fluctuation exceeds the preset threshold, local feature enhancement processing is performed to obtain the key deformation feature set. Based on the set of key deformation features, the current features are matched and analyzed with historical data to obtain a deformation trend distribution map; The deformation trend of the deformation trend distribution map is compared with a preset safety threshold, and areas exceeding the preset threshold are marked with priority to determine the disaster area distribution information.

2. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, The spatial dataset of the disaster area includes: Collect surface signal data in complex environments. For areas obscured by terrain and vegetation, obtain three-dimensional spatial information containing raw signals from at least one sensor. Use a preset threshold to filter the signal intensity to obtain a preliminary set of three-dimensional spatial signals. Based on the preliminary three-dimensional spatial signal set, the multi-source signals are integrated and processed according to the signal differences under different terrain conditions. If the signal strength is lower than the preset threshold, the missing area is supplemented with data to determine the fused three-dimensional spatial data set. By extracting terrain features from the fused 3D spatial data set, and denoising the data set to address signal interference in vegetation-covered areas, the denoised 3D spatial data units are obtained, and it is determined whether the denoised 3D spatial data units meet the preset surface information integrity standards. Based on the denoised 3D spatial data units, spatial reconstruction is performed on the denoised 3D spatial data units to meet the surface information coverage requirements in complex environments. If there are blank areas in the reconstructed data grid, data filling is performed to obtain the spatial dataset of the disaster area.

3. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, Constructing a terrain point cloud collection includes: Based on the spatial dataset of the disaster area, the data is processed in layers. Taking into account the different characteristics of surface signals and vegetation interference, the point cloud part related to terrain features is separated from at least one data layer to obtain a preliminary separated point cloud subset. For the initially separated point cloud subsets, vegetation interference and other non-topographic signals are filtered out. If the proportion of interference signals exceeds a preset threshold, the interference area is corrected to determine the filtered point cloud grouping. Based on the filtered point cloud groups, the spatial distribution is analyzed. To meet the requirement of continuity of terrain features, point cloud units related to the core data are extracted from at least one group to obtain refined terrain point cloud units. For the refined terrain point cloud units, the data is spatially reorganized. If data gaps are found in the grid, the gaps are filled to determine the terrain point cloud set.

4. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, Obtaining a 3D terrain model includes: Based on the terrain point cloud set, a comprehensive inspection of the terrain data is carried out. For at least one missing area, the relevant spatial distribution range is obtained from the information gaps to determine the priority range of area restoration. For missing areas within the priority range, data is used to fill the information gaps. Surrounding point cloud data is obtained from the gaps caused by vegetation occlusion and terrain occlusion to obtain a preliminary completed subset of point clouds. If the initially completed point cloud subset has uneven data distribution in some areas, the spatially reorganized data is adjusted to obtain a continuous distribution that conforms to the terrain features, and the adjusted point cloud grouping is determined. Based on the adjusted point cloud grouping, the terrain data is finally integrated. To ensure data integrity, relevant data units are extracted from the adjusted point cloud grouping to obtain a three-dimensional terrain model.

5. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, The segmented terrain data units include: Based on the 3D terrain model, the overall terrain is initially segmented. For at least one large area, the size of the segmented grid cells is determined according to a preset threshold, and the preliminary terrain data cells are obtained from them. For the initially divided terrain data units, the regional boundaries of each unit are identified. If uneven distribution of regional boundaries is detected, the units are re-divided to obtain terrain segmentation units with clear boundaries. Based on the clearly defined terrain segmentation units, the data distribution within each unit is detected. If the data distribution uniformity within a unit is lower than a preset threshold, the data within the unit is redistributed to determine a uniformly distributed set of block data. For a uniformly distributed set of segmented data, data integration processing is performed on all segmented units to extract unit-scale data that meets the processing scale requirements, thus obtaining segmented terrain data units.

6. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, The resulting set of denoised terrain data units includes: Based on the segmented terrain data units, the data in each unit is initially decomposed. For the data of at least one region, the part that may contain noise signals is separated out, and the initially separated data set is obtained from it. For the initially separated dataset, noise signals within each unit are identified. If the detected noise signal exceeds a preset threshold, the dataset is cleaned to obtain purified terrain data fragments. Based on the purified terrain data fragments, the regional distribution of each unit is compared. If uneven regional distribution is found, the boundaries of the fragments are adjusted to determine a reasonable combination of terrain data. For a reasonably distributed combination of terrain data, all block units are aggregated and processed, and the combination is finally calibrated to obtain a set of denoised terrain data units.

7. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, The key deformation feature set includes: Based on the denoised terrain data unit set, each unit is scanned one by one. If the detected data fluctuation exceeds the preset threshold, the terrain data unit is marked to obtain the marked unit group to be processed. For the labeled unit group to be processed, the local features within the terrain data unit are enhanced to extract hidden deformation-related information and identify feature segments containing potential deformation. Based on the feature segments containing potential deformations, the deformation information within the segments is classified and compared. If the significance of the deformation information is higher than a preset threshold, it is classified as a key deformation feature, and the classified key deformation feature group is obtained. For the classified key deformation feature groups, the key deformation features are summarized and organized, and feature sets related to terrain data are extracted from them to obtain the key deformation feature set.

8. The high-precision mapping method for disaster areas based on point cloud data as described in claim 1, characterized in that, Obtaining a deformation trend distribution map includes: Based on the key deformation feature set, relevant time series segments are extracted from the pre-established historical data archive. A preliminary comparison is made between the segments and the current features to obtain a feature difference set after the preliminary comparison. For the feature difference set, the deformation trend within the difference set is analyzed. If the direction and magnitude of change within the difference set exceed a preset threshold, it is marked as a significant change feature, and a significant change feature group is determined. Based on the significant change feature groups, the feature groups are correlated with spatial distribution information to obtain deformation trend distribution data and determine the change characteristics of the data in different regions. For deformation trend distribution data, the direction and magnitude of change are visualized to obtain a deformation trend distribution map.

Citation Information

Patent Citations

  • Geological disaster interpretation method, device, equipment and medium under steep terrain

    CN119478281A

  • Geological disaster deformation monitoring method based on airborne laser radar

    CN119714107A

  • Satellite remote sensing monitoring, early warning and forecasting methods for geological disasters

    CN119785530A

  • Landslide disaster area detecting device and landslide disaster area detecting method

    WO2023223483A1