Earthwork calculation amount dynamic monitoring and calculation method and system based on point cloud

Abstract

The application relates to the technical field of earthwork monitoring, in particular to a point cloud-based dynamic earthwork measurement monitoring method and system, which comprises the following steps: acquiring on-site point cloud data of a target construction area at a current monitoring moment; performing terrain restoration on the on-site point cloud data, removing vegetation points and temporary facility points, and obtaining target on-site point cloud; acquiring a measurement range and a design elevation of the target construction area, generating a design triangulation network model, and generating a design point cloud to construct a benchmark voxel grid model. The on-site point cloud data acquired in real time is compared with design point cloud data, the changes of excavation and filling in the construction area can be dynamically and continuously monitored, voxelization processing and height difference analysis are performed, high-precision earthwork measurement is realized, and local slight changes can be captured.

Description

Technical Field

[0001] This invention relates to the field of earthwork monitoring technology, specifically to a method and system for dynamic monitoring and calculation of earthwork volume based on point cloud. Background Technology

[0002] Currently, traditional methods mostly rely on manual surveying and measurement, which are easily affected by human factors and have large measurement errors. Especially in large-scale construction projects, the accumulation of errors may affect the final earthwork volume calculation. Moreover, traditional methods are usually based on static measurement or periodic surveys, lacking real-time monitoring of changes in the construction area, failing to dynamically and continuously reflect the progress of earthwork operations, and may not be able to detect problems in a timely manner.

[0003] Furthermore, traditional methods often rely on manual calculations or simple mathematical models, which are inefficient, especially when facing large-scale construction. The time cost of manual calculation and analysis is high, and it is difficult to quickly generate dynamic earthwork volume reports. In complex terrain or irregular construction areas, traditional methods may not be able to accurately capture subtle changes in earthwork operations, especially in variable construction environments, making it difficult to achieve refined monitoring. Traditional methods also struggle to quantify construction progress in real time. Unlike point cloud-based data monitoring methods, they cannot generate a construction progress index by comparing earthwork changes with the design earthwork volume, nor can they dynamically adjust the monitoring frequency according to the construction progress, resulting in low resource utilization efficiency and an inability to achieve accurate construction progress management. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamic monitoring and calculation of earthwork volume based on point cloud, comprising:

[0005] Acquire the on-site point cloud data of the target construction area at the current monitoring time; perform terrain restoration on the on-site point cloud data, remove vegetation points and temporary facility points, and obtain the target on-site point cloud;

[0006] Obtain the measurement range and design elevation of the target construction area, generate a design triangulation model, and use the design triangulation model to generate a design point cloud to construct a reference voxel mesh model.

[0007] The side length of the voxel unit is adjusted according to the distribution density of the target site point cloud, and the target site point cloud is voxelized and encapsulated based on the adjusted side length to obtain the first voxel mesh model.

[0008] The first voxel mesh model is compared with the reference voxel mesh model, and the first height value of each voxel unit in the first voxel mesh model and the second height value of the corresponding voxel unit in the reference voxel mesh model are extracted.

[0009] Calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than a preset height change threshold, add the voxel unit to the differential voxel set and mark the change type of the voxel unit as excavation or filling according to the positive or negative value of the height difference.

[0010] The first total volume of voxel units labeled as excavation type and the second total volume of voxel units labeled as fill type in the differential voxel set are statistically analyzed; the difference between the first total volume and the second total volume is calculated to obtain the net earthwork change, and the dynamic quantity calculation result corresponding to the target construction area is generated based on the net earthwork change.

[0011] Preferably, the on-site point cloud data includes multiple three-dimensional coordinate points on the surface of the target construction area;

[0012] The side length of the voxel unit is adjusted according to the distribution density of the target site point cloud, and the target site point cloud is voxelized and encapsulated based on the adjusted side length to obtain a first voxel mesh model, including:

[0013] Determine the reflection intensity value corresponding to each three-dimensional coordinate point in the target site point cloud;

[0014] When the reflection intensity value is greater than a preset intensity threshold, the three-dimensional coordinate point is determined as an effective surface point;

[0015] Calculate the point cloud density of the effective facets, and adjust the side length of the voxel unit in reverse according to the point cloud density. The larger the point cloud density, the smaller the side length of the voxel unit.

[0016] A three-dimensional mesh is constructed based on the adjusted side length dimensions, and each effective face point is assigned to the corresponding voxel unit to obtain the first voxel mesh model.

[0017] Preferably, the first voxel mesh model is compared with the reference voxel mesh model, and the first height value of each voxel cell in the first voxel mesh model and the second height value of the corresponding voxel cell in the reference voxel mesh model are extracted, including:

[0018] Traverse each voxel cell in the first voxel mesh model and search for the corresponding voxel cell with the same spatial index in the reference voxel mesh model;

[0019] Read the average elevation value of the current voxel element in the first voxel mesh model as the first height value;

[0020] The average elevation value of the corresponding voxel element in the reference voxel mesh model is read as the second height value;

[0021] Establish a one-to-one correspondence table between the first height value and the second height value.

[0022] Preferably, the statistical analysis of the first total volume of voxel units labeled as cut-out type and the second total volume of voxel units labeled as fill-in type in the differential voxel set includes:

[0023] Traverse each voxel unit in the set of differential voxels and identify its variation type marker;

[0024] The volume attribute values ​​of all voxel units marked as cut-out type are summed to obtain the first total volume;

[0025] The volume attribute values ​​of all voxel units marked as fill type are summed to obtain the second total volume;

[0026] Record the number of voxel units and their spatial distribution coordinates in the excavation and filling areas respectively.

[0027] Preferably, after generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes:

[0028] The dynamic quantity calculation results are compared with the preset design earthwork volume to calculate the percentage deviation.

[0029] A construction progress index is generated based on the deviation percentage, and the progress index is positively correlated with the deviation percentage.

[0030] The scanning frequency of the point cloud acquisition device is adjusted according to the construction progress index; the higher the progress index, the higher the scanning frequency.

[0031] Preferably, after generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes:

[0032] Obtain a heat map showing the distribution of the differential voxel set in the plane projection of the target construction area;

[0033] Based on the heat map, key areas for earthwork operations are determined, and these key areas are those with a color depth greater than a preset threshold.

[0034] Generate encrypted monitoring instructions for the key areas, the encrypted monitoring instructions including encrypted coordinates and encrypted time;

[0035] According to the encrypted monitoring command, the measuring equipment is controlled to perform secondary point cloud acquisition on the key area to obtain second point cloud data;

[0036] The voxel cell states in the corresponding region of the first voxel mesh model are replaced using the second point cloud data to update the local model.

[0037] Preferably, after calculating the first total volume of voxel units labeled as cut-out type and the second total volume of voxel units labeled as fill-in type in the differential voxel set, the method further includes:

[0038] Identify anomalous voxel units located at the edge of the differential voxel set, where the edge location is the position in the voxel's neighborhood where a null value exists;

[0039] Obtain the height value of the neighboring voxel units surrounding the abnormal voxel unit;

[0040] Calculate the median height value of the neighboring voxel units, and replace the height value of the abnormal voxel units with the median to obtain the corrected set of differential voxels.

[0041] The first and second total volumes are recalculated based on the corrected set of differential voxels, and the dynamic calculation results are updated.

[0042] Preferably, before acquiring the on-site point cloud data and design point cloud data of the target construction area at the current monitoring time, the method further includes:

[0043] Establish an initial point cloud model of the target construction area;

[0044] Terrain restoration is performed on the initial point cloud model to remove vegetation points and temporary facility points, resulting in a denoised initial point cloud model.

[0045] The denoised initial point cloud model is converted into a triangular mesh model, and the triangular mesh model is used as the basis for constructing the first voxel mesh model;

[0046] The design point cloud data is converted into a design triangular mesh model, and the design triangular mesh model is aligned with the initial triangular mesh model using the transformation matrix.

[0047] A point cloud-based dynamic monitoring and calculation system for earthwork volume, applicable to the aforementioned point cloud-based dynamic monitoring and calculation method for earthwork volume, includes:

[0048] The point cloud acquisition module is configured to acquire on-site point cloud data of the target construction area at the current monitoring time; perform terrain restoration on the on-site point cloud data, remove vegetation points and temporary facility points to obtain the target on-site point cloud, acquire the measurement range and design elevation of the target construction area, generate a design triangular mesh model, and use the design triangular mesh model to generate a design point cloud to construct a reference voxel mesh model.

[0049] The terrain restoration module is configured to generate a triangular network model of contour lines from the contour line point cloud and the target site point cloud; analyze the spatial distribution characteristics of the target site point cloud to distinguish between vegetation point cloud and ground points; and flatten vegetation points at the same location that are higher than the contour lines to align with the contour lines, thereby removing vegetation.

[0050] The voxel encapsulation module is configured to adjust the side length of the voxel unit according to the distribution density of the target site point cloud, and to encapsulate the target site point cloud into a voxel based on the adjusted side length to obtain a first voxel mesh model.

[0051] The voxel comparison module is configured to compare the first voxel mesh model with the reference voxel mesh model, and extract the first height value of each voxel unit in the first voxel mesh model and the second height value of the corresponding voxel unit in the reference voxel mesh model.

[0052] The difference marking module is configured to calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than a preset height change threshold, the voxel unit is added to the differential voxel set, and the change type of the voxel unit is marked as excavation or filling according to the positive or negative value of the height difference.

[0053] The dynamic quantity calculation module is configured to calculate the first total volume of voxel units marked as excavation type and the second total volume of voxel units marked as fill type in the differential voxel set; perform a difference calculation between the first total volume and the second total volume to obtain the net earthwork change; and generate the dynamic quantity calculation result corresponding to the target construction area based on the net earthwork change.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] (1) By comparing the real-time acquired field point cloud data with the design point cloud data, the present invention can dynamically and continuously monitor the changes in excavation and filling in the construction area. Through voxelization processing and height difference analysis, high-precision earthwork volume measurement can be achieved, which can capture local small changes. Through the construction of voxel grid model and voxel comparison, the volume of excavation and filling voxel units can be automatically counted, reducing manual measurement errors and workload. The calculation of net earthwork change can directly generate dynamic quantity calculation results, improving construction management efficiency.

[0056] (2) This invention achieves a balance between model accuracy and data volume by adjusting the side length of voxel units according to the density of point cloud on site. The accuracy is higher in dense areas and the computational redundancy is reduced in sparse areas. Abnormal voxels are corrected to improve the reliability of voxel models and the accuracy of calculation results. By comparing the net earthwork change with the design earthwork volume, a construction progress index can be generated to provide quantifiable indicators for project management. The scanning frequency can be dynamically adjusted according to the progress index to achieve denser monitoring of key areas, saving resources while ensuring the accuracy of construction monitoring. Furthermore, a planar heat map is generated using differential voxel sets to automatically identify key areas of earthwork operations. Encrypted monitoring instructions are generated for key areas, supporting secondary point cloud acquisition and local model updates, thereby improving the targeting and accuracy of monitoring. Attached Figure Description

[0057] Figure 1 This is a schematic flowchart of the overall method in one embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of the overall system architecture in one embodiment of the present invention.

[0059] In the diagram: 1. Point cloud acquisition module; 2. Terrain reconstruction module; 3. Voxel encapsulation module; 4. Voxel comparison module; 5. Difference marking module; 6. Dynamic quantity calculation module. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1, please refer to Figure 1 This invention provides a technical solution: a method for dynamic monitoring and calculation of earthwork volume based on point cloud, comprising:

[0062] S1. Obtain the on-site point cloud data of the target construction area at the current monitoring time; perform terrain restoration on the on-site point cloud data, remove vegetation points and temporary facility points, and obtain the target on-site point cloud;

[0063] S2. Obtain the measurement range and design elevation of the target construction area, generate a design triangulation model, and use the design triangulation model to generate a design point cloud to construct a reference voxel mesh model.

[0064] S3. Adjust the side length of the voxel unit according to the distribution density of the target site point cloud, and encapsulate the target site point cloud into voxels based on the adjusted side length to obtain the first voxel mesh model.

[0065] S4. Compare the first voxel mesh model with the reference voxel mesh model, and extract the first height value of each voxel cell in the first voxel mesh model and the second height value of the corresponding voxel cell in the reference voxel mesh model.

[0066] S5. Calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than the preset height change threshold, add the voxel unit to the differential voxel set and mark the change type of the voxel unit as either excavation or filling according to the positive or negative value of the height difference.

[0067] S6. Statistically calculate the first total volume of voxel units labeled as excavation type and the second total volume of voxel units labeled as fill type in the voxel set; perform a difference calculation between the first total volume and the second total volume to obtain the net earthwork change, and generate the dynamic quantity calculation result corresponding to the target construction area based on the net earthwork change.

[0068] It should be noted that the collection involves both on-site point cloud data and design point cloud data of the target construction area. On-site point cloud data is real-time data collected using tools such as laser scanning equipment, reflecting the actual shape and height of the land. Design point cloud data, on the other hand, is virtual data generated based on construction design drawings, representing the planned shape and height of the construction site. For example, if a construction site requires earthwork operations, the design point cloud data reflects the final leveled land height, while the on-site point cloud data reflects the current height of the land after excavation or backfilling.

[0069] Terrain reconstruction is performed on the on-site point cloud data to remove unwanted points, such as vegetation and temporary facilities, to ensure data accuracy. For example, at a construction site, laser scanning may record the tops of trees or temporary construction sheds as point cloud data. Through terrain reconstruction, the system removes these points that are irrelevant to earthwork calculations, ensuring that only ground data related to construction is considered.

[0070] Adjust the on-site point cloud data to a coordinate system consistent with the design data; adjust the on-site point cloud data to the standard coordinate system of the design drawings for comparative analysis.

[0071] Based on the design point cloud data, the system creates a baseline voxel grid model. This model divides the ground into small grid cells to more accurately describe the land's shape. For example, imagine a large rectangular plot of land. The design point cloud data represents the soil height of this plot. The baseline voxel grid model divides this plot into multiple small cells, each representing a region of the land. The height of these small cells is the soil height in the original design drawing.

[0072] The size of the voxel unit is adjusted according to the distribution density of the on-site point cloud data, and the on-site point cloud data is converted into a voxel mesh model. In this process, the size of the voxel is adjusted according to the distribution density of the data points in order to more accurately reflect the actual height of the land. For example, in some places, the point cloud data of the construction site may be relatively dense, and the system will make the voxel units in these areas smaller in order to capture more detailed ground changes; while in other areas, the voxel units may be larger.

[0073] The system compares the first voxel mesh model on-site with the reference voxel mesh model to analyze the differences between them. The height of each voxel element is extracted, and its difference from the design height is calculated. A set threshold is used to determine whether these changes are significant. For example, if the on-site soil layer height in a certain area is lower than the design soil layer height, it means that the soil in that area may have been excavated. If the on-site height is higher than the design height, it means that the soil may have been filled. By calculating these differences, the system can identify which areas are excavated and which areas are filled.

[0074] The total volume of excavation and fill voxel units in the statistical difference voxel set is used to calculate the net change in earthwork. This value represents the actual change in earthwork during earthwork operations, that is, the difference between excavation and fill during construction. For example, if the excavation volume of a region is 1000 cubic meters and the fill volume is 800 cubic meters, then the net change in earthwork is 200 cubic meters, which represents the total change in earthwork in that region.

[0075] Based on the calculated net earthwork change, dynamic earthwork volume calculation results are generated for the construction area. These results will provide construction managers with real-time reference for the progress and accuracy of earthwork operations. For example, the construction team can use these dynamic volume calculation results to understand the progress of earthwork operations, such as the amount of earthwork to be excavated and the amount of earthwork to be backfilled, to help them make adjustments and decisions.

[0076] It should be noted that, for the target site point cloud and contour point cloud, a triangular network model of contour lines is generated using the contour point cloud; the spatial distribution characteristics of the target site point cloud are analyzed to distinguish between vegetation point cloud and ground points; vegetation points at the same location that are higher than the contour lines are flattened to align with the contour lines, thereby removing vegetation.

[0077] Process point cloud data and contour data of the target site to generate an accurate 3D terrain model; contour point clouds are usually obtained through topographic surveys and represent changes in ground elevation.

[0078] A triangulated irregular network (TIN) model is generated using contour point clouds. This is a mathematical model representing the Earth's surface. It accurately reflects the undulations of the terrain by connecting points in the point cloud to form triangles. For example, suppose there is a set of contour data representing lines at different altitudes in a mountainous area. This system can transform these lines into a terrain model composed of triangles, which can clearly show terrain features such as ridges and valleys.

[0079] Analyze the point cloud data of the target site to distinguish between vegetation point clouds and ground points; vegetation point clouds usually contain points above the ground such as trees and shrubs, while ground points refer to bare ground or other low-lying objects; for example, in a forest area, there are many high points (representing tree canopies) in the point cloud data. The algorithm identifies these points as vegetation, while the lower points are identified as the ground.

[0080] Vegetation points located above the contour lines are flattened to align with them. This means that for vegetation above a certain elevation, its height will be adjusted to match the terrain's contour lines, effectively removing the vegetation. For example, if a contour line in an area is marked at 500 meters, and some trees in that area are 550 meters tall, the coordinates of these trees will be adjusted to 500 meters, thus "removing" the trees' influence and retaining only the true height of the terrain.

[0081] In one alternative embodiment, the on-site point cloud data includes multiple three-dimensional coordinate points on the surface of the target construction area;

[0082] The side length of the voxel unit is adjusted according to the distribution density of the target site point cloud, and the target site point cloud is voxelized and encapsulated based on the adjusted side length to obtain the first voxel mesh model, including:

[0083] Determine the reflection intensity value corresponding to each three-dimensional coordinate point in the target site point cloud;

[0084] When the reflection intensity value is greater than the preset intensity threshold, the three-dimensional coordinate point is determined as an effective surface point;

[0085] Calculate the point cloud density of effective facets, and adjust the side length of the voxel unit in reverse according to the point cloud density. The higher the point cloud density, the smaller the side length of the voxel unit.

[0086] A three-dimensional mesh is constructed based on the adjusted side length dimensions, and each effective face point is assigned to the corresponding voxel unit to obtain the first voxel mesh model.

[0087] It should be noted that on-site point cloud data refers to multiple three-dimensional coordinate points on the surface of the target construction area. Each point records its position in space (X, Y, Z coordinates), and some points also carry reflection intensity information. For example, when a laser scanner scans a construction site, the scanner will generate a point at each accessible location on the surface. For example, on a slope, there may be tens of thousands of points, each of which records its height and reflection signal intensity.

[0088] Point cloud data may contain unwanted points, such as buildings, construction machinery, vegetation, or temporary facilities. Therefore, each point has a reflectance value to determine if it is a valid ground point. When the reflectance value is greater than a certain set threshold, the point is considered a valid surface point. For example, if a scanner scans a mound of earth and a tree, the mound's reflectance is higher than the threshold and is considered a valid surface point; leaves and branches have lower reflectance and are ignored. This ensures that subsequent analysis is based solely on ground morphology and is not affected by interfering objects.

[0089] Point cloud density varies across different regions; some areas have more scanned points, while others have fewer. To improve mesh accuracy, the size of the voxel unit needs to be adjusted based on the point cloud density: high point cloud density → use smaller voxel units to capture terrain changes more precisely; low point cloud density → use larger voxel units to avoid generating overly sparse or hollow meshes. For example, on a flat construction site with very dense scanned points, the voxel units can be set to be smaller (e.g., a few centimeters on each side), so that each small mesh can reflect subtle changes in ground elevation. In a relatively open area with sparse points, larger voxel units (e.g., tens of centimeters on each side) can be used to reduce the amount of data computation while still roughly describing the terrain.

[0090] Based on the voxel side lengths determined in the previous step, the entire construction area is divided into a three-dimensional mesh, with each small cube being a voxel unit. Then, each effective face point is assigned to the corresponding voxel unit, thus generating the first voxel mesh model. For example, imagine the construction site as a three-dimensional chessboard divided into many small cubes, with each effective face point falling into a certain cube. Points of different heights are assigned to corresponding voxel units, so each voxel unit has a value representing its height. This meshing process makes earthwork calculation, volume measurement, and subsequent excavation and filling analysis operable and automated.

[0091] In an optional embodiment, a voxel comparison is performed between the first voxel mesh model and the reference voxel mesh model, and the first height value of each voxel cell in the first voxel mesh model and the second height value of the corresponding voxel cell in the reference voxel mesh model are extracted, including:

[0092] Traverse each voxel cell in the first voxel mesh model and find the corresponding voxel cell with the same spatial index in the reference voxel mesh model;

[0093] Read the average elevation value of the current voxel element in the first voxel mesh model as the first height value;

[0094] Read the average elevation value of the corresponding voxel element in the reference voxel mesh model as the second height value;

[0095] Establish a one-to-one correspondence table between the first altitude value and the second altitude value.

[0096] It should be noted that the coordinate system of the first voxel mesh model is unified to the coordinate system of the reference voxel mesh model. This step is to ensure that the two voxel meshes are compared in the same coordinate system. For example, suppose there are two voxel mesh models, the first voxel mesh model (field data) and the reference voxel mesh model (design data), and their origins may be different. For example, the origin of the first voxel mesh model is (0,0,0), while the origin of the reference voxel mesh model is (100,100,0). Therefore, all voxels in the first voxel mesh model need to be translated and rotated using transformation matrices to make them consistent with the coordinate system of the reference voxel mesh model.

[0097] Traverse each voxel cell in the first voxel mesh model and find the corresponding voxel cell with the same spatial position in the reference voxel mesh model. For example, suppose a voxel cell in the first voxel mesh model is located at (10,5,3), and you want to find a voxel cell in the reference voxel mesh model that is also located at (10,5,3) (Note: the coordinates here are already transformed coordinates).

[0098] For each corresponding voxel cell, the average height value of that voxel cell (i.e., the average elevation within that voxel) needs to be read from the first voxel mesh model. For example, assuming that the height value of the voxel cell (10,5,3) in the first voxel mesh model is 3.5 meters, then this 3.5 meters is the first height value.

[0099] Read the average height value of the corresponding voxel cell from the reference voxel mesh model; for example, assuming that the height value of the voxel cell (10,5,3) in the reference voxel mesh model is 3.0 meters, then this 3.0 meters is the second height value;

[0100] Establish a table mapping the first and second height values; this table records the comparison of the average height value of each voxel unit in the two models; for example, after traversing multiple voxel units, the following mapping is obtained: the first height value of the first voxel mesh model (10,5,3) is 3.5 meters, and the second height value of the corresponding voxel unit in the reference voxel mesh model is 3.0 meters; the first height value of the first voxel mesh model (12,6,4) is 4.2 meters, and the second height value of the corresponding voxel unit in the reference voxel mesh model is 4.0 meters.

[0101] In an optional embodiment, the first total volume of voxel cells labeled as cut-out type and the second total volume of voxel cells labeled as fill-in type in the statistically differential voxel set include:

[0102] Traverse each voxel unit in the set of differential voxels and identify its variation type label;

[0103] The volume attribute values ​​of all voxel units marked as cut-out type are summed to obtain the first total volume;

[0104] The volume attribute values ​​of all voxel units marked as fill type are summed to obtain the second total volume;

[0105] Record the number of voxel units and their spatial distribution coordinates in the excavation and filling areas respectively.

[0106] It should be noted that the differential voxel set refers to the set of voxel elements whose heights have changed after comparing the first voxel mesh model and the reference voxel mesh model. Each voxel element in the set usually has a change type label indicating whether it is "cut" or "fill". For example, if a voxel in the construction site scan has a height of 4 meters in the design model but an actual height of 3 meters, then this voxel is a fill; if the actual height is 5 meters, then it is a cut.

[0107] Examine each voxel in the differential voxel set and read its change type; for example, the differential voxel set may contain these voxel units: Voxel A: Change type = cut, volume = 1 cubic meter; Voxel B: Change type = fill, volume = 0.8 cubic meters; Voxel C: Change type = cut, volume = 1.2 cubic meters;

[0108] Total excavation volume: The total volume of the excavation area is obtained by summing the volumes of all voxel units marked as excavation. Total fill volume: The total volume of the fill area is obtained by summing the volumes of all voxel units marked as fill. For example: Excavation voxels: A and C, total volume = 1 + 1.2 = 2.2 cubic meters; Fill voxel: B, total volume = 0.8 cubic meters.

[0109] In addition to the total volume, the following also need to be recorded: Voxel count: the number of voxel units in the excavation and filling; Spatial distribution coordinates: the position of each voxel unit in space, so as to know the location and range of the excavation and filling; For example: the number of voxels in the excavation = 2, and the coordinates are (10,5,3) and (12,6,4) respectively; the number of voxels in the filling = 1, and the coordinates are (11,5,4); By recording this information, a spatial distribution map of the excavation and filling can be generated to understand where there is more excavation and where there is more filling.

[0110] In an optional embodiment, after generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes:

[0111] The dynamic quantity calculation results are compared with the preset design earthwork volume, and the deviation percentage is calculated.

[0112] A construction progress index is generated based on the percentage of deviation, and the progress index is positively correlated with the percentage of deviation.

[0113] The scanning frequency of the point cloud acquisition equipment is adjusted according to the construction progress index; the higher the progress index, the higher the scanning frequency.

[0114] It should be noted that dynamic quantity calculation results refer to the actual earthwork volume obtained by monitoring earthwork changes at the construction site in real time; while the design earthwork volume is the pre-set earthwork plan volume before construction. Comparing these two values ​​can reveal the gap between the construction site and the design target. For example, suppose the design earthwork volume is 1000 cubic meters, meaning the planned excavation or filling volume is 1000 cubic meters; while the dynamic quantity calculation result obtained through on-site scanning is 950 cubic meters. By comparing the actual earthwork volume (950 cubic meters) with the design earthwork volume (1000 cubic meters), the difference between them can be calculated, i.e., the deviation percentage.

[0115] The percentage deviation measures the difference between the construction progress and the planned progress based on the discrepancy between the dynamic calculation result and the designed earthwork volume. A larger percentage deviation indicates a more significant delay or advancement in construction progress. For example: Designed earthwork volume: 1000 cubic meters; Dynamic calculation result: 950 cubic meters; Percentage deviation = (1000 - 950) / 1000 × 100% = 5%; This 5% deviation means that the actual construction progress is 5% slower than the design plan.

[0116] The construction progress index is an indicator generated based on the percentage deviation, reflecting the speed of construction progress. The progress index is positively correlated with the percentage deviation; the larger the percentage deviation, the higher the progress index, indicating slower progress, which may require more resources and attention. For example, assuming the construction progress index ranges from 0 to 10, where 0 represents progress completely on schedule and 10 represents the slowest progress, if the percentage deviation is 5%, according to preset rules, the construction progress index can be set to 3, indicating that there is some delay in progress, but it is still within a controllable range.

[0117] Adjust the scanning frequency of the point cloud acquisition equipment according to the construction progress index. The higher the progress index, the slower the construction progress, which may require more frequent scanning to better monitor the progress. Conversely, if the progress is good, the scanning frequency can be appropriately reduced to avoid wasting resources. For example, if the construction progress index is 8, it means the progress is slow, so the scanning frequency can be increased, for example, twice a week. If the construction progress index is 2, it means the progress is good, so the scanning frequency can be reduced, for example, once a month.

[0118] In an optional embodiment, after generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes:

[0119] Obtain a heat map showing the distribution of the differential voxel set in the plane projection of the target construction area;

[0120] The key areas for earthwork operations are determined based on the distribution heat map. The key areas are those with a color depth greater than a preset threshold.

[0121] Generate encrypted monitoring instructions for key areas, including encrypted coordinates and encrypted time.

[0122] According to the encrypted monitoring instructions, the measuring equipment is controlled to perform secondary point cloud acquisition on key areas to obtain second point cloud data;

[0123] The voxel cell states in the corresponding region of the first voxel mesh model are replaced using the second point cloud data, and the local model is updated.

[0124] It should be noted that during construction, the set of differential voxels reflects the changes in earthwork. By analyzing these changes, a heatmap can be generated. A heatmap is a graphical method that uses color depth to display data distribution, representing the amount of earthwork change in different areas. Here, the heatmap shows the distribution of earthwork changes within the target construction area. For example, suppose a construction project has uneven terrain. By comparing the differences between the design model and the actual situation, the generated set of differential voxels can be displayed on the plane of the construction area using a heatmap. The depth of the color indicates the degree of earthwork change in different areas; for example, darker areas may indicate larger amounts of excavation or filling, while lighter areas may indicate less earthwork change.

[0125] Based on the generated heatmap, the color depth can be used to determine which areas require special attention. Generally, areas with a color depth greater than a preset threshold indicate significant changes in earthwork, and these areas are key areas that need to be monitored and inspected first. For example, suppose the heatmap of several areas in the construction zone shows a very dark color (indicating significant changes in earthwork in these areas). These areas may be critical parts of the construction, such as foundation pits or road intersections; therefore, these areas will be identified as key areas.

[0126] For identified key areas, encrypted monitoring instructions need to be issued for more precise monitoring. These instructions consist of two main parts: encrypted coordinates, indicating the location of the key area; and encrypted time, indicating the time points at which monitoring is conducted in these areas. For example, if significant changes in earthwork are detected around a foundation pit at a certain stage, this area is identified as a key area. The system will then generate an encrypted monitoring instruction, directing the measuring equipment to scan more frequently at specific coordinates of the foundation pit, such as once per hour.

[0127] According to the encrypted monitoring instructions, the control and measurement equipment performs secondary point cloud acquisition in key areas; the second scan will provide more detailed data, which will help to update the earthwork operation model more accurately; this secondary acquisition data is called second point cloud data; for example: suppose in the foundation pit area, the point cloud data obtained by the first scan only roughly describes the earthwork changes, but does not have enough precision; according to the encrypted monitoring instructions, the measurement equipment performs a second scan of the area to obtain more detailed point cloud data;

[0128] The point cloud data obtained from the second scan is used to update the previous model. Specifically, the voxel unit states of the corresponding area in the first voxel mesh model are replaced by the second point cloud data, thereby updating the local earthwork model and making it more accurately reflect the actual construction progress. For example, in the foundation pit area, the voxel mesh model obtained from the first scan may not be accurate enough because it does not have sufficiently detailed data. The second point cloud data obtained from the second scan can replace the voxel unit states (such as earthwork volume, terrain changes, etc.) in that area of ​​the model, updating the local model and making it closer to the actual construction situation.

[0129] In an optional embodiment, after determining the first total volume of voxel cells labeled as cut-out type and the second total volume of voxel cells labeled as fill-in type in the statistically differential voxel set, the method further includes:

[0130] Identify anomalous voxel units located at the edges of the differential voxel set, where the edge position is the location in the voxel's neighborhood where a null value exists;

[0131] Obtain the height value of the neighboring voxel cells surrounding the abnormal voxel cell;

[0132] Calculate the median height value of the neighboring voxel units and replace the height value of the abnormal voxel units with the median to obtain the corrected set of differential voxels;

[0133] The first and second total volumes are recalculated based on the corrected set of differential voxels, and the dynamic calculation results are updated.

[0134] It should be noted that in the differential voxel set, anomalous voxel cells at the edge are those located at the edge of the voxel network and have null values ​​(i.e., missing data or incomplete areas) in their surrounding neighborhood. These anomalous voxel cells may lead to inaccurate calculations because their values ​​are incomplete or cannot correctly reflect actual earthwork changes. For example, suppose that in the differential voxel set at a construction site, earthwork data for some areas is obtained from sensors or measuring equipment, and some voxels in these areas are at the edge, with insufficient data in their surrounding neighborhood (e.g., a sensor may not have covered them); these voxel cells are labeled as "anomalous voxel cells".

[0135] For these anomalous voxel units, it is necessary to examine their surrounding neighboring voxel units; neighboring voxel units refer to other voxel units adjacent to the anomalous voxel unit; by examining the height values ​​of these neighboring voxel units, it is helpful to determine the correction value of the anomalous voxel unit; for example: Suppose that an anomalous voxel unit is located at the edge of the pit, and there are multiple neighboring voxel units whose height values ​​have been measured; for example, the heights of the surrounding voxel units are 5 meters, 5.1 meters, and 5.2 meters, respectively;

[0136] By calculating the median height values ​​of these neighboring voxel units, a standard value representing the surrounding environment is obtained. The median is the value in the middle after sorting the height values. Here, the median is chosen instead of the average to avoid the influence of some extreme data. For example, based on the previous example, the height values ​​of the neighboring voxel units are 5 meters, 5.1 meters, and 5.2 meters. After sorting, the median of these three values ​​is 5.1 meters. Therefore, the height value of the abnormal voxel unit will be corrected to 5.1 meters.

[0137] Replace the original height value of the anomalous voxel with the calculated median value; in this way, the anomalous voxel is corrected and becomes more consistent with the surrounding environment; for example: if the height value of the anomalous voxel could not be calculated or was incorrect (e.g., its value was empty or the data was inaccurate), now replace it with 5.1 meters to ensure that it is consistent with the height value of the surrounding neighborhood.

[0138] By correcting the height values ​​of the anomalous voxel elements, the content of the entire differential voxel set has changed. Now, the total volume of excavation and fill can be recalculated based on the corrected differential voxel set. This results in more accurate calculations that reflect actual earthwork changes. For example, suppose the changes in voxel elements in the excavation area are adjusted in the corrected differential voxel set, which may cause a change in the first total volume (total excavation). The voxel elements in the fill area may also be adjusted, affecting the second total volume (total fill). After correction, the dynamic volume calculation results will be more accurate.

[0139] In an optional embodiment, before acquiring the on-site point cloud data and design point cloud data of the target construction area at the current monitoring time, the method further includes:

[0140] Establish an initial point cloud model of the target construction area;

[0141] Terrain reconstruction is performed on the initial point cloud model, and vegetation points and temporary facility points are removed to obtain a denoised initial point cloud model.

[0142] The denoised initial point cloud model is converted into a triangular mesh model, and the triangular mesh model is used as the basis for constructing the reference voxel mesh model.

[0143] The design point cloud data is converted into a design triangular mesh model, and the design triangular mesh model is aligned with the initial triangular mesh model using a transformation matrix.

[0144] It should be noted that the acquisition of three-dimensional spatial information of the construction area; the initial point cloud model is a set of three-dimensional points of the construction site obtained through scanning or measurement. It records the spatial location of the ground, existing buildings, earthwork accumulation, etc. For example, suppose a new road is to be built, and point cloud data of the construction area is obtained by using drone laser scanning; the initial point cloud model is to organize these scanned points according to spatial coordinates to form a complete three-dimensional model, which shows the ground undulations, embankments and the outline of surrounding buildings.

[0145] Raw point cloud data usually contains some noisy points, such as vegetation (trees, grass) or temporary facilities (construction sheds, machinery and equipment). If these points are not processed, they will affect subsequent cut and fill analysis. Terrain restoration is to remove these irrelevant points and retain only the points of the real ground or construction foundation. For example, there are some trees and temporary construction warehouses in the scanned road area. After terrain restoration, these points will be deleted, leaving only the point cloud of the ground, roadbed and soil mound.

[0146] Point clouds are just discrete points, making direct volume calculations difficult. To facilitate analysis, point clouds need to be converted into triangular mesh models. Triangular mesh models connect adjacent points with triangles to form continuous surfaces, which can more accurately represent terrain undulations. For example, after denoising a road point cloud and converting it into a triangular mesh model, you can see that the entire construction area is covered by small triangles, and the changes in the height of the triangles reflect the undulations of the road terrain. This allows you to visually see the distribution of potholes and roadbed fill.

[0147] Using the triangular mesh model as a foundation, a voxel mesh model can be constructed. The voxel model divides space into small cubic units, with each voxel recording the height value or whether it is earthwork at its location. The voxel model is very suitable for calculating cut and fill volumes and performing differential analysis because each cube can be quantified. For example, a road construction area is divided into small cubes with sides of 1 meter, and each voxel records the height at that location. In this way, a construction area model represented by a "cube mesh" is obtained, which is convenient for calculation and analysis.

[0148] Construction projects all have design drawings, and the design point cloud is an ideal terrain model generated based on the planning and design. The design point cloud is also converted into a triangular mesh model, which can be compared with the initial point cloud model. For example, the road design specifies that the roadbed height is 2 meters and the road surface is flat. After the design point cloud is converted into a triangular mesh model, each triangle reflects the design height and slope.

[0149] Because scanning and design may use different coordinate systems, alignment is required through a transformation matrix to ensure that the design model and the initial scan model are completely overlapped in space. This is necessary to accurately compare the differences between the actual construction and the design. For example, the initial point cloud model and the design model may have different scanning coordinates, possibly offset by a few meters. After alignment, the road terrain of the scan model will be found to be completely superimposed on the design model. At this point, it is possible to accurately calculate which areas need to be excavated and which areas need to be filled.

[0150] Example 2, please refer to Figure 2 This invention provides a technical solution: a point cloud-based dynamic monitoring and calculation system for earthwork volume, applicable to the aforementioned point cloud-based dynamic monitoring and calculation method for earthwork volume, comprising:

[0151] Point cloud acquisition module 1 is configured to acquire on-site point cloud data of the target construction area at the current monitoring time; perform terrain restoration on the on-site point cloud data, remove vegetation points and temporary facility points to obtain the target on-site point cloud, acquire the measurement range and design elevation of the target construction area, generate a design triangular mesh model, and use the design triangular mesh model to generate a design point cloud to construct a reference voxel mesh model.

[0152] The terrain restoration module 2 is configured to generate a triangular network model of contour lines from the contour line point cloud and the target site point cloud; analyze the spatial distribution characteristics of the target site point cloud to distinguish between vegetation point cloud and ground points; and flatten vegetation points at the same location that are higher than the contour lines to align with the contour lines, thereby removing vegetation.

[0153] The voxel encapsulation module 3 is configured to adjust the side length of the voxel unit according to the distribution density of the target site point cloud, and to encapsulate the target site point cloud into a voxel based on the adjusted side length to obtain the first voxel mesh model.

[0154] Voxel comparison module 4 is configured to compare the first voxel mesh model with the reference voxel mesh model, and extract the first height value of each voxel cell in the first voxel mesh model and the second height value of the corresponding voxel cell in the reference voxel mesh model.

[0155] The difference marking module 5 is configured to calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than the preset height change threshold, the voxel unit is added to the differential voxel set, and the change type of the voxel unit is marked as excavation or filling according to the positive or negative value of the height difference.

[0156] The dynamic quantity calculation module 6 is configured to calculate the first total volume of voxel units marked as excavation type and the second total volume of voxel units marked as fill type in the statistical difference voxel set; the difference between the first total volume and the second total volume is calculated to obtain the net earthwork change, and the dynamic quantity calculation result corresponding to the target construction area is generated based on the net earthwork change.

[0157] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A dynamic monitoring and calculation method for earthwork volume based on point cloud, characterized in that, include: Acquire the on-site point cloud data of the target construction area at the current monitoring time; The terrain is restored from the site point cloud data, and vegetation points and temporary facility points are removed to obtain the target site point cloud; Obtain the measurement range and design elevation of the target construction area, generate a design triangulation model, and use the design triangulation model to generate a design point cloud to construct a reference voxel mesh model. The side length of the voxel unit is adjusted according to the distribution density of the target site point cloud, and the target site point cloud is voxelized and encapsulated based on the adjusted side length to obtain the first voxel mesh model. The first voxel mesh model is compared with the reference voxel mesh model, and the first height value of each voxel unit in the first voxel mesh model and the second height value of the corresponding voxel unit in the reference voxel mesh model are extracted. Calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than a preset height change threshold, add the voxel unit to the differential voxel set and mark the change type of the voxel unit as excavation or filling according to the positive or negative value of the height difference. The first total volume of voxel units labeled as excavation type and the second total volume of voxel units labeled as fill type in the differential voxel set are statistically analyzed; the difference between the first total volume and the second total volume is calculated to obtain the net earthwork change, and the dynamic quantity calculation result corresponding to the target construction area is generated based on the net earthwork change.

2. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 1, characterized in that, The on-site point cloud data includes multiple three-dimensional coordinate points on the surface of the target construction area; The side length of the voxel unit is adjusted according to the distribution density of the target site point cloud, and the target site point cloud is voxelized and encapsulated based on the adjusted side length to obtain a first voxel mesh model, including: Determine the reflection intensity value corresponding to each three-dimensional coordinate point in the target site point cloud; When the reflection intensity value is greater than a preset intensity threshold, the three-dimensional coordinate point is determined as an effective surface point; Calculate the point cloud density of the effective facets, and adjust the side length of the voxel unit in reverse according to the point cloud density. The larger the point cloud density, the smaller the side length of the voxel unit. A three-dimensional mesh is constructed based on the adjusted side length dimensions, and each effective face point is assigned to the corresponding voxel unit to obtain the first voxel mesh model.

3. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 2, characterized in that, The first voxel mesh model is compared with the reference voxel mesh model, and the first height value of each voxel cell in the first voxel mesh model and the second height value of the corresponding voxel cell in the reference voxel mesh model are extracted, including: Traverse each voxel cell in the first voxel mesh model and search for the corresponding voxel cell with the same spatial index in the reference voxel mesh model; Read the average elevation value of the current voxel element in the first voxel mesh model as the first height value; The average elevation value of the corresponding voxel element in the reference voxel mesh model is read as the second height value; Establish a one-to-one correspondence table between the first height value and the second height value.

4. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 3, characterized in that, The statistics include the first total volume of voxel units labeled as cut-out type and the second total volume of voxel units labeled as fill-in type in the differential voxel set, including: Traverse each voxel unit in the set of differential voxels and identify its variation type marker; The volume attribute values ​​of all voxel units marked as cut-out type are summed to obtain the first total volume; The volume attribute values ​​of all voxel units marked as fill type are summed to obtain the second total volume; Record the number of voxel units and their spatial distribution coordinates in the excavation and filling areas respectively.

5. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 4, characterized in that, After generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes: The dynamic quantity calculation results are compared with the preset design earthwork volume to calculate the percentage deviation. A construction progress index is generated based on the deviation percentage, and the progress index is positively correlated with the deviation percentage. The scanning frequency of the point cloud acquisition device is adjusted according to the construction progress index; the higher the progress index, the higher the scanning frequency.

6. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 5, characterized in that, After generating the dynamic quantity calculation results corresponding to the target construction area based on the net earthwork change, the method further includes: Obtain a heat map showing the distribution of the differential voxel set in the plane projection of the target construction area; Based on the heat map, key areas for earthwork operations are determined, and these key areas are those with a color depth greater than a preset threshold. Generate encrypted monitoring instructions for the key areas, the encrypted monitoring instructions including encrypted coordinates and encrypted time; According to the encrypted monitoring command, the measuring equipment is controlled to perform secondary point cloud acquisition on the key area to obtain second point cloud data; The voxel cell states in the corresponding region of the first voxel mesh model are replaced using the second point cloud data to update the local model.

7. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 6, characterized in that, After calculating the first total volume of voxel units labeled as cut-out type and the second total volume of voxel units labeled as fill-in type in the differential voxel set, the method further includes: Identify anomalous voxel units located at the edge of the differential voxel set, where the edge location is the position in the voxel's neighborhood where a null value exists; Obtain the height value of the neighboring voxel units surrounding the abnormal voxel unit; Calculate the median height value of the neighboring voxel units, and replace the height value of the abnormal voxel units with the median to obtain the corrected set of differential voxels. The first and second total volumes are recalculated based on the corrected set of differential voxels, and the dynamic calculation results are updated.

8. The method for dynamic monitoring and calculation of earthwork volume based on point cloud as described in claim 7, characterized in that, Before acquiring the on-site point cloud data and design point cloud data of the target construction area at the current monitoring time, the method further includes: Establish an initial point cloud model of the target construction area; The initial point cloud model is subjected to terrain restoration, and vegetation points and temporary facility points are removed to obtain a denoised initial point cloud model. The denoised initial point cloud model is converted into a triangular mesh model, and the triangular mesh model is used as the basis for constructing the first voxel mesh model; The design point cloud data is converted into a design triangular mesh model, and the design triangular mesh model is aligned with the initial triangular mesh model using the transformation matrix.

9. A point cloud-based dynamic monitoring and calculation system for earthwork volume, applicable to the point cloud-based dynamic monitoring and calculation method for earthwork volume as described in any one of claims 1-8, characterized in that, include: The point cloud acquisition module is configured to acquire on-site point cloud data of the target construction area at the current monitoring time; The site point cloud data is used to perform terrain restoration, remove vegetation points and temporary facility points to obtain the target site point cloud, obtain the measurement range and design elevation of the target construction area, generate a design triangular mesh model, and use the design triangular mesh model to generate the design point cloud to construct a reference voxel mesh model. The terrain reconstruction module is configured to generate a triangular mesh model of contour lines from the contour point cloud and contour point cloud of the target site. Analyze the spatial distribution characteristics of the target site point cloud to distinguish vegetation point cloud from ground points; flatten vegetation points at the same location that are above the contour line to align with the contour line to remove vegetation. The voxel encapsulation module is configured to adjust the side length of the voxel unit according to the distribution density of the target site point cloud, and to encapsulate the target site point cloud into a voxel based on the adjusted side length to obtain a first voxel mesh model. The voxel comparison module is configured to compare the first voxel mesh model with the reference voxel mesh model, and extract the first height value of each voxel unit in the first voxel mesh model and the second height value of the corresponding voxel unit in the reference voxel mesh model. The difference marking module is configured to calculate the height difference between the first height value and the second height value. When the absolute value of the height difference is greater than a preset height change threshold, the voxel unit is added to the differential voxel set, and the change type of the voxel unit is marked as excavation or filling according to the positive or negative value of the height difference. The dynamic quantity calculation module is configured to calculate the first total volume of voxel units marked as excavation type and the second total volume of voxel units marked as fill type in the differential voxel set; perform a difference calculation between the first total volume and the second total volume to obtain the net earthwork change; and generate the dynamic quantity calculation result corresponding to the target construction area based on the net earthwork change.

Images