Water conservancy earthwork dynamic monitoring method and system based on multi-source point cloud data fusion

Abstract

The invention relates to the technical field of water conservancy projects, in particular to a water conservancy earthwork dynamic monitoring method and system based on multi-source point cloud data fusion, and the method comprises the steps: carrying out the identification and standardization processing of airborne LiDAR, ground scanning and oblique photography point clouds, and generating a standard point cloud data set with uniform spatial positions and balanced density; dividing an abrupt slope area, a gentle slope area and a flat area based on topographic gradients, and constructing a grid framework adaptive to topographic features; carrying out weighted fusion by combining the point cloud coverage rate and the normalized terrain gradient to generate high-precision grid elevation data; calculating an earthwork variation distribution matrix by comparing historical elevation models; dynamically analyzing the change trend of the excavation and filling area, and synchronously evaluating the construction progress deviation and the geological disaster risk. According to the method, the problems of insufficient multi-source point cloud fusion precision, poor terrain adaptability and dynamic evaluation splitting in the prior art can be solved, and the monitoring precision and the engineering safety early warning efficiency under the complex terrain condition are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent measurement technology, in particular to a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method and system. BACKGROUND

[0002] Water conservancy earthwork project dynamic monitoring is a key link in large-scale water conservancy project construction and operation management, which provides important basis for engineering safety evaluation, construction quality control and disaster warning by continuously tracking the surface shape change. With the development of remote sensing technology, the monitoring method based on point cloud data gradually becomes mainstream, which obtains high-precision surface three-dimensional information through laser radar, photogrammetry and other devices, and realizes quantitative analysis of earthwork excavation, backfilling and other engineering activities. This monitoring technology not only needs to capture millimeter-level changes in the surface, but also needs to maintain data stability in complex terrain conditions to provide reliable basis for engineering decision-making.

[0003] However, the existing technology still faces significant challenges in practical application: on the one hand, multi-source point cloud data such as airborne LiDAR, ground scanning and oblique photography data have problems such as inconsistent coordinate systems and uneven density distribution due to differences in collection equipment and time phase, resulting in insufficient fusion result precision; on the other hand, traditional monitoring methods are difficult to effectively distinguish the influence of different terrain features such as steep slopes and gentle slopes on earthwork changes, and are prone to calculation deviations in complex terrain areas. In addition, existing methods often separate construction progress monitoring and geological disaster risk assessment, lack dynamic comprehensive analysis capability of earthwork change trend, and are difficult to find engineering progress deviation and potential landslide risk in time. These technical limitations make it difficult for current water conservancy project monitoring to meet the modern management needs of high precision and high efficiency. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method and system which can realize multi-source point cloud high-precision fusion, adaptive terrain feature analysis, and synchronous evaluation of construction progress and geological disaster risk.

[0005] The purpose of the present application is achieved by the following scheme:

[0006] In a first aspect, the present application provides a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method, comprising the following steps:

[0007] S1: identifying and standardizing the input original multi-source point cloud data set to generate a standard point cloud data set, the original multi-source point cloud data set including airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud;

[0008] S2: Perform terrain partitioning processing based on the standard point cloud dataset to generate a terrain classification distribution map, which is used to indicate the spatial distribution range of steep slope area, gentle slope area and flat area;

[0009] S3: Perform grid division processing on the terrain classification distribution map, adjust the grid density of the grid unit according to the terrain gradient difference of the terrain feature unit, and generate an adaptive grid framework with terrain labels;

[0010] S4: Fuse the standard point cloud dataset using the adaptive grid framework, calculate the weight coefficient combining the point cloud coverage and normalized terrain gradient of the grid unit, and generate a fused grid elevation dataset through fusion calculation;

[0011] S5: Calculate the earthwork change based on the fused grid elevation dataset, construct an elevation model for the engineering area, and compare the historical data of the historical monitoring period to generate an earthwork change distribution matrix;

[0012] S6: Perform dynamic analysis on the earthwork change distribution matrix, identify the excavation and filling area and calculate the change trend of the earthwork, and generate a dynamic monitoring evaluation report, which is used to indicate the construction progress deviation and geological disaster risk.

[0013] In a second aspect, the present application provides a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring system, which is configured with the following modules:

[0014] A multi-source point cloud preprocessing module is configured to perform identification and standardization processing on the input original multi-source point cloud dataset to generate a standard point cloud dataset, wherein the original multi-source point cloud dataset includes airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud.

[0015] A terrain partitioning processing module is configured to perform terrain partitioning processing based on the standard point cloud dataset to generate a terrain classification distribution map, which is used to indicate the spatial distribution range of steep slope area, gentle slope area and flat area.

[0016] An adaptive grid generation module is configured to perform grid division processing on the terrain classification distribution map, adjust the grid density of the grid unit according to the terrain gradient difference of the terrain feature unit, and generate an adaptive grid framework with terrain labels.

[0017] A point cloud fusion calculation module is configured to fuse the standard point cloud dataset using the adaptive grid framework, calculate the weight coefficient combining the point cloud coverage and normalized terrain gradient of the grid unit, and generate a fused grid elevation dataset through fusion calculation.

[0018] The earthwork change calculation module is configured to calculate the earthwork change based on the fused grid elevation data set, construct an elevation model of the engineering area, and compare historical data of a historical monitoring period to generate an earthwork change distribution matrix.

[0019] The dynamic monitoring and evaluation module is configured to perform dynamic analysis on the earthwork change distribution matrix, identify the excavation and filling area, and calculate the change trend of the earthwork volume to generate a dynamic monitoring and evaluation report, which is configured to indicate the construction progress deviation and geological disaster risk.

[0020] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements any of the above-mentioned water conservancy earthwork dynamic monitoring methods of multi-source point cloud data fusion when executing the computer program.

[0021] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement any of the above-mentioned water conservancy earthwork dynamic monitoring methods of multi-source point cloud data fusion.

[0022] In summary, the water conservancy earthwork dynamic monitoring method of multi-source point cloud data fusion provided by the present application can systematically solve the three core problems of insufficient data fusion accuracy, poor terrain adaptability and fragmented dynamic evaluation in the existing point cloud monitoring technology through the deep collaborative processing of multi-source point cloud data and the terrain adaptive analysis mechanism. At the data fusion level, the processing procedure based on device identification and coordinate standardization can achieve spatial consistency control of multi-source point cloud data, effectively eliminating the system error of airborne LiDAR, ground scanning and oblique photography data; the adaptive grid framework combined with the terrain gradient difference can realize fine partition management of steep slope area, gentle slope area and flat area, and significantly improve the complex terrain representation capability through different density grid division. At the core technology link, the dual weight fusion mechanism based on point cloud coverage rate and normalized terrain gradient can guarantee the accuracy and reliability of the fused grid elevation data set, so as to realize millimeter-level earthwork change detection; and the dynamic comparison and analysis of historical and current elevation models can capture the micro deformation characteristics of the engineering area. Finally, through the coupling analysis of excavation and filling area identification and change trend, the method can simultaneously generate construction progress deviation warning and geological disaster risk evaluation, so as to realize the transformation of engineering safety control from passive response to active intervention, thereby effectively improving the monitoring accuracy, enhancing the terrain adaptability, optimizing the dynamic decision, and providing full-cycle, multi-dimensional and quantifiable safety monitoring support for large-scale water conservancy projects.

[0023] In order to better understand and implement, the present application will be described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1A flowchart of a water conservancy earthwork dynamic monitoring method of multi-source point cloud data fusion provided by an embodiment of the application is shown in the figure.

[0025] Figure 2 A flowchart of generating a fused mesh elevation dataset provided by an embodiment of the application is shown in the figure.

[0026] Figure 3 A flowchart of generating an earthwork change amount distribution matrix provided by an embodiment of the application is shown in the figure.

[0027] Figure 4 A structural diagram of a water conservancy earthwork dynamic monitoring system of multi-source point cloud data fusion provided by another embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0028] In order to facilitate the understanding of the present application, the present application will be described more fully below with reference to the accompanying drawings. The preferred embodiments of the present application are shown in the drawings. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the application more thorough and comprehensive.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application belongs. The terms used in the specification of the application are only for the purpose of describing the specific embodiments and are not intended to limit the application. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0030] In one embodiment, as shown in Figure 1 A water conservancy earthwork dynamic monitoring method of multi-source point cloud data fusion is provided. In this embodiment, the method is applied to a terminal for illustration. It should be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and can be realized through the interaction of the terminal and the server. In this embodiment, the method includes the following steps:

[0031] S1: Identifying and standardizing the input original multi-source point cloud dataset to generate a standard point cloud dataset, wherein the original multi-source point cloud dataset includes airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud.

[0032] Specifically, the system performs identification processing on the airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud, records the corresponding acquisition device parameters of each data type, such as laser emission frequency, scanning angle, camera model, focal length, pixel size, etc., as well as acquisition timestamp and spatial position information, including latitude and longitude, altitude and projection coordinates, etc. At the same time, the system comprehensively cleans the obtained multi-source data. Preferably, the system can use statistical analysis algorithms to identify and eliminate noise points, which are usually represented as outliers with significantly abnormal distance from surrounding point clouds; at the same time, the system removes duplicate points in multi-source data, constructs a spatial index structure, and quickly searches and compares point cloud data to determine and delete duplicate point cloud data.

[0033] In terms of coordinate systems, the system can convert point cloud data in different coordinate systems one by one into a unified geographic coordinate system according to preset geographic coordinate conversion parameters, to ensure that all data are processed in the same spatial reference system. For data formats, the system converts airborne LiDAR point cloud from common LAZ or LAS format to universal XYZ or PCD format, ground scanning point cloud from E57 or PTX format, and oblique photography point cloud from OBJ or OSGB format to a unified format, so that subsequent software can be compatible.

[0034] In addition, the system standardizes the color information of the ground scanning point cloud and the oblique photography point cloud, adjusts the brightness, contrast and saturation of the color channel through a color correction algorithm, so that the color information of different data sources has consistency and comparability, providing conditions for subsequent color feature-based analysis. After a series of rigorous processing steps, the system generates a standard point cloud data set, providing high-quality basic data for subsequent fusion processing and analysis, and ensuring the accuracy and reliability of the entire monitoring process.

[0035] S2: Perform terrain partitioning processing based on the standard point cloud data set to generate a terrain classification distribution map, which is used to indicate the spatial distribution range of steep slope area, gentle slope area and flat area.

[0036] Specifically, the system extracts and analyzes the terrain feature information such as elevation, slope, and aspect in the standard point cloud dataset. Specifically, the system constructs an initial elevation model by calculating the elevation value of each point cloud data point, providing a basis for subsequent slope and aspect calculations. Subsequently, the system uses a differential algorithm based on the elevation model to calculate the slope value of each point cloud data point, i.e., the elevation change rate of the point relative to its surrounding neighborhood points. At the same time, by analyzing the directionality of the elevation change, the aspect information of each point is determined, i.e., the inclination direction of the slope surface. In the terrain classification process, the system divides the monitoring area into different terrain type regions such as steep slope area, gentle slope area, and flat area according to the preset terrain classification standard. For example, set the area with a slope greater than 30 degrees as the steep slope area, the area with a slope between 10 degrees and 30 degrees as the gentle slope area, and the area with a slope less than 10 degrees as the flat area. The system judges the terrain features of each point cloud data point through computer algorithm, and statistically analyzes the spatial range and distribution of each region.

[0037] In the judgment process, the system can use a rule-based classification algorithm to match the slope, aspect, and other features of each point cloud data point with the preset classification standard to determine its terrain type. At the same time, the system performs spatial connectivity analysis on the classification results to ensure that the same terrain type region has spatial coherence and integrity, avoiding isolated classification points or regions. Finally, the system generates a terrain classification distribution map that intuitively and accurately reflects the spatial distribution range of different terrain types in the monitoring area. The map clearly identifies the location and range of various terrain regions such as steep slope area, gentle slope area, and flat area with different colors or legends, providing a key basis for subsequent differentiated grid division and fusion processing according to different terrain features, ensuring that subsequent steps can be targeted according to the characteristics of different terrain types for processing and analysis, improving the accuracy and efficiency of the entire monitoring system.

[0038] S3: Perform grid division processing on the terrain classification distribution map, adjust the grid density of the grid unit according to the terrain gradient difference of the terrain feature unit, and generate an adaptive grid framework with terrain labels.

[0039] Specifically, since the terrain gradient difference of different terrain feature units has a certain impact on data accuracy and analysis result accuracy, the system appropriately reduces the size of the grid unit and increases the grid density for steep slope areas with larger terrain gradients during the division process, in order to more meticulously capture terrain changes and earthwork changes and improve monitoring accuracy. For gentle slope areas and flat areas with smaller terrain gradients, the size of the grid unit is appropriately increased and the grid density is reduced, ensuring a certain monitoring accuracy while reducing data volume and computational complexity, and improving processing efficiency.

[0040] The system determines the grid division strategy for each region by analyzing the terrain gradient information in the terrain classification distribution map. In steep slope areas, the system uses a high-resolution grid division algorithm to divide the region into smaller grid cells, such as setting the grid size to 1 meter x 1 meter or smaller, to accurately capture the subtle changes of the slope surface. In gentle slope areas, the grid size can be appropriately increased to 2 meters x 2 meters to 3 meters x 3 meters, balancing the accuracy and efficiency of different regional units. In flat areas, the grid size can be further expanded to 5 meters x 5 meters or larger to reduce data volume and computational load. At the same time, the system ensures that the boundaries of the grid cells adapt to the terrain classification boundaries during the grid division process, avoiding the situation where grid cells cross different terrain types, thereby ensuring the relative consistency of terrain features within each grid cell.

[0041] After completing the grid division, the system associates each grid cell with the corresponding terrain type label, such as steep slope, gentle slope, or flat, and generates an adaptive grid framework with terrain labels. This framework not only contains the geographical location information and size attributes of the grid cells, but also records the terrain type of each grid cell in detail, providing a refined spatial framework structure for subsequent point cloud data fusion processing using the adaptive grid framework, ensuring that the fusion processing can be optimized according to the characteristics of different terrain types, improving the adaptability and accuracy of the entire monitoring system.

[0042] S4: Use the adaptive grid framework to perform fusion processing on the standard point cloud dataset, calculate the weight coefficients based on the point cloud coverage and normalized terrain gradient of the grid cells, and generate a fused grid elevation dataset through fusion calculation.

[0043] Specifically, the system classifies the point cloud data in the standard point cloud dataset according to the grid cells they belong to, and counts the distribution of point cloud data from various data sources such as airborne LiDAR point cloud, ground scanning point cloud, and oblique photography point cloud within each grid cell. The system quickly determines the grid cell to which each point cloud data point belongs through a spatial indexing algorithm, and establishes the correspondence between the data and the grid cells. During the classification process, the system records the number, coordinates, elevation, and other information of each data source point cloud data within each grid cell, providing data support for subsequent coverage calculation. The system calculates the coverage of each data source point cloud in the grid cell, and the formula for calculating the coverage is the number of point cloud data of each data source in the grid cell divided by the total number of all point cloud data in the grid cell. For example, for a certain grid cell, if there are 100 data points of airborne LiDAR point cloud, 150 data points of ground scanning point cloud, and 50 data points of oblique photography point cloud, then the coverage of airborne LiDAR point cloud is 100 / (100+150+50) = 0.333, the coverage of ground scanning point cloud is 0.5, and the coverage of oblique photography point cloud is 0.167.

[0044] At the same time, the system combines the terrain label of the grid cell to calculate the normalized terrain gradient value of each grid cell by using the normalization method. The calculation method of the normalized terrain gradient value is to calculate the ratio of the terrain gradient value of the grid cell to the maximum terrain gradient value of the entire monitoring area, so that the normalized terrain gradient value is between 0 and 1. For example, if the maximum terrain gradient of the entire monitoring area is 45 degrees, and the terrain gradient of a certain grid cell is 20 degrees, then the normalized terrain gradient value of the grid cell is 20 / 45≈0.444. Subsequently, the system calculates the weight coefficient of each data source in the grid cell according to the preset weight coefficient calculation model, taking the point cloud coverage and the normalized terrain gradient of the grid cell as input variables. The calculation model of the weight coefficient considers factors such as the coverage, accuracy, and terrain adaptability of the data source.

[0045] For example, for the grid cell in the steep slope area, the ground scanning point cloud has higher detail capture capability and accuracy in the steep slope area, so its weight coefficient is relatively high; for the grid cell in the flat area, the oblique photography point cloud has a wide data coverage and rich texture information, so its weight coefficient may be larger. Finally, the system performs weighted fusion calculation on the point cloud data of each data source according to the calculated weight coefficient. The system performs fusion calculation on the point cloud data in each grid cell to generate a fused grid elevation dataset. This dataset integrates the advantages of multi-source point cloud data in each grid cell, effectively improving the data accuracy and reliability, providing a high-quality data basis for subsequent earthwork change calculation processing, and ensuring that the calculation result of the earthwork change can accurately reflect the actual earthwork change in the engineering area, providing reliable data support for engineering decision-making.

[0046] S5: Calculate the earthwork change based on the fused grid elevation dataset, construct an elevation model of the engineering area, and compare the historical data of the historical monitoring period to generate an earthwork change distribution matrix.

[0047] Specifically, the system uses an elevation interpolation algorithm to construct an elevation model of the engineering area. The elevation interpolation algorithm analyzes the elevation value of each grid cell in the fused grid elevation dataset, combines the geographical location and spatial relationship of the grid cell, and uses methods such as polynomial fitting or spline interpolation to generate a continuous elevation model that can accurately reflect the surface morphology and subtle changes of the engineering area. The resolution of the model matches the density of the adaptive grid framework, ensuring that it can capture the details of surface changes in different terrain areas. At the same time, the system retrieves historical data of the historical monitoring period from the database. These historical data are also stored in the form of elevation models or earthwork data. The system accurately compares and analyzes the currently constructed elevation model with the historical data, and calculates the earthwork change distribution matrix using numerical integration methods.

[0048] During the calculation process, the system fully considers the characteristics and laws of earthwork changes in different terrain areas. For steep slope areas, smaller integration step and finer calculation grid are adopted to accurately capture the dramatic earthwork changes. For gentle slope and flat areas, the integration step is appropriately increased to improve the processing efficiency while ensuring the calculation accuracy. The system also classifies and analyzes the calculation results according to the terrain type and the size of earthwork change, and generates a detailed earthwork change distribution matrix. This matrix intuitively shows the earthwork change situation at different positions in the entire monitoring area, including the specific values and spatial distribution of excavation and filling quantities, providing key data support for subsequent dynamic analysis and evaluation, and ensuring that the monitoring results can accurately reflect the actual situation of earthwork changes in the engineering area, providing reliable basis for engineering management and decision-making.

[0049] S6: Perform dynamic analysis on the earthwork change distribution matrix, identify excavation and filling areas, and calculate the change trend of earthwork quantity, and generate a dynamic monitoring and evaluation report. The dynamic monitoring and evaluation report is used to indicate the construction progress deviation and geological disaster risk.

[0050] Specifically, the system performs cluster analysis on the earthwork change data in the matrix to identify excavation and filling areas. This algorithm aggregates regions with similar change characteristics by calculating the local density of earthwork change data for each grid cell. According to the positive and negative values of the earthwork change, the system divides the regions into excavation and filling areas and determines the boundaries and ranges of each area. When determining the boundaries, the system can use a boundary tracking algorithm to track each point along the edge of the clustered region, ensuring the accuracy and integrity of the boundary delineation.

[0051] On this basis, the system combines with the monitoring time series, and uses linear regression, moving average or exponential smoothing method to calculate the change trend of earthwork. The system analyzes the change rate of earthwork over time, that is, the increase or decrease amplitude of earthwork per unit time; the change amplitude, that is, the absolute value of the change of earthwork; and the periodic characteristics of the change, such as whether there is seasonal fluctuation or construction stage change rule, etc. In order to combine the analysis results with the engineering construction plan and the geological disaster risk assessment, the system reads the construction progress plan data from the engineering management database, compares the actual earthwork change with the plan, calculates the construction progress deviation, and determines which areas are ahead of or lagging behind the construction progress. At the same time, the system calls the geological disaster risk assessment model, and according to the earthwork change rate, the change amount size and the topographic and geological conditions, etc., the potential landslide and other geological disaster risk areas are warned and assessed, and the risk level and warning information are generated. Finally, the system integrates all the analysis results to generate a comprehensive dynamic monitoring and evaluation report. The report records in detail the specific location and degree of construction progress deviation, as well as the warning information of potential geological disaster risk areas, providing scientific, comprehensive and timely decision-making basis for the construction and operation management of water conservancy projects, helping engineering management personnel to accurately control the engineering progress and safety risk, and ensuring the smooth construction and long-term stable operation of water conservancy projects.

[0052] In summary, the multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the present application can systematically solve the three core problems of insufficient data fusion accuracy, poor terrain adaptability and fragmented dynamic evaluation in the existing point cloud monitoring technology through deep collaborative processing of multi-source point cloud data and terrain adaptive analysis mechanism. At the data fusion level, the processing flow based on device identification and coordinate standardization can achieve spatial consistency control of multi-source point cloud data, effectively eliminating the system error of airborne LiDAR, ground scanning and oblique photography data; the adaptive grid framework combined with terrain gradient difference can realize fine partition management of steep slope area, gentle slope area and flat area, and significantly improve the complex terrain representation capability through different density grid division. At the core technology link, the dual weight fusion mechanism based on point cloud coverage rate and normalized terrain gradient can guarantee the accuracy and reliability of the fusion grid elevation data set, in order to realize millimeter-level earthwork change detection; and the dynamic comparison and analysis of historical and current elevation models can capture the micro deformation characteristics of the engineering area. Finally, through the coupling analysis of excavation and filling area recognition and change trend, the method can simultaneously generate construction progress deviation warning and geological disaster risk assessment, in order to realize the transformation of engineering safety control from passive response to active intervention, thereby effectively improving the monitoring accuracy, enhancing the terrain adaptability, optimizing the dynamic decision-making, and providing full-cycle, multi-dimensional and quantifiable safety monitoring support for large-scale water conservancy projects.

[0053] In one of the embodiments, the application provides a water conservancy earthwork dynamic monitoring method based on multi-source point cloud data fusion, which specifically comprises the following steps:

[0054] S11: Each set of point cloud data in the input original multi-source point cloud data set is identified and processed, a unique identifier is constructed according to the type of the collection device and the collection time, and a traceable point cloud data set is generated.

[0055] Specifically, in the process of water conservancy earthwork dynamic monitoring, the system receives an original multi-source point cloud data set containing airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud. The system constructs a unique identifier for each set of point cloud data according to its collection device type and collection time. For airborne LiDAR point cloud, the system extracts the serial number of the device, the laser emission frequency, the scanning angle and other parameters, and combines the time stamp of the collection task to generate an identifier according to the preset encoding rule; for ground scanning point cloud, the system encodes the model of the scanner, the coordinates of the erected position and the scanning start time and other information; and for oblique photography point cloud, the corresponding identifier is generated according to the serial number of the camera, the flight attitude parameters and the shooting time. The system binds these identifiers with the corresponding point cloud data to generate a traceable point cloud data set. This step not only ensures the traceability of the data, facilitates subsequent data quality inspection and problem tracing, but also provides accurate data source information for subsequent data fusion processing, ensuring the reliability of the fusion results. When generating the traceable point cloud data set, the system also establishes a detailed metadata database to record the identifier of each set of point cloud data, the collection device parameters, the collection time, the collection position and other information, providing support for subsequent data management and analysis.

[0056] S12: The traceable point cloud data set is processed in the coordinate system alignment, and the airborne LiDAR point cloud, the ground scanning point cloud and the oblique photography point cloud are unified to the engineering coordinate system to generate a spatially aligned point cloud data set.

[0057] Specifically, the system first reads the original coordinate system information of each set of point cloud data, including the geographic coordinate system of airborne LiDAR point cloud (such as WGS-84), the local coordinate system of ground scanning point cloud, and the image space coordinate system of oblique photography point cloud, etc. The system converts the point cloud data in different coordinate systems into a unified engineering coordinate system one by one according to the preset coordinate conversion parameters. For airborne LiDAR point cloud, the system uses a seven-parameter conversion model to accurately calculate and adjust its coordinates in combination with control point data; for ground scanning point cloud, the system uses a similarity transformation algorithm for coordinate conversion through matching with known control points; and for oblique photography point cloud, the system uses aerial triangulation results and ground control points for strict geometric rectification and coordinate conversion. During the conversion process, the system accurately calculates the coordinate values of each point cloud point to ensure that the converted coordinate accuracy meets the engineering requirements. The system also checks the quality of the converted point cloud data, verifies the accuracy of the coordinate conversion by calculating the residual error of the point cloud data before and after conversion, checking the integrity and consistency of the data, etc. After a series of rigorous processing steps, the system generates a spatially aligned point cloud dataset, providing a unified spatial reference framework for subsequent density adjustment processing and fusion analysis, and ensuring the accurate spatial alignment of point cloud data from different sources.

[0058] S13: Perform density adjustment processing on the spatially aligned point cloud dataset to balance the density differences of point clouds from different sources, and generate a standard point cloud dataset indicating point cloud information with unified spatial position and balanced density.

[0059] Specifically, after completing the coordinate system alignment, the system performs density adjustment processing on the spatially aligned point cloud dataset to balance the density differences of point clouds from different sources. The system analyzes the density distribution characteristics of airborne LiDAR point cloud, ground scanning point cloud, and oblique photography point cloud. Airborne LiDAR point cloud generally has a relatively uniform density distribution, but its density is relatively low due to the influence of flight height and laser emission frequency; ground scanning point cloud has a relatively high density within the scanning range, but may have local density unevenness; the density distribution of oblique photography point cloud is greatly affected by flight attitude and shooting angle.

[0060] Preferably, the system can apply an adaptive filtering algorithm to the ground scanning point cloud with high density for down-sampling processing, determine the down-sampled point cloud points according to the preset density threshold and the spatial distribution characteristics of the point cloud, and ensure that the density of the ground scanning point cloud matches the density of the airborne LiDAR point cloud. At the same time, the system performs up-sampling processing on the airborne LiDAR point cloud with low density, generates new point cloud points in the sparse area of the point cloud data by using an interpolation-based algorithm, and fills the density gap. For the oblique photography point cloud, the system combines its image texture information and depth information, and uses a multi-view geometry algorithm to optimize and adjust the point cloud density, so that the density distribution is more uniform and coordinated with the densities of the airborne LiDAR point cloud and the ground scanning point cloud. During the adjustment process, the system monitors the density change of the point cloud data in real time, calculates the density statistical indicators and visualizes the density distribution diagram to ensure that the adjusted point cloud data density is balanced and meets the engineering requirements. After the density adjustment processing, the system generates a standard point cloud data set, which not only realizes the unity in spatial position, but also achieves the balance in density distribution, providing a high-quality data basis for subsequent fusion processing and analysis, and ensuring the accuracy and reliability of the entire monitoring process.

[0061] In one of the embodiments, the S2 of the multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the application specifically comprises the following steps:

[0062] S21: Calculate the terrain gradient of the standard point cloud data set, calculate the slope value of each point based on the elevation of the spatial neighborhood points, and generate terrain gradient data.

[0063] Preferably, the system first preprocesses the standard point cloud data set and constructs a spatial index structure to improve the neighborhood point retrieval efficiency. For each point cloud point, the system defines a spatial neighborhood with a certain radius centered on the point, usually set to 1-2 meters, to balance the calculation accuracy and efficiency. Within the neighborhood, the system extracts the elevation values of all adjacent point cloud points. In order to calculate the slope value, the system uses the elevation difference method to calculate the elevation difference between the center point and each adjacent point in the neighborhood, and combines the horizontal distance between the two points to calculate the slope value through the trigonometric function.

[0064] In order to improve the reliability of the calculation results, the system statistically analyzes the slope values of all adjacent points in the neighborhood, and calculates the average slope value as the final slope value of the center point. In this process, the system also performs quality control on the calculation results to eliminate abnormal slope values caused by outlier points or noise points. After this series of rigorous calculation steps, the system generates detailed terrain gradient data, which provides accurate slope information for subsequent terrain zoning processing, and ensures that the slope value of each point cloud point can accurately reflect the inclination of the ground surface.

[0065] S22: Region division processing is performed on the slope value of each point in the terrain gradient data, and the region with a slope value less than a preset first threshold value is divided into a flat area, the region with a slope value between the preset first threshold value and a preset second threshold value is divided into a gentle slope area, and the region with a slope value greater than the preset second threshold value is divided into a steep slope area, to generate a terrain partition division result.

[0066] Preferably, the system performs region division processing on the slope value of each point based on the generated terrain gradient data. Before division, the construction personnel sets two key slope threshold values according to engineering experience and field terrain characteristics: the first threshold value is usually set at 5-10 degrees to distinguish between flat areas and gentle slope areas; and the second threshold value is set at about 30-40 degrees to distinguish between gentle slope areas and steep slope areas. In the division process, the system analyzes the slope value of each point cloud point point by point. For the region with a slope value less than the first threshold value, the system determines it as a flat area, and the ground surface of this region is relatively flat, and the earthwork changes relatively small; for the region with a slope value between the first threshold value and the second threshold value, the system divides it into a gentle slope area, and the ground surface of this region has a certain inclination, but the earthwork changes relatively gently; and for the region with a slope value greater than the second threshold value, it is determined as a steep slope area, and the ground surface of this region is steep, and the earthwork changes significantly, which is usually the key area for monitoring.

[0067] In the division process, the system can use a spatial connectivity analysis algorithm to ensure that the point cloud points in the same region have spatial connectivity, avoiding isolated division results. At the same time, the system optimizes the division boundary, uses moving average filtering and other methods to smooth the boundary transition area, so that the region division result is more in line with the actual terrain characteristics. After this processing step, the system generates a detailed terrain partition division result, which provides an accurate region division basis for subsequent terrain classification distribution map generation, and ensures that the division result of different terrain regions can accurately guide the subsequent monitoring and analysis work.

[0068] S23: Based on the terrain partition division result, a standard point cloud data set is used to generate a distribution map, and the boundaries of the steep slope area, the gentle slope area and the flat area are labeled to generate a terrain classification distribution map.

[0069] Preferably, the system calls geographic information system (GIS) software to combine the terrain zoning division results with the spatial coordinate information of the standard point cloud dataset. Specifically, the system assigns each point cloud point with a corresponding terrain type attribute, i.e., steep slope area, gentle slope area, or flat area, according to the terrain zoning division results. When generating the distribution map, the system can use vector rasterization technology to aggregate point cloud points with the same terrain type attribute into continuous areas and distinguish them with different colors or patterns. Steep slope areas are usually represented in red, gentle slope areas in yellow, and flat areas in green. At the same time, the system also accurately depicts the boundaries of each area by calculating the spatial positional relationship of adjacent area point cloud points to determine the orientation and shape of the boundary line and store and display it using line elements.

[0070] To improve the readability and practicality of the distribution map, the system adds map elements such as legend, scale, and compass to the map and labels each terrain area to display basic information such as area and average slope. In addition, the system checks the quality of the generated terrain classification distribution map by comparing the original point cloud data and the terrain zoning division results to verify the accuracy and completeness of the distribution map. If there is a deviation, the system will automatically adjust the color and boundary to ensure that the distribution map can truly and intuitively reflect the terrain classification of the monitoring area, providing reliable spatial data support for subsequent earthwork change analysis and engineering decision-making.

[0071] In one embodiment, the present application provides a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method S3 specifically includes the following steps:

[0072] S31: Perform dense grid division processing on the steep slope area in the terrain classification distribution map, using grid cells smaller than a preset first size to generate steep slope area grid cells.

[0073] Preferably, the construction personnel pre-set a smaller grid size in the system according to the complexity of the terrain in the steep slope area and the density of the point cloud data, usually set at about 0.5-1 meter, to ensure that subtle changes in the terrain of the steep slope area can be captured. The system calls an adaptive grid division algorithm that subdivides the steep slope area into multiple grid cells based on the boundary information of the steep slope area and the internal point cloud distribution characteristics. The size of each grid cell is strictly controlled below the pre-set first size. During the division process, the system fully considers the terrain discontinuity and elevation abruptness of the steep slope area, and generates smaller size grid cells at the terrain abruptness by introducing buffer zone analysis technology, to better adapt to the terrain changes. The system also constructs the topological relationship of the divided grid cells, records the adjacent relationship and spatial position information of each grid cell, and ensures the integrity and connectivity of the grid division. The system monitors the point cloud data distribution in the grid cell in real time during the division process, and verifies the rationality of the grid division by calculating the point cloud coverage rate and elevation standard deviation in each grid cell. If necessary, the system will automatically adjust the grid boundary to optimize the division result. After this processing step, the system generates a steep slope area grid cell, which provides a fine spatial framework for subsequent high-precision earthwork change analysis, ensuring that the earthwork change in the steep slope area can be accurately captured and quantified. Through accurate grid division, the system can achieve high-resolution terrain modeling in the steep slope area, providing a reliable data basis for subsequent earthwork calculation and dynamic monitoring.

[0074] S32: Perform medium-density grid division processing on the gentle slope area in the terrain classification distribution map, divide it with grid cells larger than the pre-set first size and smaller than the pre-set second size, and generate gentle slope area grid cells.

[0075] Preferably, the system performs medium-density grid division processing on the gentle slope area in the terrain classification distribution map. The construction personnel pre-sets a medium size range in the system according to the terrain characteristics of the gentle slope area and the monitoring accuracy requirements, which is larger than the pre-set first size and smaller than the pre-set second size, usually set between 1-3 meters. The system uses an optimized quadtree grid division algorithm to divide the gentle slope area into multiple grid cells in combination with the point cloud data density and terrain gradient information of the gentle slope area. The algorithm dynamically adjusts the grid size during the division process to ensure that larger grid sizes are used in areas with relatively slow terrain changes, and smaller grid sizes are used in areas with slightly larger local terrain changes, to balance data accuracy and processing efficiency.

[0076] During the partitioning process, the system monitors the distribution of point cloud data within each grid cell in real-time. By calculating the point cloud coverage and elevation standard deviation within each grid cell, the system verifies the rationality of the grid partitioning. If necessary, the system will automatically adjust the grid boundaries to optimize the partitioning results. The system also constructs the topological relationship of the partitioned grid cells, records the adjacent relationship and spatial location information of each grid cell, and ensures the integrity and connectivity of the grid partitioning. Finally, the system generates the gentle slope area grid cells, which can meet the accuracy requirements of the gentle slope area earthwork change monitoring, effectively control the data volume, and improve the subsequent processing efficiency, providing an adaptive spatial framework for the gentle slope area earthwork change analysis. Through this medium-density grid partitioning, the system can efficiently model the terrain in the gentle slope area, providing a reliable data foundation for subsequent earthwork calculation and dynamic monitoring.

[0077] S33: Perform sparse grid partitioning processing on the flat area in the terrain classification distribution map, divide it into grid cells with a size larger than the preset second size, and generate flat area grid cells.

[0078] Preferably, according to the characteristics of the flat area, such as relatively flat ground surface and small earthwork change, the construction personnel pre-set a larger grid size in the system, which is usually larger than the preset second size and is set in the range of 3-10 meters. The system calls the rule-based grid partitioning algorithm to divide the flat area into regular grid cells according to the pre-set grid size. During the partitioning process, the system fully considers the boundary conditions and internal point cloud distribution uniformity of the flat area, calculates the spatial density and elevation consistency of the point cloud data to ensure the rationality of the grid partitioning. The system also simplifies the partitioned grid cells to remove redundant grid boundary information, further reducing data complexity and improving processing efficiency.

[0079] During the partitioning process, the system monitors the distribution of point cloud data within each grid cell in real-time. By calculating the point cloud coverage and elevation standard deviation within each grid cell, the system verifies the rationality of the grid partitioning. If necessary, the system will automatically adjust the grid boundaries to optimize the partitioning results. The system also constructs the topological relationship of the partitioned grid cells, records the adjacent relationship and spatial location information of each grid cell, and ensures the integrity and connectivity of the grid partitioning. After this processing step, the system generates the flat area grid cells, which have a sparse grid layout that can reflect the basic terrain characteristics of the flat area while significantly reducing the data volume, providing an efficient foundation framework for subsequent large-scale data processing and analysis. Through this sparse grid partitioning, the system can efficiently manage and analyze data in flat areas, providing a reliable data foundation for subsequent earthwork calculation and dynamic monitoring.

[0080] S34: The steep slope area grid cells, the gentle slope area grid cells and the flat area grid cells are integrated, the topographic type labels of each grid cell are marked, and the adaptive grid framework with topographic labels is generated.

[0081] Preferably, the system accurately splices the grids of different areas according to the geographical position information and topological relationship of each grid cell, ensuring the continuity and integrity of the entire grid framework. The system uses spatial indexing technology to quickly search and manage the integrated grid framework, improving data processing efficiency. Subsequently, the system assigns each grid cell with a corresponding topographic type label, extracts the corresponding topographic type label from the predefined label library according to the area to which the grid cell belongs, and binds it with the grid cell. The label information not only contains the topographic type, but also includes the attribute information of the grid cell, such as grid size, center coordinates, elevation range, etc. The system also conducts quality inspection on the labeled grid framework, verifies the accuracy and integrity of the labeling by comparing the original topographic classification distribution map and the labeling information. If there is a deviation, the system will automatically adjust the label information to ensure that the topographic type label of each grid cell is consistent with the actual topographic features. Through this integration and labeling process, the system generates an adaptive grid framework with topographic labels, providing a unified, accurate and richly topographic information spatial data basis for subsequent point cloud fusion processing and earthwork change analysis, ensuring that the monitoring results can fully consider the influence of different topographic features, improving the scientificity and reliability of the analysis. The system strictly follows the preset algorithm and rules in the entire processing process, ensuring the accuracy and efficiency of each step, providing solid data support for subsequent earthwork calculation and dynamic monitoring.

[0082] In one embodiment, as shown in Figure 2 S4 of the multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the application specifically includes the following steps:

[0083] S41: Each grid cell in the adaptive grid framework is subjected to weight calculation processing, and a fusion weight coefficient is generated.

[0084] Specifically, the fusion weight coefficient is obtained by the following calculation formula:

[0085]

[0086] wherein ω is a fusion weight coefficient, α and β are preset weight factors, and α + β = 1, C is an effective point cloud coverage in the grid cell, determined by the ratio of the number of effective point cloud points in the grid cell to the total number of point cloud points, reflecting the density and integrity of the point cloud data in the region; G is a normalized terrain gradient value of the grid cell, obtained by calculating the average value of the terrain gradient in the grid cell and normalizing it to the interval [0, 1], used to measure the relative degree of change of the terrain. In the calculation process, the system first obtains the effective point cloud coverage C of each grid cell, then calculates the normalized terrain gradient value G, and finally performs weighted summation on C and 1 / G according to the preset weight factors α and β to obtain the fusion weight coefficient ω. The weight coefficient comprehensively considers the coverage and terrain gradient factors of the point cloud data, providing a scientific weight basis for subsequent point cloud fusion processing, and ensuring that different grid cells can play a reasonable role in the fusion process according to their data quality and terrain characteristics.

[0087] S42: Perform weighted fusion processing on the standard point cloud dataset to calculate the fusion elevation value of each grid cell.

[0088] Specifically, preferably, the calculation formula of the fusion elevation value of the grid cell is as follows:

[0089]

[0090] wherein H is the fusion elevation value, w is the weight coefficient of the i th point cloud in the standard point cloud dataset, H is the elevation value of the i th point cloud, and n is the number of point clouds in the grid. In the calculation process, the system first obtains the weight coefficient w of each point cloud data, which is calculated according to the fusion weight coefficient formula and reflects the reliability and representativeness of the point cloud data in the grid cell. Then, the system multiplies the elevation value H of each point cloud data with its corresponding weight coefficient w to obtain the weighted elevation value. Next, the system sums all the weighted elevation values and divides by the sum of all the weight coefficients to obtain the fusion elevation value H. The fusion elevation value comprehensively considers the weights of different point cloud data and can more accurately reflect the actual terrain elevation in the grid cell, providing a high-precision elevation data basis for subsequent earthwork change analysis. fused i i i i i fused

[0091] S43: Perform dataset generation processing on the fusion elevation values of all grid cells to form a fusion grid elevation dataset.

[0092] ​​​​​​​Specifically, the system combines the fused elevation values of each grid cell with the spatial location information of that grid cell to construct a complete dataset. During the generation process, the system assigns each fused elevation value with the geographic coordinate information of its corresponding grid cell, ensuring that the dataset has accurate spatial reference. At the same time, the system performs format conversion and organization on the data to meet the requirements of subsequent analysis and storage. The system also performs quality checks on the generated fused grid elevation dataset to verify the integrity and consistency of the data. By comparing the original point cloud data and the fused elevation values, it checks for the presence of outliers or missing values and makes necessary corrections and supplements to the data. Finally, the system generates a fused grid elevation dataset that not only contains the elevation information of each grid cell but also has a unified spatial reference and data format, providing a high-quality data foundation for subsequent earthwork volume calculation and dynamic analysis, ensuring the accuracy and reliability of the entire monitoring process.

[0093] The multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided above can significantly improve the monitoring accuracy under complex terrain conditions through the weight optimization mechanism of multi-source point cloud fusion. In the weight calculation stage, the double-factor fusion formula based on point cloud coverage and normalized terrain gradient can achieve adaptive weight distribution of the terrain, effectively balancing the contradiction between sparse data in steep slope areas and redundant data in flat areas; the weighted fusion processing can realize the collaborative optimization of multi-source point cloud data through normalized weight coefficients, so as to eliminate the system deviation of airborne LiDAR, ground scanning and oblique photography data. The final generated elevation dataset can guarantee the millimeter-level precision of the grid elevation representation, providing a reliable data foundation for earthwork volume calculation. Through the coupling of mathematical modeling and terrain features, the present application can realize efficient fusion of multi-source heterogeneous point clouds in complex water conservancy engineering scenarios, overcoming the inherent defects of existing point cloud monitoring methods such as accuracy drop in steep slope areas and data oversaturation in flat areas, and establishing an accurate spatial reference for subsequent dynamic analysis.

[0094] In one embodiment, as shown in Figure 3 The multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the present application specifically includes the following steps in S5:

[0095] S51: Process the fused grid elevation dataset to construct an elevation model with the elevation values and spatial coordinates of the grid cells, and generate the current digital elevation model of the engineering area.

[0096] Specifically, the system first reads the elevation values and spatial coordinate information of each grid cell in the dataset, taking these data points as the basis for constructing a digital elevation model (DEM). Preferably, the system can employ an interpolation algorithm, such as Kriging interpolation or inverse distance weighted interpolation, to estimate the elevations of the areas between grid cells, generating a continuous elevation surface. During the interpolation process, the system compares the generated elevation surface with the original grid cell data, taking into account the elevation values of multiple surrounding grid cells, to ensure that the generated elevation surface is smooth and consistent with the actual topographic features, ensuring the accuracy and reliability of the model. Finally, the system constructs the current digital elevation model of the engineering area, which presents the surface morphology of the engineering area in two or three dimensions, providing basic data support for subsequent elevation change analysis. The system automatically records the time of model generation and the algorithm parameters used, for subsequent tracing and optimization.

[0097] S52: Change quantity calculation and processing of the current digital elevation model and the digital elevation model of the historical monitoring period, generating the elevation change quantity of each grid cell.

[0098] Specifically, the system performs spatial registration on the current digital elevation model and the digital elevation model of the historical monitoring period, ensuring that they are compared under the same coordinate system and spatial resolution. This process involves coordinate transformation and resampling of both models to eliminate matching errors caused by inconsistent coordinate systems or resolution differences. The system uses high-precision coordinate transformation algorithms, such as a seven-parameter transformation model, to convert the historical model and the current model to a unified coordinate system. At the same time, in order to make the two models have the same resolution, the system will downsample the model with higher resolution, or upsample the model with lower resolution, to ensure that each grid cell can be accurately corresponded in subsequent comparative analysis. After completing the spatial registration, the system calculates the elevation change quantity of each grid cell. Preferably, the formula for calculating the elevation change quantity of a grid cell is as follows:

[0099] ΔH = H current -H historical

[0100] Where ΔH is the elevation change quantity of the grid cell, H current is the current monitoring period elevation value, H historicalFor historical monitoring period elevation values. In the calculation process, the system will strictly check the accuracy of the elevation value of each grid cell, and through comparison with the original point cloud data, exclude the calculation deviation caused by data error or abnormal value. For example, if the elevation change of a certain grid cell exceeds the preset reasonable range (such as ±5 meters), the system will automatically mark the unit and determine whether it is an abnormal value through cross verification with the original point cloud data. For grid cells determined to be abnormal values, the system will use the average value of adjacent grid cells for correction to ensure the reliability of the calculation results. Finally, the system generates a dataset containing the elevation change of all grid cells, which records the ground elevation change of the engineering area in the monitoring period in detail, providing key data support for subsequent earthwork change analysis. The system automatically records the process and results of the change calculation, including the algorithm used, parameter settings and abnormal value processing, ensuring the traceability of the data and the accuracy of the analysis.

[0101] S53: Perform distribution matrix generation processing on the engineering area based on the elevation change of all grid cells, map the elevation change data according to the grid location, and form the earthwork change distribution matrix.

[0102] Specifically, the system creates a matrix structure that matches the spatial layout of the engineering area, with rows and columns corresponding to the coordinate axes in geographical space, and each element corresponding to the elevation change of a grid cell. The system accurately maps and organizes the elevation change data according to the spatial location of each grid cell, filling the matrix with the corresponding position.

[0103] To ensure the integrity and consistency of the data, the system will check each element in the matrix to ensure there are no missing values or incorrect data. If it is found that the data of a certain grid cell is missing or abnormal, the system will automatically use the average value or interpolation result of adjacent grid cells to supplement and correct it. During the process of filling the matrix, the system also standardizes the data, converting the elevation change to a unified dimension and unit to facilitate subsequent analysis and visualization. To enhance the readability and analyzability of the distribution matrix, the system visualizes the matrix data, using different colors or symbols to represent different degrees of elevation change.

[0104] For example, the system defines a color mapping table, representing areas with increased elevation in blue, areas with decreased elevation in red, and setting the color depth according to the size of the change. The deeper the color, the greater the change. At the same time, the system also generates related legends and annotations to clearly indicate the range of elevation change represented by the color or symbol. In addition, the system will perform data validation on the distribution matrix, comparing it with the original elevation change data set to check for inconsistencies or excessive errors in the matrix. If data quality problems are found, the system updates and improves the matrix. The final generated earthwork change distribution matrix not only visually displays the earthwork change in different areas of the project area, but also provides detailed data support for subsequent construction progress assessment and geological disaster risk analysis. The system automatically saves the distribution matrix and generates a report containing statistical information and visual charts for further analysis and decision-making by users. The report will list the statistical indicators of earthwork change, such as total fill volume, total excavation volume, average change, and spatial distribution characteristics of the change, helping users to fully understand the dynamic changes of earthwork in the project area.

[0105] The multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the present application can realize high-precision dynamic monitoring of water conservancy engineering deformation through the space-time coupled earthwork change calculation mechanism. Based on the gridded elevation model construction process, millimeter-level representation of surface deformation in the project area can be achieved, and through the difference calculation of the current and historical digital elevation models, the elevation change of each grid unit can be accurately quantified to realize the space-time continuity capture of micro earthwork movement. The earthwork change distribution matrix generated by mapping the grid position can construct a coupling expression of three-dimensional space and change intensity, and can realize the automatic identification and visual positioning of excavation and filling areas. By integrating the time and space dimension analysis framework, the present application can overcome the defects of low utilization rate of historical data and insufficient visualization of change trend in existing point cloud monitoring methods, and provide quantitative basis for construction progress deviation warning and geological disaster risk assessment, and finally achieve dynamic perception and early risk judgment of engineering safety situation.

[0106] In one embodiment, the multi-source point cloud data fusion water conservancy earthwork dynamic monitoring method provided by the present application comprises the following steps:

[0107] S61: Perform excavation area identification processing on the earthwork change distribution matrix, extract the negative value area and calculate the total excavation volume, generate excavation volume data, and the excavation volume data is used to indicate the earthwork reduction in the project area.

[0108] Specifically, the system traverses each element in the matrix, identifies all areas with negative elevation change values, which are the excavation areas. The system further spatially aggregates these negative value areas, merging adjacent negative grids into continuous excavation domains. In calculating the total excavation volume, the system integrates according to the area and elevation change of each grid cell. Specifically, the system multiplies the area of each grid cell by the elevation change to obtain the earthwork change volume of that cell, and then accumulates the earthwork change volumes of all negative cells to finally obtain the total excavation volume of the entire project area. The system integrates these information into excavation volume data, which not only includes the total excavation volume, but also records in detail the location, range and corresponding earthwork reduction of each excavation domain. These data provide accurate quantitative basis for subsequent earthwork balance analysis, and help construction managers intuitively grasp the earthwork excavation situation in the project area, ensuring the consistency of construction progress and design requirements.

[0109] S62: Perform fill area identification processing on the earthwork change volume distribution matrix, extract positive value areas and calculate the total fill volume, generate fill volume data, which is used to indicate the earthwork increase of the project area.

[0110] Specifically, the system again traverses the earthwork change volume distribution matrix to accurately extract the areas with positive elevation change values, i.e. fill areas. Similarly, the system spatially aggregates these positive value areas to form continuous fill domains. In calculating the total fill volume, the system uses a similar integration method as excavation calculation, multiplies the area of each positive grid cell by the elevation change to obtain the earthwork change volume of that cell, and accumulates the earthwork change volumes of all positive cells to obtain the total fill volume. The system generates fill volume data, which records in detail the location, range and corresponding earthwork increase of the fill area. By combining the excavation volume data and fill volume data, the system provides comprehensive quantitative basis for subsequent earthwork balance analysis and construction progress evaluation. This not only helps to optimize construction resource allocation, but also ensures the rationality and efficiency of construction activities, providing data support for project management.

[0111] S63: Based on the excavation volume data and fill volume data, perform trend analysis processing on the earthwork change volume of multiple monitoring periods, calculate the earthwork volume change rate, and generate the evaluation analysis results of the earthwork volume change trend.

[0112] Specifically, the system reads the earthwork change data of historical monitoring periods and integrates it with the current period's excavation and filling amount. Preferably, the system can use time series analysis algorithms such as linear regression or moving average to analyze the data of multiple periods, thereby calculating the earthwork change rate. This process involves detailed analysis of time series data to determine the trend line of earthwork changes, revealing the dynamic change law of earthwork over time. The system calculates the earthwork change rate, i.e., the increase or decrease of earthwork per unit time, providing a key indicator for dynamic assessment of construction progress. By analyzing the earthwork change trend, the system can identify abnormal fluctuations in construction progress in a timely manner, providing early warning of potential delay risks, and providing a scientific basis for optimizing construction plans and resource allocation.

[0113] S64: Process the evaluation analysis results, combine the preset construction plan and the geological safety threshold of the engineering area to conduct construction progress and geological disaster risk assessment, and generate a dynamic monitoring evaluation report.

[0114] Specifically, after generating the evaluation analysis results of the earthwork change trend, the system further combines the preset construction plan and the geological safety threshold of the engineering area to conduct construction progress and geological disaster risk assessment. The system automatically compares the actual earthwork change with the expected value in the construction plan, calculates the progress deviation, and assesses the construction progress status according to the deviation degree. For example, the system determines whether the project progress is lagging or ahead by calculating the ratio of the actual completed earthwork to the planned earthwork. At the same time, the system automatically compares the earthwork change rate with the geological safety threshold to identify high-risk areas that may trigger geological disasters. Based on these analysis results, the system generates a dynamic monitoring evaluation report, which details the current status of construction progress, deviation from the plan, and early warning information of geological disaster risks. The report may also include the positioning of potential landslide risk areas and risk level assessment, providing comprehensive and timely decision support for engineering management personnel. In this way, the system helps engineering management personnel accurately control project progress and safety risks, ensuring the smooth progress and long-term stable operation of water conservancy construction.

[0115] Preferably, as shown in Figure 4 The present application provides a multi-source point cloud data fusion water conservancy earthwork dynamic monitoring system 700, which is configured with the following modules:

[0116] A multi-source point cloud preprocessing module 710 is used to identify and standardize the input raw multi-source point cloud data set, generate a standard point cloud data set, and the raw multi-source point cloud data set includes airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud.

[0117] The terrain partition processing module 720 is configured to perform terrain partition processing based on the standard point cloud dataset, and generate a terrain classification distribution map, which is used to indicate the spatial distribution range of steep slope areas, gentle slope areas and flat areas.

[0118] The adaptive grid generation module 730 is configured to perform grid division processing on the terrain classification distribution map, adjust the grid density of the grid unit according to the terrain gradient difference of the terrain feature unit, and generate an adaptive grid framework with a terrain label.

[0119] The point cloud fusion calculation module 740 is configured to perform fusion processing on the standard point cloud dataset by using the adaptive grid framework, calculate a weight coefficient by combining the point cloud coverage of the grid unit and the normalized terrain gradient, and generate a fusion grid elevation dataset through fusion calculation.

[0120] The earthwork change calculation module 750 is configured to perform earthwork change calculation processing based on the fusion grid elevation dataset, construct an elevation model of the engineering area, and compare historical data of a historical monitoring period, to generate an earthwork change distribution matrix.

[0121] The dynamic monitoring and evaluation module 760 is configured to perform dynamic analysis processing on the earthwork change distribution matrix, identify excavation and filling areas and calculate the change trend of the earthwork volume, and generate a dynamic monitoring and evaluation report, which is used to indicate the construction progress deviation and the geological disaster risk.

[0122] In summary, the water conservancy earthwork dynamic monitoring system based on multi-source point cloud data fusion provided by the present application can systematically solve the three core problems of insufficient data fusion accuracy, poor terrain adaptability and fragmented dynamic evaluation in the existing point cloud monitoring technology through deep collaborative processing of multi-source point cloud data and a terrain adaptive analysis mechanism. At the data fusion level, the spatial consistency control of multi-source point cloud data can be achieved based on the device identification and coordinate standardization processing procedure, effectively eliminating the system error of the airborne LiDAR, ground scanning and oblique photography data; the adaptive grid framework based on the terrain gradient difference can realize fine partition management of steep slope areas, gentle slope areas and flat areas, and significantly improve the complex terrain representation capability through differential density grid division. At the core technology link, the dual weight fusion mechanism based on the point cloud coverage and the normalized terrain gradient can guarantee the accuracy and reliability of the fusion grid elevation dataset, to realize millimeter-level earthwork change detection; and the dynamic comparison and analysis of the historical and current elevation models can capture the micro deformation characteristics of the engineering area. Finally, through the coupling analysis of the excavation and filling area identification and the change trend, the system can simultaneously generate a construction progress deviation warning and a geological disaster risk assessment, to realize the transformation of engineering safety control from passive response to active intervention, thereby effectively improving the monitoring accuracy, enhancing the terrain adaptability and optimizing the dynamic decision-making, and providing full-cycle, multi-dimensional and quantifiable safety monitoring support for large-scale water conservancy projects.

[0123] Preferably, the multi-source point cloud preprocessing module 710 provided by the application is configured with the following units:

[0124] A point cloud identification processing unit is configured to perform identification processing on each set of point cloud data in the input original multi-source point cloud data set, construct a unique identifier according to the type of acquisition equipment and the acquisition time, and generate a traceable point cloud data set.

[0125] A coordinate system alignment unit is configured to perform coordinate system alignment processing on the traceable point cloud data set, unify the airborne LiDAR point cloud, the ground scanning point cloud and the oblique photography point cloud to the engineering coordinate system, and generate a spatially aligned point cloud data set.

[0126] A point cloud density balancing unit is configured to perform density adjustment processing on the spatially aligned point cloud data set, balance the density difference of point clouds of different sources, and generate a standard point cloud data set indicating point cloud information unified in space and balanced in density.

[0127] Preferably, the terrain partition processing module 720 provided by the application is configured with the following units:

[0128] A terrain gradient calculation unit is configured to perform terrain gradient calculation on the standard point cloud data set, calculate the slope value of each point based on the elevation of the spatial neighborhood points, and generate terrain gradient data.

[0129] A terrain region division unit is configured to perform region division processing on the slope value of each point in the terrain gradient data, divide the region with a slope value less than a preset first threshold value into a flat area, divide the region with a slope value between the preset first threshold value and a preset second threshold value into a gentle slope area, and divide the region with a slope value greater than the preset second threshold value into a steep slope area, to generate a terrain partition division result.

[0130] A terrain distribution map generation unit is configured to generate a distribution map based on the terrain partition division result, label the boundaries of the steep slope area, the gentle slope area and the flat area, and generate a terrain classification distribution map.

[0131] Preferably, the adaptive grid generation module 730 provided by the application is configured with the following units:

[0132] A steep slope area grid division unit is configured to perform dense grid division processing on the steep slope area in the terrain classification distribution map, divide the area using grid units smaller than a preset first size, and generate steep slope area grid units.

[0133] A gentle slope area grid division unit is configured to perform medium-density grid division processing on the gentle slope area in the terrain classification distribution map, divide the area using grid units greater than the preset first size and smaller than a preset second size, and generate gentle slope area grid units.

[0134] The flat area grid division unit is configured to perform sparse grid division processing on the flat area in the terrain classification distribution map, divide the flat area using a grid unit larger than a preset second size, and generate a flat area grid unit;

[0135] The grid integration labeling unit is configured to perform integration processing on the steep slope area grid unit, the gentle slope area grid unit, and the flat area grid unit, label the terrain type label of each grid unit, and generate an adaptive grid framework with a terrain label.

[0136] Preferably, the point cloud fusion calculation module 740 provided by the application is configured with the following units:

[0137] The fusion weight calculation unit is configured to perform weight calculation processing on each grid unit in the adaptive grid framework, and generate a fusion weight coefficient;

[0138] The weighted elevation fusion unit is configured to perform weighted fusion processing on the standard point cloud dataset, and calculate the fusion elevation value of each grid unit;

[0139] The elevation dataset generation unit is configured to perform dataset generation processing on the fusion elevation value of all grid units, and form a fusion grid elevation dataset.

[0140] Preferably, the earthwork change calculation module 750 provided by the application is configured with the following units:

[0141] The elevation model construction unit is configured to process the fusion grid elevation dataset to construct an elevation model using the elevation value and spatial coordinates of the grid unit, and generate a current digital elevation model of the engineering area;

[0142] The elevation change calculation unit is configured to perform change amount calculation processing on the current digital elevation model and the digital elevation model of the historical monitoring period, and generate the elevation change amount of each grid unit;

[0143] The distribution matrix generation unit is configured to perform distribution matrix generation processing on the engineering area based on the elevation change amount of all grid units, map the elevation change amount data according to the grid position, and form an earthwork change amount distribution matrix.

[0144] Preferably, the dynamic monitoring evaluation module 760 provided by the application is configured with the following units:

[0145] The excavation area identification unit is configured to perform excavation area identification processing on the earthwork change amount distribution matrix, extract the negative value area and calculate the total excavation amount, generate excavation amount data, and the excavation amount data is used to indicate the earthwork reduction amount of the engineering area;

[0146] The fill area identification unit is configured to perform fill area identification processing on the earthwork change amount distribution matrix, extract a positive value area, and calculate a total fill amount, to generate fill amount data, which is used to indicate an earthwork increase amount of the engineering area.

[0147] The earthwork trend analysis unit is configured to perform trend analysis processing on earthwork change amounts of multiple monitoring periods based on the cut amount data and the fill amount data, calculate an earthwork amount change rate, and generate an evaluation analysis result of an earthwork amount change trend.

[0148] The dynamic monitoring evaluation unit is configured to perform processing on the evaluation analysis result, perform construction progress and geological disaster risk evaluation in combination with a preset construction plan and a geological safety threshold of the engineering area, and generate a dynamic monitoring evaluation report.

[0149] In one embodiment, the present application further provides a computer device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned water conservancy earthwork dynamic monitoring method of multi-source point cloud data fusion when executing the computer program.

[0150] In one embodiment, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned water conservancy earthwork dynamic monitoring method of multi-source point cloud data fusion.

[0151] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0152] For the device embodiment, since it basically corresponds to the method embodiment, the relevant part is described in the part of the method embodiment. The device embodiment described above is only schematic, and the components described as separate components can or can not be physically separate, and the components displayed as a unit can or can not be a physical unit, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the present disclosure according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0153] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various changes or replacements within the technical range disclosed by the present application, and these should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A dynamic monitoring method for water conservancy earthwork based on multi-source point cloud data fusion, characterized in that: The following steps are involved: S1: Identify and standardize the input original multi-source point cloud dataset to generate a standard point cloud dataset, wherein the original multi-source point cloud dataset includes airborne LiDAR point cloud, ground scanning point cloud, and oblique photography point cloud; S2: performing terrain zoning processing based on the standard point cloud dataset to generate a terrain classification distribution map, wherein the terrain classification distribution map is used to indicate the spatial distribution range of steep slope areas, gentle slope areas, and flat areas; S3: performing grid division processing on the terrain classification distribution map, adjusting the grid density of the grid cells according to the terrain gradient differences of the terrain feature cells, and generating an adaptive grid framework with terrain labels; S4: using the adaptive grid framework to perform fusion processing on the standard point cloud dataset, combining the point cloud coverage of the grid unit and the normalized terrain gradient to calculate the weight coefficient, and generating a fused grid elevation dataset through fusion calculation; S5: performing earthwork change calculation based on the fused grid elevation dataset, constructing an elevation model of the project area and comparing it with historical data of historical monitoring periods to generate an earthwork change distribution matrix; S6: Dynamically analyzing and processing the earthwork change distribution matrix, identifying the excavation and filling areas and calculating the change trend of the earthwork volume, and generating a dynamic monitoring and evaluation report, which is used to indicate construction progress deviations and geological disaster risks.

2. The method according to claim 1, characterized in that Said S1 comprises: S11: perform identification processing on each set of point cloud data in the input original multi-source point cloud dataset, construct a unique identifier based on the acquisition device type and acquisition time, and generate a traceable point cloud dataset; S12: performing coordinate system alignment processing on the traceable point cloud dataset, unifying the airborne LiDAR point cloud, the ground scanning point cloud, and the oblique photography point cloud into an engineering coordinate system, and generating a spatially aligned point cloud dataset; S13: Density adjustment processing is performed on the spatially aligned point cloud dataset to balance density differences between point clouds from different sources, thereby generating a standard point cloud dataset. The standard point cloud dataset is used to indicate point cloud information with uniform spatial position and balanced density.

3. The method according to claim 1, characterized in that The S2 includes: S21: performing terrain gradient calculation on the standard point cloud dataset, calculating the slope value of each point based on the elevation of the spatial neighborhood points, and generating terrain gradient data; S22: performing region division processing on the slope value of each point in the terrain gradient data, dividing the area with a slope value less than a preset first threshold into a flat area, dividing the area with a slope value between the preset first threshold and a preset second threshold into a gentle slope area, and dividing the area with a slope value greater than the preset second threshold into a steep slope area, thereby generating a terrain zoning result; S23: Generating a distribution map of the standard point cloud dataset based on the terrain zoning result, marking the boundaries of steep slope areas, gentle slope areas and flat areas, and generating a terrain classification distribution map.

4. The method according to claim 1, wherein The S3 includes: S31: performing dense grid division processing on the steep slope area in the terrain classification distribution map, using grid units smaller than a preset first size for division, to generate steep slope area grid units; S32: performing medium-density grid division processing on the gentle slope area in the terrain classification distribution map, using grid cells larger than a preset first size and smaller than a preset second size for division, to generate grid cells for the gentle slope area; S33: performing sparse grid division processing on the flat area in the terrain classification distribution map, using grid units larger than a preset second size for division, to generate flat area grid units; S34: Integrate the steep slope area grid cells, the gentle slope area grid cells, and the flat area grid cells, mark the terrain type label of each grid cell, and generate an adaptive grid framework with terrain labels.

5. The method according to claim 1, wherein The S4 includes: S41: Perform weight calculation processing on each grid unit in the adaptive grid framework to generate a fusion weight coefficient using the following formula: Where ω is the fusion weight coefficient, α and β are preset weight factors, and α + β = 1, C is the effective point cloud coverage rate within the grid cell, and G is the normalized terrain gradient value of the grid cell; S42: Perform weighted fusion processing on the standard point cloud dataset to calculate the fused elevation value of each grid cell using the following formula: Among them, H fused is the fused elevation value, w i is the weight coefficient of the i-th point cloud in the standard point cloud dataset, H i is the elevation value of the i-th point cloud, and n is the number of point clouds in the grid; S43: Performing dataset generation processing on the fused elevation values ​​of all grid cells to form a fused grid elevation dataset.

6. The method according to claim 1, characterized in that The S5 includes: S51: Processing the fused grid elevation dataset, constructing an elevation model using the elevation values ​​and spatial coordinates of the grid cells, and generating a current digital elevation model of the project area; S52: Calculate the change of the current digital elevation model and the digital elevation model of the historical monitoring period using the following formula to generate the elevation change of each grid cell: ΔH=H current -H historical Where ΔH is the elevation change of the grid cell, H current is the elevation value of the current monitoring period, H historical is the elevation value during the historical monitoring period; S53: Based on the elevation changes of all grid cells, a distribution matrix is ​​generated for the project area, and elevation change data is mapped according to grid positions to form an earthwork change distribution matrix.

7. The method according to claim 1, characterized in that The S6 includes: S61: performing excavation area identification processing on the earthwork change distribution matrix, extracting negative value areas and calculating the total excavation volume to generate excavation volume data, wherein the excavation volume data is used to indicate the earthwork reduction amount in the project area; S62: performing fill area identification processing on the earthwork change distribution matrix, extracting positive value areas and calculating the total fill volume to generate fill volume data, wherein the fill volume data is used to indicate the earthwork increase in the project area; S63: performing trend analysis on the earthwork change amount of multiple monitoring periods based on the excavation amount data and the filling amount data, calculating the earthwork amount change rate, and generating an evaluation analysis result of the earthwork amount change trend; S64: Processing the evaluation and analysis results, conducting construction progress and geological disaster risk assessment in combination with a preset construction plan and geological safety thresholds of the project area, and generating a dynamic monitoring and assessment report.

8. A dynamic monitoring system for water conservancy and earthwork based on multi-source point cloud data fusion, characterized in that: The system comprises: A multi-source point cloud preprocessing module is used to identify and standardize the input original multi-source point cloud dataset to generate a standard point cloud dataset. The original multi-source point cloud dataset includes airborne LiDAR point cloud, ground scanning point cloud and oblique photography point cloud; A terrain partitioning processing module is used to perform terrain partitioning processing based on the standard point cloud data set to generate a terrain classification distribution map, wherein the terrain classification distribution map is used to indicate the spatial distribution range of steep slope areas, gentle slope areas and flat areas; An adaptive grid generation module is used to perform grid division processing on the terrain classification distribution map, adjust the grid density of the grid unit according to the terrain gradient difference of the terrain feature unit, and generate an adaptive grid framework with terrain labels; a point cloud fusion calculation module, configured to fuse the standard point cloud dataset using the adaptive grid framework, calculate weight coefficients based on the point cloud coverage of the grid cells and the normalized terrain gradient, and generate a fused grid elevation dataset through fusion calculation; an earthwork change calculation module for calculating earthwork changes based on the fused grid elevation dataset, constructing an elevation model of the project area and comparing it with historical data from historical monitoring periods to generate an earthwork change distribution matrix; The dynamic monitoring and evaluation module is used to dynamically analyze and process the earthwork change distribution matrix, identify the excavation and filling areas, calculate the change trend of the earthwork volume, and generate a dynamic monitoring and evaluation report. The dynamic monitoring and evaluation report is used to indicate construction progress deviations and geological disaster risks.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images