Building three-dimensional model construction method and system based on digital measurement

Abstract

The invention relates to the technical field of three-dimensional model construction, in particular to a building three-dimensional model construction method and system based on digital measurement, and the method comprises the following steps: carrying out the region separation of building wall surfaces, windows and column structures in data based on collected point cloud data, screening the geometric attributes of each region, and carrying out the matching processing, extracting space contour points, and calculating geometric morphology parameters of multiple regions to obtain a geometric feature mapping table; according to the method, the hierarchy and operability of the data are improved through region separation and geometric attribute matching, so that the three-dimensional model is more accurate when presenting complex building details, design errors caused by insufficient precision of the three-dimensional model are avoided when the practicability and aesthetic property of a building design scheme are evaluated, and the accuracy of the three-dimensional model is improved. Through efficient fitting of the point cloud data and optimization processing of geometric errors, the time cost and potential errors caused by manual adjustment are reduced, and the overall quality of the building model is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional model construction, and in particular to a method and system for constructing a three-dimensional building model based on digital measurement. Background Art

[0002] 3D model construction technology involves the use of computer-aided design (CAD) and other digital tools to generate accurate virtual models from 2D or 3D data, including point cloud processing, surface reconstruction, voxelization, and geometric optimization of these models. In the construction industry, this technology is used to design complex building structures, conduct urban planning, and digitally protect historical buildings.

[0003] Among them, the digital measurement method of building three-dimensional models is to use digital measurement technologies such as laser scanning and optical measurement to create accurate three-dimensional models of buildings. It can capture detailed data from real-world buildings or structures, and then process the data through software tools to form a three-dimensional space representation. The three-dimensional model is widely used in architectural design visualization, structural analysis, building maintenance planning and renovation projects. By providing highly accurate three-dimensional models, architects and engineers can more effectively evaluate the practicality and aesthetics of design schemes.

[0004] Existing technologies are insufficient in accurately capturing and expressing details of highly complex structures, especially when identifying small or subtle architectural features. Key information is often ignored due to technical and methodological limitations. Error handling of point cloud data often lacks efficient automated processes, resulting in more manual intervention in the model building process, increasing time costs and reducing work efficiency. In building maintenance and historical building protection, it is impossible to accurately restore the historical state of the building, affecting the accuracy and effectiveness of restoration and protection work. Summary of the invention

[0005] In order to solve the problems that the existing technologies are insufficient in accurately capturing and expressing details of highly complex structures, especially when identifying small or subtle architectural features, key information is ignored due to technical and methodological limitations, and error processing of point cloud data often lacks efficient automated processes, resulting in more manual intervention in the model building process, increasing time costs and reducing work efficiency. In building maintenance and historical building protection, it is impossible to accurately restore the historical state of the building, affecting the accuracy and effectiveness of restoration and protection work. The embodiment of the present invention provides a method and system for building three-dimensional models based on digital measurement. The technical solution is as follows:

[0006] On the one hand, a method for constructing a three-dimensional building model based on digital measurement is provided, comprising the following steps:

[0007] S1: Based on the collected point cloud data, perform regional separation on the building wall surfaces, windows, and column structures in the data, screen the geometric attributes of each region and perform matching processing, then extract spatial contour points, calculate the geometric shape parameters of multiple regions, and obtain a geometric feature mapping table;

[0008] S2: Based on the geometric feature mapping table, extract the point cloud data of the wall surfaces, windows, and columns according to the features, perform surface geometric equation fitting on each type of data, calculate the surface contour point set, and perform smoothing processing on the boundary points of the 3D model to optimize the geometric error at the model connection, and obtain the initial surface model data;

[0009] S3: Based on the initial surface model data, perform geometric comparison with the original point cloud data, mark the point cloud coordinates deviating from the 3D model surface, calculate the distance between the deviation points and the model surface, and adjust the coordinate positions of the deviation points to obtain a model geometric correction set;

[0010] S4: Based on the model geometric correction set, extract the 3D model boundary and key structure point sets, calculate the cumulative error for the deviation of the structure point set, adjust the geometric position parameters in the point set according to the error results, and analyze the offset rate threshold of each point in the model, and iteratively optimize the positions of the boundary points to obtain a model error correction configuration;

[0011] S5: Based on the model error correction configuration, collect the global feature points of the 3D model, calculate the matching ratio between the feature points and the existing model points, obtain the corresponding comparison matrix, and test the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

[0012] On the other hand, the geometric feature mapping table includes region identification labels, attribute matching indexes, and morphological analysis data. The initial surface model data specifically refers to boundary smoothness, geometric fitting degree, and contour point accuracy. The model geometric correction set includes coordinate adjustment records, distance error data, and adjusted coordinate sets. The model error correction configuration includes error analysis data, point offset rate data, and boundary point optimization results. The building model completion information specifically refers to feature point matching information, comparison matrix data, and consistency test results.

[0013] On the other hand, the steps of, based on the collected point cloud data, performing regional separation on the building wall surfaces, windows, and column structures in the data, screening the geometric attributes of each region and performing matching processing, then extracting spatial contour points, and calculating the geometric shape parameters of multiple regions to obtain a geometric feature mapping table are specifically as follows:

[0014] S101: Based on the collected point cloud data, identify the positions of the building wall surfaces, windows, and columns according to the physical boundaries of the building, divide the spatial positions of different building regions, and perform denoising processing on the point set of each region to screen out error points to obtain a region separation data set;

[0015] S102: Based on the region-separated data set, screen the geometric attributes of the point cloud within each region, perform point cloud clustering through proximity judgment and spatial consistency analysis, aggregate the matching points according to the spatial distribution to form a contour, and re-calibrate the point cloud parameters of the aggregation result to obtain contour point integrated data;

[0016] S103: Based on the contour point integrated data, analyze the coordinate distribution of each point in the point cloud, adjust the shape parameters by point-by-point verification, and detect extreme value deviations in the geometric parameter distribution to obtain a geometric feature mapping table.

[0017] On the other hand, based on the geometric feature mapping table, the steps of extracting the point cloud data of the wall surface, windows and columns according to the features, fitting the surface geometric equation for each type of data, calculating the surface contour point set, and smoothing the boundary points of the three-dimensional model to optimize the geometric error at the model connection to obtain the initial surface model data are specifically as follows:

[0018] S201: Based on the geometric feature mapping table, extract the point cloud data of the wall surface, windows and columns according to the features, screen the points in the point cloud whose error range exceeds the standard value, remove the exceeding points, and at the same time adjust the position information in the point cloud to obtain optimized point cloud data;

[0019] S202: Based on the optimized point cloud data, generate a surface geometric equation through the coordinate distribution of points in three-dimensional space, calculate the distribution range of the surface point set for each building area, and perform point connection to form a continuous geometric surface to obtain a surface geometric model;

[0020] S203: Based on the surface geometric model, calculate the boundary points of the geometric surface point by point, smooth the joint part of the three-dimensional model boundary, remove the error points detected in the calculation, and at the same time correct the error points to obtain the initial surface model data.

[0021] On the other hand, for generating the surface geometric equation through the coordinate distribution of points in three-dimensional space, for each building area, use the formula:

[0022]

[0023] Calculate the distribution range of the surface point set, and perform point connection to form a continuous geometric surface to obtain a surface geometric model, where D i represents the point set distribution diameter of the i-th building area, x j , y j and z j respectively represent the x, y and z coordinates of the j-th point in three-dimensional space, and are the mean values of the points in the x, y and z directions respectively, nD is the total number of points.

[0024] On the other hand, based on the initial draft data of the surface model, geometric comparison is carried out with the original point cloud data, the point cloud coordinates deviating from the surface of the three-dimensional model are marked, the distance between the deviating points and the model surface is calculated, and the coordinate positions of the deviating points are adjusted to obtain the steps of the model geometric correction set as follows:

[0025] S301: Based on the initial draft data of the surface model, geometric comparison is carried out with the original point cloud data, the point cloud coordinates deviating from the surface of the three-dimensional model are marked, and the correctness of each marked point cloud coordinate is verified to obtain the deviating point coordinate set;

[0026] S302: Based on the deviating point coordinate set, calculate the straight-line distance from each marked point to the nearest model surface, measure the distance of each deviating point, and adjust the position of each point one by one according to the measurement result to obtain the position adjustment information;

[0027] S303: Based on the position adjustment information, re-integrate the adjusted point cloud data, and verify whether each point cloud coordinate is adjusted to the target position to obtain the model geometric correction set.

[0028] On the other hand, based on the model geometric correction set, extract the three-dimensional model boundary and key structure point set, calculate the cumulative error for the deviation of the structure point set, adjust the geometric position parameters in the point set according to the error result, and analyze the offset rate threshold of each point in the model, and iteratively optimize the position of the boundary points to obtain the steps of the model error correction configuration as follows:

[0029] S401: Based on the model geometric correction set, extract the three-dimensional model boundary and key structure point set, measure the position deviation of the point set, and analyze the geometric error value of each point to obtain the error measurement data;

[0030] S402: Based on the error measurement data, analyze the offset rate threshold of each point, gradually optimize the geometric position parameters of each point through iterative adjustment, and reduce the position error of the boundary points to obtain the model error correction configuration.

[0031] On the other hand, for extracting the three-dimensional model boundary and key structure point set, measuring the position deviation of the point set, the formula is used:

[0032]

[0033] Analyze the geometric error value of each point to obtain the error measurement data, where, ∈ i represents the geometric error value of the i-th point, p k represents the current coordinate of the measurement point, q k represents the theoretical coordinate, and n represents the total number of key structure points.

[0034] On the other hand, based on the model error correction configuration, global feature points of the three-dimensional model are collected, the matching ratio between the feature points and the existing model points is calculated, the corresponding comparison matrix is obtained, and the matching consistency between the rows and columns of the comparison matrix is tested to obtain the building model completion information. The specific steps are as follows:

[0035] S501: Based on the model error correction configuration, use a three-dimensional scanning device to collect global feature points of the three-dimensional model, detect the geometric features of the model surface point by point, and identify each key geometric feature point to obtain a feature point data directory;

[0036] S502: Based on the feature point data directory, perform one-by-one matching comparison between each feature point and the existing model points, analyze the similarity between each pair of feature points, perform hierarchical evaluation on the matching points, calculate the matching ratio between each pair of feature points, and obtain the matching efficiency analysis result;

[0037] S503: Based on the matching efficiency analysis result, check the matching relationship in the matrix row by row and column by column, and verify the matching consistency of the feature points by analyzing the error distribution in the matrix to obtain the building model completion information.

[0038] On the other hand, a building three-dimensional model construction system based on digital measurement is provided. This system is applied to the building three-dimensional model construction method based on digital measurement and includes:

[0039] The point cloud data analysis module, based on the collected point cloud data, separates the areas of building walls, windows, and column structures in the data, screens the geometric attributes of each area and performs matching processing, calculates the geometric shape parameters of multiple areas, and obtains a geometric feature mapping table;

[0040] The surface fitting module, based on the geometric feature mapping table, extracts the point cloud data of walls, windows, and columns according to the features, calculates the surface contour point set, and smooths the boundary points of the three-dimensional model to obtain the initial surface model data;

[0041] The geometric correction module, based on the initial surface model data, marks the point cloud coordinates deviating from the three-dimensional model surface, calculates the distance between the deviating points and the model surface, and adjusts the coordinate positions of the deviating points to obtain a model geometric correction set;

[0042] The error correction module, based on the model geometric correction set, calculates the cumulative error for the deviation of the structural point set, adjusts the geometric position parameters in the point set according to the error result, and analyzes the offset rate threshold of each point in the model to obtain the model error correction configuration;

[0043] The completion verification module calculates the matching ratio between the feature points and the existing model points based on the model error correction configuration, obtains the corresponding comparison matrix, and tests the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

[0044] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0045] Through region separation and geometric attribute matching, the hierarchy and operability of the data are improved, making the 3D model more accurate when presenting complex building details. Therefore, when evaluating the practicability and aesthetics of the building design scheme, design errors caused by insufficient 3D model accuracy can be avoided. Through the efficient fitting of point cloud data and the optimization of geometric errors, the time cost and potential errors caused by manual adjustment are reduced, and the overall quality of the building model is improved. The automation of geometric correction and error correction greatly reduces the workload and supports the refined management of complex projects and the protection of historical buildings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 is the main step flow chart of the present invention;

[0048] Figure 2 is the step flow chart of S1 of the present invention;

[0049] Figure 3 is the step flow chart of S2 of the present invention;

[0050] Figure 4 is the step flow chart of S3 of the present invention;

[0051] Figure 5 is the step flow chart of S4 of the present invention;

[0052] Figure 6 is the step flow chart of S5 of the present invention;

[0053] Figure 7 is the system block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] The following describes the technical solutions in the present invention with reference to the drawings.

[0055] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to give examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0056] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same. "(of)", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same.

[0057] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0058] To make the technical problems to be solved, technical solutions and advantages of the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0059] The embodiments of the present invention provide a method for constructing a three-dimensional model of a building based on digital measurement, as Figure 1 shown, including the following steps:

[0060] S1: Based on the collected point cloud data, separate the regions of the building walls, windows and column structures in the data, screen the geometric attributes of each region and perform matching processing, then extract the spatial contour points, calculate the geometric shape parameters of multiple regions, and obtain a geometric feature mapping table;

[0061] S2: Based on the geometric feature mapping table, extract the point cloud data of the walls, windows and columns according to the features, fit the surface geometric equations for each type of data, calculate the surface contour point set, and smooth the boundary points of the three-dimensional model to optimize the geometric error at the model connection, and obtain the initial surface model data;

[0062] S3: Based on the initial surface model data, perform geometric comparison with the original point cloud data, mark the point cloud coordinates deviating from the surface of the three-dimensional model, calculate the distance between the deviating points and the model surface, and adjust the coordinate positions of the deviating points to obtain a model geometric correction set;

[0063] S4: Based on the model geometric correction set, extract the three-dimensional model boundary and the key structure point set, calculate the cumulative error for the deviation of the structure point set, adjust the geometric position parameters in the point set according to the error results, analyze the offset rate threshold for each point in the model, and iteratively optimize the positions of the boundary points to obtain the model error correction configuration;

[0064] S5: Based on the model error correction configuration, collect the global feature points of the three-dimensional model, calculate the matching ratio between the feature points and the existing model points, obtain the corresponding comparison matrix, and test the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

[0065] The geometric feature mapping table includes region recognition labels, attribute matching indexes, and morphological analysis data. The surface model draft data specifically includes boundary smoothness, geometric fitting degree, and contour point accuracy. The model geometric correction set includes coordinate adjustment records, distance error data, and the adjusted coordinate set. The model error correction configuration includes error analysis data, point offset rate data, and boundary point optimization results. The building model completion information specifically refers to feature point matching information, comparison matrix data, and consistency test results.

[0066] As Figure 2 shown, based on the collected point cloud data, the steps of separating the regions of the building walls, windows, and columns in the data, screening the geometric attributes of each region and performing matching processing, then extracting the spatial contour points, and calculating the geometric shape parameters of multiple regions to obtain the geometric feature mapping table are specifically as follows:

[0067] S101: Based on the collected point cloud data, identify the positions of the building walls, windows, and columns according to the physical boundaries of the building, divide the spatial positions of different building regions, and perform denoising processing on the point set of each region to filter out error points and obtain the region separation data set;

[0068] By gradually analyzing the original data points in the point cloud and extracting their three-dimensional coordinates, reflection intensities, and corresponding scanning trajectories, identify the physical boundaries of the building. First, select the points in the point cloud that conform to the plane characteristics of the building wall, gradually eliminate non-wall points through normal vector calculation and plane fitting methods, and classify the points that meet geometric continuity into the wall set. At the same time, analyze the linear distribution characteristics perpendicular to the wall in the point cloud, extract the column positions through curve fitting and spatial gradient calculation, then calculate the window contour using the differences in point cloud density and smoothness, combine the change in the reflection intensity of the point cloud to extract the boundary point set of the window. Finally, divide the wall, column, and window regions into independent spatial positions and generate a point set, and combine the threshold screening algorithm based on the Euclidean distance between points to filter out isolated error points and outliers to obtain the separated point cloud data set for each region.

[0069] S102: Based on the region separation data set, the geometric attributes of the point cloud in each region are screened, and the point cloud is clustered through proximity judgment and spatial consistency analysis. The matched points are aggregated according to the spatial distribution to form a contour, and the point cloud parameters of the aggregation result are recalibrated to obtain contour point integration data;

[0070] For the point cloud set in each area, its geometric properties are extracted, including the three-dimensional coordinates of the points, spatial distribution density and surface normal vector. By setting the threshold for point cloud proximity judgment, the proximity relationship is calculated according to the Euclidean distance between points. The points that meet the proximity are clustered using the spatial consistency analysis method. The points of the same category are grouped into a local contour set according to density aggregation. The central coordinates and boundary point distribution of the point cloud in the aggregation result are calculated, and the point cloud parameters are recalibrated to eliminate the offset caused by measurement errors. Accurate contour point integration data is formed by matching the spatial relationship between contour boundary points and the continuity of point cloud projection.

[0071] S103: Based on the contour point integrated data, the coordinate distribution of each point in the point cloud is analyzed, the morphological parameters are adjusted by point-by-point verification, the extreme value deviation in the geometric parameter distribution is detected, and a geometric feature mapping table is obtained;

[0072] First, the point cloud coordinates are gridded and partitioned, and the spatial position relationship of adjacent points is extracted point by point. The morphological parameters of each point are gradually verified by setting detection conditions, including the geometric distance between points, normal vector change and point density distribution. The extreme points in the coordinate distribution that deviate from the set range are eliminated through the median filtering method. Then, the geometric distribution characteristics in the grid area are analyzed, and the morphological parameters of each point are readjusted to make it meet the contour requirements. Finally, a geometric feature mapping table containing areas such as walls, columns and windows is generated.

[0073] like Figure 3 As shown in the figure, based on the geometric feature mapping table, point cloud data of walls, windows and columns are extracted according to the features, surface geometric equations are fitted for each type of data, surface contour point sets are calculated, and the boundary points of the three-dimensional model are smoothed to optimize the geometric errors at the model connection. The specific steps to obtain the preliminary draft data of the surface model are as follows:

[0074] S201: Based on the geometric feature mapping table, point cloud data of the wall, window and column are extracted according to the features, points in the point cloud whose error range exceeds the standard value are screened, and the exceeding points are removed, and the position information in the point cloud is adjusted to obtain optimized point cloud data;

[0075] Perform point-by-point screening on the extracted data set. By analyzing the position coordinates of each point and its spatial relationship with surrounding points, compare whether the position deviation of the point exceeds the preset error range, remove the points that exceed the range, mark and remove the removed points from the data set. At the same time, readjust the position relationship of the point cloud according to the distribution of the remaining points, recalculate the distribution center of the remaining points in the area and the tightness between neighborhoods, and finally form the optimized point cloud data after removing error points and adjusting positions.

[0076] S202: Based on the optimized point cloud data, generate a surface geometry equation through the coordinate distribution of points in three-dimensional space. Calculate the distribution range of the surface point set for each building area, and perform point connection to form a continuous geometric surface to obtain a surface geometry model;

[0077] By analyzing the spatial coordinates of points, gradually generate the surface geometry description of each building area. First, group the points according to their distribution characteristics into different areas. For each grouped area, select the points that conform to the plane or surface characteristics. Gradually generate surface data by analyzing the position distribution and surface continuity of the points. Perform point-to-point connection operations on the surfaces of each area to form a continuous geometric surface. At the same time, check whether there are cases of missing point sets or connection errors in the generated surface data, and complete the optimization of the geometric model by adjusting the boundary positions of the points, thereby forming a surface geometry model that describes the walls, windows, and columns.

[0078] Generate a surface geometry equation through the coordinate distribution of points in three-dimensional space. For each building area, use the formula:

[0079]

[0080] Calculate the distribution range of the surface point set and perform point connection to form a continuous geometric surface to obtain a surface geometry model, where D i represents the diameter of the point set distribution of the i-th building area, which is a measure of the degree of diffusion of the point set in three-dimensional space and is used to estimate the spatial size of the point cloud data in a certain area. x j , y j and z j respectively represent the x, y, and z coordinates of the j-th point in three-dimensional space, and are the means of the points in the x, y, and z directions respectively, and n D is the total number of points;

[0081] There is a point cloud data sample of a building area, and the coordinates of 2 points are collected:

[0082] P1 = (2, 3, 1)

[0083] P2 = (5, 6, 4)

[0084] First, calculate the mean in the x, y, and z directions:

[0085]

[0086] Calculate the sum of the squares of the distances from each point to the mean point:

[0087] For P1:

[0088]

[0089] P1=(-1.5) 2 +(-1.5) 2 +(-1.5) 2 =2.25+2.25+2.25=6.75;

[0090] For P2:

[0091]

[0092] P2=1.5 2 +1.5 2 +1.5 2 =2.25+2.25+2.25=6.75;

[0093] Sum the squares of the total distances:

[0094]

[0095] Substitute the values ​​into the formula to calculate D i :

[0096]

[0097] Get D i The value is 3.674, which represents the average distribution diameter of two points in three-dimensional space relative to their mean point. Calculations show that the point set is highly dispersed in three-dimensional space, that is, the points are relatively dispersed in space. Such measurements help to understand the characteristics of spatial distribution within the building area and have practical reference significance for architectural design and space planning.

[0098] S203: Based on the surface geometric model, the boundary points of the geometric surface are calculated point by point, the joint part of the boundary of the three-dimensional model is smoothed, and the error points detected in the calculation are eliminated and corrected at the same time, so as to obtain the preliminary draft data of the surface model;

[0099] Analyze the boundary characteristics in the point-by-point analysis model, extract the set of points that conform to the boundary features. By analyzing the boundary position relationships of the points, adjust and smooth the boundary point set, eliminate the error points found in the calculation process. By comparing the spatial relationships between the error points and their neighboring points, re-correct the offset points to reasonable positions. Further check the connection parts of the model boundary, refine the joints to ensure a smooth transition of the boundary part. Finally, form a preliminary surface model data to describe the main area boundaries and shapes of the building.

[0100] As Figure 4 shown, based on the preliminary draft data of the surface model, conduct a geometric comparison with the original point cloud data, mark the point cloud coordinates that deviate from the surface of the three-dimensional model, calculate the distance between the deviation points and the model surface, and adjust the coordinate positions of the deviation points. The steps to obtain the model geometric correction set are specifically as follows:

[0101] S301: Based on the preliminary draft data of the surface model, conduct a geometric comparison with the original point cloud data, mark the point cloud coordinates that deviate from the surface of the three-dimensional model, and verify the correctness of each marked point cloud coordinate to obtain the deviation point coordinate set;

[0102] Perform point-by-point matching on the three-dimensional coordinate distributions of the two. By analyzing the spatial relationships between each point in the original point cloud data and the surface model, mark the points that deviate from the model surface by a certain range, classify the points as deviation points. At the same time, utilize the neighborhood information of the original point cloud data to check the spatial coherence and position rationality of each marked point, verify the accuracy of the deviation point markings, exclude the mis-marked points or the points with position deviations less than the acceptable range. Finally, obtain a set of accurate deviation point coordinate sets.

[0103] S302: Based on the deviation point coordinate set, calculate the straight-line distance from each marked point to the nearest model surface, measure the distance for each deviation point, and adjust the position of each point one by one according to the measurement results to obtain the position adjustment information;

[0104] Extract the coordinate information of the deviation points and the coordinate data of the neighboring surface points. By analyzing the spatial vertical projection from the deviation points to the surface, determine the shortest distance for each point. Subsequently, gradually adjust the coordinate positions of the deviation points according to the calculated distance magnitudes. Use the adjusted coordinates to conduct successive comparisons with the surface model to ensure that each point gradually approaches the correct position on the model surface. Finally, output the accurate positions and correction information of each adjusted point to form a complete data set of the position adjustment information.

[0105] S303: Based on the position adjustment information, re-integrate the adjusted point cloud data, and verify whether each point cloud coordinate is adjusted to the target position to obtain the model geometric correction set;

[0106] Re - integrate the adjusted point cloud data. By analyzing the distribution of the adjusted point cloud, compare the matching degree between the target position and the actual position of each point, verify whether each point is successfully adjusted to the target position, and gradually adjust the points whose deviation still exceeds the target range again. Finally, ensure that the adjusted point cloud is consistent with the surface model, forming a complete geometric point cloud data set that has been corrected and calibrated, and finally obtaining a model geometric correction set to describe the optimized building area.

[0107] As Figure 5 shown, based on the model geometric correction set, extract the three - dimensional model boundary and the key structure point set, calculate the cumulative error for the deviation of the structure point set, adjust the geometric position parameters in the point set according to the error result, and analyze the offset rate threshold of each point in the model, and iteratively optimize the position of the boundary points. The specific steps for obtaining the model error correction configuration are as follows:

[0108] S401: Based on the model geometric correction set, extract the three - dimensional model boundary and the key structure point set, measure the position deviation of the point set, and analyze the geometric error value of each point to obtain error metric data;

[0109] Extract the boundary points and the key structure point set of the three - dimensional model. By analyzing the spatial distribution characteristics of the boundary points in the point cloud data, gradually extract the points at the transition positions between regions, sort and group them according to their coordinate values in three - dimensional space, and then select the points with significant geometric characteristics in the model as key structure points. Analyze the spatial position relationship between the key points and the boundary points point by point. By calculating the spatial distance between the boundary points and the key points, compare whether there is a deviation between the geometric distribution of each point and the model structure, record the offset value of the point, and generate geometric error information according to the difference between it and the target position. Finally, organize the position deviation values of each point into an error metric data set.

[0110] Extract the three - dimensional model boundary and the key structure point set, measure the position deviation of the point set, and use the formula:

[0111]

[0112] Analyze the geometric error value of each point to obtain error metric data, where, ∈ i represents the geometric error value of the i - th point, p k represents the current coordinate of the measurement point, q k represents the theoretical coordinate, and n represents the total number of key structure points;

[0113] There is a set of data, where the actual coordinates of three key points are (2, 3, 5), (1, 2, 2), (4, 6, 8) respectively, and the theoretical coordinates are (1, 2, 4), (2, 3, 3), (5, 7, 9) respectively. n is the total number of key structure points, n = 3, substitute into the calculation:

[0114] For the first point, the deviation is

[0115] For the second point, the deviation is

[0116] For the third point, the deviation is

[0117] Summing the squared deviations and dividing by the total number of points n gives:

[0118]

[0119] This result shows that for these three point sets, the root mean square error of the geometric error value is approximately 1.732, representing the average geometric deviation between the measurement result and the theoretical model. Through this error value, the geometric accuracy of the 3D model can be evaluated.

[0120] S402: Based on the error metric data, analyze the offset rate threshold for each point, gradually optimize the geometric position parameters of each point through iterative adjustment, and reduce the position error of the boundary points to obtain the model error correction configuration;

[0121] Analyze the offset rate threshold of the key points point by point, optimize the geometric position parameters of the key points through successive iterative adjustment, classify the points whose offset rate exceeds the set threshold, mark them as high-offset points, and record their initial positions and target positions. Subsequently, adjust the spatial coordinates of each point one by one, limit the adjustment range near the model surface or boundary line, calculate the adjustment direction and step size of each point using the relative position relationship with neighboring points, re-measure and verify the position change after each adjustment to ensure that the offset rate of the points gradually decreases after each iterative optimization. At the same time, handle the adjustment of the boundary points separately, correct the position distribution of the boundary points by increasing the point cloud density to obtain the model error correction configuration.

[0122] As Figure 6 shown, based on the model error correction configuration, the steps to collect the global feature points of the 3D model, calculate the matching ratio between the feature points and the existing model points, obtain the corresponding comparison matrix, and test the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information are specifically as follows:

[0123] S501: Based on the model error correction configuration, use a 3D scanning device to collect the global feature points of the 3D model, detect the geometric features on the model surface point by point, and identify each key geometric feature point to obtain the feature point data directory;

[0124] The global feature points of the 3D model are collected using a 3D scanning device. First, the entire model surface is divided into grids for zoning to facilitate the collection of feature points zone by zone. Then, key points are identified within each zone according to the significance of geometric features, such as sharp edges, surface intersection points, and regions with significant concave and convex changes. The 3D coordinates of each feature point are recorded point by point. Subsequently, the distribution of feature points in the entire model is analyzed, and feature points with repeatability or those that can represent the geometric shape of the model are screened. Finally, the feature points are classified by region to generate a feature point data directory.

[0125] S502: Based on the feature point data directory, each feature point is matched and compared with the existing model points one by one. The similarity between each pair of feature points is analyzed, and the matching points are evaluated hierarchically. The matching ratio between each pair of feature points is calculated to obtain the matching efficiency analysis result.

[0126] By normalizing the 3D coordinates of the feature points to make their coordinate systems consistent with those of the existing model points, the spatial distance between two points is calculated point by point. Points with a distance less than the preset threshold are determined as matching points. At the same time, combining the local geometric features of the two points, including the normal vector direction and curvature change, the geometric similarity of each pair of feature points is further analyzed. The matching points are classified, sorted from high to low according to similarity, and the matching level is marked. The matching ratio of all feature points is gradually counted, and the overall matching ratio is calculated to generate the matching efficiency analysis result.

[0127] S503: Based on the matching efficiency analysis result, the matching relationships in the matrix are checked row by row and column by column. By analyzing the error distribution in the matrix, the matching consistency of the feature points is verified to obtain the building model completion information.

[0128] The similarity distribution of the matching point pairs is mapped into the matrix. By analyzing the matching level of each point pair row by row, point pairs with significant errors or low matching levels are screened out and marked as error points separately. Subsequently, the overall matching trend and the error concentration distribution area of the point pairs are determined through column - by - column comparison. The distribution range of the error points is analyzed spatially, and the abnormal areas are recorded. Combining the analysis, the matching consistency of the feature points is verified, and finally, the building model completion information is generated to describe the global geometric characteristics of the 3D model and the matching situation of the key structural points.

[0129] As Figure 7 shown, a building 3D model construction system based on digital measurement includes:

[0130] The point cloud data analysis module separates the building wall surfaces, windows, and column structures in the data based on the collected point cloud data, screens the geometric attributes of each region, performs matching processing, calculates the geometric shape parameters of multiple regions, and obtains a geometric feature mapping table.

[0131] Based on the geometric feature mapping table, the surface fitting module extracts the point cloud data of the wall, window and column according to the features, calculates the surface contour point set, and smooths the boundary points of the three-dimensional model to obtain the initial data of the surface model;

[0132] Based on the initial data of the surface model, the geometric correction module marks the point cloud coordinates deviating from the surface of the three-dimensional model, calculates the distance between the deviating points and the model surface, and adjusts the coordinate positions of the deviating points to obtain the model geometric correction set;

[0133] Based on the model geometric correction set, the error correction module calculates the cumulative error for the deviation of the structural point set, adjusts the geometric position parameters in the point set according to the error results, and analyzes the offset rate threshold of each point in the model to obtain the model error correction configuration;

[0134] Based on the model error correction configuration, the completion verification module calculates the matching ratio between the feature points and the existing model points, obtains the corresponding comparison matrix, and tests the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

[0135] It should be understood that the term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Among them, A and B can be singular or plural. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0136] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single item (s) or plural item (s). For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0137] It should be understood that in various embodiments of the present invention, the magnitudes of the serial numbers of the above processes do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0138] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0139] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0140] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0141] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0142] In addition, the functional units in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0143] When the above-described function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0144] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for constructing a three-dimensional building model based on digital measurement, characterized in that: The method comprises: S1: Based on the collected point cloud data, the building walls, windows and column structures in the data are separated into regions, the geometric attributes of each region are screened and matched, and then the spatial contour points are extracted, the geometric morphological parameters of multiple regions are calculated, and the geometric feature mapping table is obtained; S2: Based on the geometric feature mapping table, point cloud data of walls, windows and columns are extracted according to the features, surface geometric equations are fitted for each type of data, surface contour point sets are calculated, and boundary points of the three-dimensional model are smoothed to optimize the geometric errors at the model connection points, thereby obtaining preliminary surface model data; S3: Based on the draft surface model data, a geometric comparison is performed with the original point cloud data, the point cloud coordinates that deviate from the three-dimensional model surface are marked, the distance between the deviated point and the model surface is calculated, and the coordinate position of the deviated point is adjusted to obtain a model geometry correction set; S4: Based on the model geometry correction set, extract the boundary and key structure point set of the three-dimensional model, calculate the cumulative error for the deviation of the structure point set, adjust the geometric position parameters in the point set according to the error result, analyze the offset rate threshold of each point in the model, iteratively optimize the position of the boundary point, and obtain the model error correction configuration; S5: Based on the model error correction configuration, collect the global feature points of the three-dimensional model, calculate the matching ratio between the feature points and the existing model points, obtain the corresponding comparison matrix, and test the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

2. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: The geometric feature mapping table includes area identification labels, attribute matching indexes, and morphological analysis data; the surface model draft data specifically includes boundary smoothness, geometric fit, and contour point accuracy; the model geometry correction set includes coordinate adjustment records, distance error data, and adjusted coordinate sets; the model error correction configuration includes error analysis data, point offset rate data, and boundary point optimization results; the building model completion information specifically refers to feature point matching information, comparison matrix data, and consistency test results.

3. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: Based on the collected point cloud data, the building walls, windows and column structures in the data are separated into regions, the geometric attributes of each region are screened and matched, and then the spatial contour points are extracted, and the geometric morphological parameters of multiple regions are calculated. The specific steps to obtain the geometric feature mapping table are as follows: S101: Based on the collected point cloud data, the locations of building walls, windows and columns are identified according to the physical boundaries of the building, the spatial locations of different building areas are divided, and the point set of each area is denoised to filter out error points to obtain a regional separation data set; S102: Based on the region separation data set, the geometric attributes of the point cloud in each region are screened, the point cloud is clustered through proximity judgment and spatial consistency analysis, the matching points are aggregated according to the spatial distribution to form a contour, and the point cloud parameters of the aggregation result are recalibrated to obtain contour point integration data; S103: Based on the contour point integrated data, the coordinate distribution of each point in the point cloud is analyzed, the morphological parameters are adjusted by point-by-point verification, the extreme value deviation in the geometric parameter distribution is detected, and a geometric feature mapping table is obtained.

4. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: Based on the geometric feature mapping table, point cloud data of walls, windows and columns are extracted according to the features, surface geometric equations are fitted for each type of data, surface contour point sets are calculated, and the boundary points of the three-dimensional model are smoothed to optimize the geometric errors at the model connection points. The specific steps for obtaining the preliminary draft data of the surface model are as follows: S201: Based on the geometric feature mapping table, extract point cloud data of walls, windows and columns according to the features, filter out points in the point cloud whose error range exceeds the standard value, remove the points that exceed the standard value, and adjust the position information in the point cloud to obtain optimized point cloud data; S202: Based on the optimized point cloud data, a surface geometric equation is generated through the coordinate distribution of points in the three-dimensional space, the distribution range of the surface point set is calculated for each building area, and the points are connected to form a continuous geometric surface to obtain a surface geometric model; S203: Based on the surface geometric model, the boundary points of the geometric surface are calculated point by point, the joint parts of the three-dimensional model boundary are smoothed, and the error points detected in the calculation are eliminated and corrected to obtain the preliminary draft data of the surface model.

5. The method for constructing a three-dimensional building model based on digital measurement according to claim 4, characterized in that: The surface geometric equation is generated by the coordinate distribution of the points in the three-dimensional space, and the formula is used for each building area: ； Calculate the distribution range of the surface point set, connect the points to form a continuous geometric surface, and obtain the surface geometric model, where: Representative The point set distribution diameter of the building area, and Respectively represent The x, y, and z coordinates of a point in three-dimensional space, and are the means of the points in the x, y and z directions respectively, is the total number of points.

6. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: Based on the draft surface model data, geometric comparison is performed with the original point cloud data, the point cloud coordinates that deviate from the three-dimensional model surface are marked, the distance between the deviated point and the model surface is calculated, and the coordinate position of the deviated point is adjusted to obtain the model geometry correction set. Specifically, the steps are as follows: S301: Based on the draft surface model data, geometric comparison is performed with the original point cloud data, the point cloud coordinates that deviate from the three-dimensional model surface are marked, and the correctness of each point cloud coordinate mark is verified to obtain a deviated point coordinate set; S302: Based on the deviation point coordinate set, calculate the straight-line distance from each marked point to the nearest model surface, measure the distance of each deviation point, and adjust the position of each point one by one according to the measurement result to obtain position adjustment information; S303: Based on the position adjustment information, the adjusted point cloud data is reintegrated, and it is verified whether each point cloud coordinate is adjusted to the target position to obtain a model geometry correction set.

7. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: Based on the model geometry correction set, the three-dimensional model boundary and key structure point set are extracted, the cumulative error is calculated for the deviation of the structure point set, the geometric position parameters in the point set are adjusted according to the error result, and the offset rate threshold of each point in the model is analyzed. The position of the boundary point is iteratively optimized, and the steps of obtaining the model error correction configuration are specifically as follows: S401: extracting the boundary and key structure point set of the three-dimensional model based on the model geometry correction set, measuring the position deviation of the point set, and analyzing the geometric error value of each point to obtain error measurement data; S402: Based on the error measurement data, the offset rate threshold of each point is analyzed, the geometric position parameters of each point are gradually optimized through iterative adjustment, and the position error of the boundary points is reduced to obtain a model error correction configuration.

8. The method for constructing a three-dimensional building model based on digital measurement according to claim 7, characterized in that: The three-dimensional model boundary and key structure point set are extracted, and the position deviation of the point set is measured using the formula: ； Analyze the geometric error value of each point to obtain error measurement data, where: Representative The geometric error value of each point, Represents the current coordinates of the measurement point, represents the theoretical coordinates, Represents the total number of key structural points.

9. The method for constructing a three-dimensional building model based on digital measurement according to claim 1, characterized in that: Based on the model error correction configuration, the global feature points of the three-dimensional model are collected, the matching ratio between the feature points and the existing model points is calculated, the corresponding comparison matrix is ​​obtained, and the matching consistency between the rows and columns of the comparison matrix is ​​tested to obtain the building model completion information. Specifically, the steps are as follows: S501: Based on the model error correction configuration, a three-dimensional scanning device is used to collect global feature points of the three-dimensional model, the geometric features of the model surface are detected point by point, and each key geometric feature point is identified to obtain a feature point data directory; S502: Based on the feature point data directory, each feature point is matched and compared with the existing model points one by one, the similarity between each pair of feature points is analyzed, and the matching points are graded and evaluated, and the matching ratio between each pair of feature points is calculated to obtain the matching efficiency analysis result; S503: Based on the matching efficiency analysis result, the matching relationship in the matrix is ​​checked row by row and column by column, and the matching consistency of the feature points is verified by analyzing the error distribution in the matrix to obtain the building model completion information.

10. A system for constructing a three-dimensional building model based on digital measurement, wherein the system is used to implement the method for constructing a three-dimensional building model based on digital measurement as claimed in any one of claims 1 to 9, characterized in that: The system comprises: The point cloud data analysis module separates the building walls, windows and column structures in the data based on the collected point cloud data, selects the geometric attributes of each area and performs matching processing, calculates the geometric morphological parameters of multiple areas, and obtains a geometric feature mapping table; The surface fitting module extracts point cloud data of walls, windows and columns according to the features based on the geometric feature mapping table, calculates the surface contour point set, and smoothes the boundary points of the three-dimensional model to obtain the preliminary draft data of the surface model; The geometric correction module marks the point cloud coordinates that deviate from the surface of the three-dimensional model based on the surface model draft data, calculates the distance between the deviated point and the model surface, and adjusts the coordinate position of the deviated point to obtain a model geometric correction set; The error correction module calculates the cumulative error of the deviation of the structure point set based on the model geometry correction set, adjusts the geometric position parameters in the point set according to the error result, and analyzes the offset rate threshold of each point in the model to obtain the model error correction configuration; The completion verification module calculates the matching ratio between the feature points and the existing model points based on the model error correction configuration, obtains the corresponding comparison matrix, and tests the matching consistency between the rows and columns of the comparison matrix to obtain the building model completion information.

Images