Construction wall inclination detection method and system suitable for construction engineering

Abstract

The application relates to the technical field of wall body inclination detection, in particular to a construction wall body inclination detection method and system suitable for building engineering, which comprises the following steps: acquiring a three-dimensional point cloud model of the building engineering, and identifying each structural surface in the three-dimensional point cloud model; determining a ground surface structural surface in all the structural surfaces, and fitting a ground surface reference plane; carrying out object segmentation on the three-dimensional point cloud model, and determining the wall body confidence of each object; carrying out threshold segmentation on the wall body confidence of all the objects, and acquiring the construction wall body; determining the concave-convex coefficient of each point in each construction wall body, and obtaining the concave-convex points in each construction wall body; eliminating the concave-convex points in each construction wall body, determining the reference plane of each construction wall body, and obtaining the perpendicularity deviation of each construction wall body through the normal vector angle between the reference plane of each construction wall body and the ground surface reference plane, and combining the actual height of the construction wall body, so that the inclination of the construction wall body is detected. Therefore, the precision of the construction wall body inclination detection is improved.

Description

Technical Field

[0001] This application relates to the field of wall tilt detection technology, specifically to a method and system for detecting the tilt of construction walls applicable to building engineering. Background Technology

[0002] The quality of a building structure is crucial to the personal and property safety of residents, and structural health monitoring and assessment is a vital part of construction engineering. As a key load-bearing component of a building structure, the tilt of walls directly affects the stability and safety of the building. Accurate wall tilt detection can promptly identify construction deviations, effectively preventing serious problems such as deformation and collapse. Therefore, tilt detection of walls during construction helps ensure the safety of residents and the building itself, while also reducing economic losses caused by quality issues.

[0003] Current technologies typically employ 3D laser scanning to scan and locate walls within point cloud data of building structures. The angle between the fitted wall plane and the ground normal vector is then directly calculated to assess wall tilt. However, these technologies fail to adequately consider the presence of debris at wall corners on construction sites and the potential for uneven wall surfaces due to carving or uneven application. Therefore, directly calculating the angle between the fitted wall plane and the ground normal vector leads to significant errors in the final tilt detection results, reducing the accuracy of wall tilt detection. Summary of the Invention

[0004] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for detecting the tilt of construction walls in building engineering. The specific technical solution adopted is as follows:

[0005] In a first aspect, embodiments of this application provide a method for detecting the tilt of construction walls in building engineering, the method comprising the following steps:

[0006] Obtain a 3D point cloud model of the building project and identify the structural surfaces in the 3D point cloud model;

[0007] Based on the numerical distribution of z-coordinate values ​​of all points within each structural plane, the surface structural plane is determined among all structural planes, and the surface reference plane is fitted.

[0008] The wall confidence of each object is determined by the magnitude of the interior point ratio of all structural surfaces in each object, the proportion of points belonging to structural surfaces in each object, and the fluctuation of the proportion of the proportion of each object and its connected objects.

[0009] Threshold segmentation is performed on the wall confidence of all objects to obtain the construction walls; based on the degree of curvature fluctuation of each point in each construction wall and its nearest neighbor, and the local in-point rate of each point obtained by plane fitting of each point in each construction wall and its nearest neighbor, the concavity and convexity coefficients of each point in each construction wall are determined to obtain the concavity and convexity points in each construction wall.

[0010] After removing any uneven points in each construction wall, the reference surface of each construction wall is determined. By using the angle between the normal vectors of each construction wall's reference surface and the ground reference surface, combined with the actual height of the construction wall, the verticality offset of each construction wall is obtained, and the tilt of the construction wall is detected.

[0011] Calculate the mean and dispersion of the z-coordinate values ​​of all points within each structural plane, and take the structural plane with the smallest fusion result of the mean and dispersion as the surface structural plane.

[0012] In one embodiment, fitting the surface reference surface includes:

[0013] The x and y coordinates of all points within the surface structure are fitted using a plane fitting algorithm, and the mean value of the surface structure is used as the uniform height value of the fitted plane to obtain the surface reference surface.

[0014] In one embodiment, determining the wall confidence level includes:

[0015] Calculate the mean of the interior point ratio of all structural surfaces in each object, denoted as the first mean; calculate the dispersion of the quantity ratio of each object and its connected objects, denoted as the first dispersion; combine the first mean, the quantity ratio, and the first dispersion to determine the confidence level of the wall.

[0016] The confidence level of the wall is positively correlated with the first mean and the proportion of quantities, and negatively correlated with the first dispersion.

[0017] In one embodiment, further determination of the wall confidence level includes:

[0018] Calculate the product of the first mean and the quantity ratio, calculate the sum of the first dispersion and a preset value greater than 0, and determine the ratio of the product to the sum as the confidence level of the wall.

[0019] In one embodiment, the construction wall is an object with a wall confidence level greater than or equal to a segmentation threshold.

[0020] In one embodiment, determining the unevenness coefficient at each point in each constructed wall includes:

[0021] The degree of dispersion of the curvature of each point in each construction wall and its nearest neighbor is calculated and denoted as the second dispersion. The concavity coefficient of each point in each construction wall is the ratio of the second dispersion to the local in-point rate.

[0022] In one embodiment, the bumps and depressions in each construction wall are defined by thresholding the bump and depression coefficients of all points in each construction wall, with the bump and depression coefficients being greater than or equal to the threshold. The reference surface of each construction wall is defined by filling the construction wall with point cloud after removing the bumps and depressions, and using the two largest faces in the three-dimensional minimum bounding rectangle of the construction wall after point cloud filling as the reference surface of the construction wall.

[0023] In one embodiment, the process of obtaining the verticality offset is as follows:

[0024] Calculate the angle between the normal vectors of the two reference planes of each construction wall and the ground reference plane. Record the absolute value of the difference between the mean of the two angles of each construction wall and 90° as the verticality of each construction wall.

[0025] The product of the sine function value of the verticality of each construction wall and its actual height is used as the verticality offset of each construction wall.

[0026] Secondly, embodiments of this application also provide a construction wall tilt detection system suitable for building engineering, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.

[0027] This application has at least the following beneficial effects:

[0028] This application constructs a ground reference surface by acquiring the surface structure surface from the 3D point cloud model of the building project. This effectively avoids interference from construction debris on the surface fitting plane, ensuring that the measurement of tilt angles is no longer affected by surface irregularities. This provides a reliable foundation for subsequent verticality offset analysis and ensures high reliability of subsequent calculation results. By constructing the wall confidence level of each object, walls in the point cloud data can be accurately located, effectively distinguishing walls from interfering debris and avoiding misidentification and missed detection of walls. This ensures that only highly reliable walls are extracted, reducing potential errors in subsequent tilt detection. Furthermore, by determining the concavity coefficient of each point in each construction wall, this application can accurately assess the concavity and convexity of the wall surface. This ensures that the irregularity of the wall surface is taken into account during tilt detection, avoiding misjudgments caused by uneven surfaces. This allows for the removal of points in concave and convex areas of the wall, thereby constructing a more reliable wall reference surface.

[0029] This application addresses the problem that existing technologies do not adequately consider the interference of construction site debris on the ground and wall fitting plane, leading to significant errors in the calculation of wall tilt. This application constructs a wall reference plane and combines it with a ground surface reference plane to calculate the vertical offset of each construction wall, thereby accurately assessing the tilt of the construction wall. This eliminates the influence of ground debris and uneven wall surfaces on wall tilt detection, improving the accuracy and reliability of wall tilt detection and ensuring the safety of building structures. Attached Figure Description

[0030] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A flowchart illustrating the steps of a construction wall tilt detection method applicable to building engineering, provided as an embodiment of this application;

[0032] Figure 2 Flowchart for determining the verticality offset of the construction wall. Detailed Implementation

[0033] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the construction wall tilt detection method and system applicable to building engineering proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0035] The following, in conjunction with the accompanying drawings, details the specific scheme of the construction wall tilt detection method and system applicable to building engineering provided in this application.

[0036] Please see Figure 1 The diagram illustrates a flowchart of a construction wall tilt detection method applicable to building engineering provided in an embodiment of this application. The method includes the following steps:

[0037] S1: Obtain the 3D point cloud model of the building project and identify the structural surfaces in the 3D point cloud model.

[0038] Multiple ground-based laser scanners are deployed around the construction site to comprehensively collect 3D point cloud data of the building. In this embodiment, the distance between each laser scanner is 8 meters, and the overlap rate is 50%. Implementers can set this distance according to their actual needs; this embodiment does not impose any restrictions. After scanning, the point cloud data collected by adjacent scanners are stitched together. Then, a filtering algorithm is used to filter the point cloud data, resulting in a filtered 3D point cloud model of the building. In this embodiment, the point cloud stitching method uses the ICP registration algorithm, and the filtering algorithm uses Gaussian filtering. Both ICP registration and Gaussian filtering are existing, well-known technologies. Implementers can choose other feasible point cloud stitching methods and filtering algorithms.

[0039] The 3D point cloud model of the building is used as input to the point cloud structural surface segmentation algorithm, thereby identifying multiple structural surfaces in all point cloud data within the 3D point cloud model. Point cloud structural surface segmentation methods are not limited to the RANSAC (Random Sample Consensus) algorithm or the Region Growing method of the PCL point cloud library. This embodiment uses the RANSAC algorithm to obtain each structural surface in the 3D point cloud model. The RANSAC algorithm is a well-known existing technology, and its specific process will not be elaborated upon.

[0040] S2, based on the numerical distribution of z-coordinate values ​​of all points within each structural plane, determines the surface structural plane among all structural planes and fits the surface reference plane.

[0041] Because construction sites typically contain a large amount of debris, such as construction equipment, building materials, and broken bricks and stones, and debris easily accumulates at the corners of walls, errors can easily occur in the fitted ground surface reference plane and the fitted wall surface, thus affecting the accuracy of the wall tilt calculation. Therefore, to obtain a more accurate wall tilt, it is first necessary to construct a flat ground surface reference plane.

[0042] Since the z-coordinate of the surface reference plane is the smallest and relatively consistent among all structural planes, this embodiment locates the surface structural planes by analyzing the z-coordinates of each structural plane and the differences between the z-coordinates.

[0043] Taking the j-th structural surface in the 3D point cloud model as an example, the mean and dispersion of the z-coordinates of all points in the j-th structural surface are calculated respectively, and the fusion result of the mean and the dispersion is obtained. The structural surface corresponding to the minimum value of the fusion result of all structural surfaces in the 3D point cloud model is taken as the surface structural surface.

[0044] It should be noted that the degree of dispersion can be calculated using methods such as variance, standard deviation, and coefficient of variation; fusion means combining multiple variables, which can be calculated using methods such as addition, multiplication, and averaging.

[0045] In this embodiment, the sum of the mean and variance of the z-coordinates of all points within the j-th structural plane is calculated and denoted as the first sum. The first sum reflects whether the height of each point within the j-th structural plane is generally low in three-dimensional space. The smaller the first sum, the lower the overall height of the j-th structural plane and the more consistent the height.

[0046] The first sum of all structural surfaces in the 3D point cloud model is calculated in the same way, and the structural surface with the smallest first sum is recorded as the ground structural surface. Since the ground may have slight undulations due to dust, materials, etc., in construction scenarios, these slight undulations may cause errors in the calculation of wall tilt, thus requiring further processing.

[0047] A plane fitting algorithm is used to fit the x and y coordinates of all points within the surface structure plane, and the mean value of the surface structure plane is used as the uniform height value of the fitted plane, thereby constructing an ideal surface reference plane. The plane fitting algorithm is not limited to the least squares method or the RANSAC algorithm; this embodiment uses the least squares method for plane fitting, which is a well-known existing technique, and the specific process will not be described in detail.

[0048] S3. Perform object segmentation on the 3D point cloud model. Determine the wall confidence of each object by the magnitude of the interior point ratio of all structural surfaces in each object, the proportion of points belonging to structural surfaces in each object, and the fluctuation of the proportion of the proportion of each object and its connected objects.

[0049] After constructing the ground reference surface, it is necessary to perform planar fitting on each wall to calculate the wall tilt. Therefore, it is necessary to locate and identify the walls in the 3D point cloud data.

[0050] The PointNet network is used to segment the 3D point cloud model of a building into multiple objects; the probability of the i-th object being a wall is analyzed. The dataset used for training the PointNet network in this embodiment is the S3DIS dataset. The training and use of the PointNet network are well-known techniques, and will not be described in detail here.

[0051] Although personalized carving and uneven application can cause unevenness on the wall surface, the surface of construction equipment and materials is even more uneven. Therefore, this embodiment still analyzes the possibility that the object is a wall based on its flatness.

[0052] The mean of the interior point ratios of all structural surfaces in the i-th object is obtained and denoted as the first mean of the i-th object. A higher interior point ratio indicates a more concentrated distribution of points on the structural surface, resulting in a smoother surface. Therefore, a larger first mean indicates a smoother structural surface in the i-th object, increasing the likelihood that it is a wall. It should be noted that the interior point ratios of the structural surfaces are obtained when fitting the structural surfaces using the RANSAC algorithm.

[0053] Simultaneously, the ratio between the number of all points belonging to structural surfaces in the i-th object and the total number of points in the i-th object is calculated and denoted as the first ratio of the i-th object. The first ratio reflects the proportion of points belonging to structural surfaces in the i-th object; the larger the first ratio, the more likely that most points of the i-th object are distributed on a flat plane, the greater the probability that the i-th object has a regular geometric shape, and the smaller the probability that it is a miscellaneous object.

[0054] Furthermore, since walls are usually connected to each other, and the first ratio between walls is relatively consistent, this embodiment analyzes the degree of difference between the first ratio of the i-th object and its connected objects to assess whether the i-th object may be a wall.

[0055] Get the maximum coordinates of a point inside the i-th object on each coordinate axis. minimum coordinates The bounding box of the i-th object is obtained using the AABB bounding box algorithm. Similarly, the bounding boxes of all objects in the 3D point cloud model are obtained. The bounding box of the i-th object and the bounding boxes of every other object are used as inputs to the calOverTotal method, which outputs the overlap values ​​between the i-th object and all other objects. Objects with positive overlap values ​​are designated as connected objects of the i-th object. Both the AABB bounding box algorithm and the calOverTotal method are well-known existing technologies, and their specific processes are not detailed here.

[0056] Calculate the first ratio between the i-th object and each of its connected objects, and then denote the degree of dispersion of the first ratio between the i-th object and all its connected objects as the first dispersion of the i-th object. The first dispersion reflects whether the i-th object and its connected objects have relatively consistent regularity. The smaller the first dispersion, the smaller the difference in geometric regularity between the i-th object and its connected objects, and the greater the probability that the i-th object is a wall. In this embodiment, the first dispersion of the i-th object is the variance of the first ratio between the i-th object and all its connected objects.

[0057] Construct the wall confidence of the i-th object The specific expression is:

[0058]

[0059] In the formula, Let be the first mean of the i-th object; The first ratio of the i-th object; Let be the first discreteness of the i-th object; To avoid a denominator of 0, the preset value is greater than 0. In this embodiment, The implementer can set it according to the actual situation, and this embodiment does not impose any restrictions on it.

[0060] The wall confidence score reflects the probability that the i-th object is a wall; the higher the wall confidence score, the more consistent the characteristics of the i-th object are with the characteristics of a wall, and the greater the probability that the i-th object is a wall.

[0061] S4. Threshold segmentation is performed on the wall confidence of all objects to obtain the construction walls. Based on the degree of curvature fluctuation of each point in each construction wall and its nearest neighbor, and the local in-point rate of each point obtained by plane fitting of each point in each construction wall and its nearest neighbor, the concavity and convexity coefficients of each point in each construction wall are determined to obtain the concavity and convexity points in each construction wall.

[0062] Considering that the wall surface may be uneven due to personalized carving or poor construction results, and that a large protrusion would directly lead to a large deviation in the fitted wall surface, we cannot directly perform planar fitting using all the point cloud data of the wall and further analysis is required.

[0063] First, the wall confidence score of each object is calculated, and all wall confidence scores are used as input to the Otsu threshold segmentation algorithm. The output is the segmentation threshold. Objects with wall confidence scores greater than or equal to the segmentation threshold are identified as construction walls, thereby locating the wall positions in the point cloud data.

[0064] Furthermore, taking the v-th point in the u-th construction wall as an example, we evaluate the confidence level that it is a point in a flat region. If the v-th point is a point in a flat region, then the curvature of the v-th point and all its surrounding points are relatively consistent, and the better the plane fitting effect is achieved through points in this local region.

[0065] Taking point v as the center, obtain the nearest neighbor points of point v. In this embodiment, calculate the Euclidean distance between point v and the remaining points in the same construction wall, and sort the Euclidean distances in ascending order. The first q points after sorting are taken as the nearest neighbors of point v. In this embodiment, q=20, but the implementer can set it according to the actual situation; this embodiment does not impose any restrictions. Calculate the degree of dispersion of the curvature between point v and all its nearest neighbors, denoted as the second dispersion of point v. The second dispersion reflects whether point v is located in a locally flat region. The larger the second dispersion, the greater the probability that point v is located in a locally flat region. The calculation of curvature is a well-known existing technique and will not be elaborated further. In this embodiment, the second dispersion of point v is the variance of the curvature between point v and all its nearest neighbors.

[0066] The v-th point and all its nearest neighbors are used as inputs to the RANSAC algorithm for plane fitting. The inlier rate of the plane fitting is then denoted as the local inlier rate of the v-th point.

[0067] The ratio between the second dispersion of point v and the local inlier rate of point v is denoted as the concavity coefficient of point v. The concavity coefficient reflects the probability that point v is located in a flat region; the smaller the concavity coefficient, the greater the probability that point v is located in a flat region.

[0068] The concavity / convexity coefficients of all points in the u-th construction wall are calculated in the same way and used as input to the Otsu threshold segmentation algorithm. The segmentation threshold is output, and all points with concavity / convexity coefficients greater than or equal to the segmentation threshold are regarded as concavity / convex points.

[0069] S5. Eliminate the uneven points in each construction wall, determine the reference surface of each construction wall, and obtain the verticality offset of each construction wall by combining the normal vector angle between the reference surface of each construction wall and the ground reference surface, and the actual height of the construction wall, and perform tilt detection on the construction wall.

[0070] In the u-th construction wall, all bumps and depressions are removed. After removing the bumps and depressions, holes may appear in the point cloud model of the u-th construction wall. Also, considering that if there are debris at the corners of the u-th construction wall, these corners will be missing after object segmentation, a point cloud filling method is used to fill the point cloud model of the u-th construction wall after removing the bumps and depressions. The point cloud filling method is not limited to methods based on plane fitting or curvature continuity; this embodiment uses a plane fitting method. The point cloud filling method is a known technique and will not be described further here.

[0071] Furthermore, the minimum 3D bounding rectangle of the u-th construction wall after point cloud filling is obtained, and then the two faces with the largest areas in the minimum 3D bounding rectangle are used as the reference faces of the u-th construction wall. This removes the error influence of protruding parts in the original wall surface on the fitted plane, thereby improving the accuracy of subsequent wall tilt calculations.

[0072] Continuing with the example of the u-th construction wall, we obtain the normal vectors of the two reference planes and the ground reference plane of the u-th construction wall. Then, we calculate the angle between the normal vectors of the two reference planes and the ground reference plane, and record the absolute value of the difference between the mean of the two angles and 90° as the verticality of the u-th construction wall. The calculation of the angle between the normal vectors is a well-known technique and will not be elaborated here.

[0073] Finally, the verticality offset of the u-th construction wall is calculated. The specific expression is:

[0074]

[0075] In the formula, The actual height of the uth construction wall Let be the verticality of the u-th construction wall, and sin() be the trigonometric sine function. Determine the verticality offset of the u-th construction wall. Is it less than The tilt of the u-th construction wall is determined by comparing it to the minimum value of 30mm. If the tilt is less than this value, the tilt of the u-th construction wall is considered to meet the building specifications. If it is not less than this value, the u-th construction wall has a relatively serious tilt and needs to be reinforced immediately. The tilt of all construction walls in the building project is checked in the same way. The verticality offset is also determined. Is it less than The minimum value among 30mm is a well-known technique in the field of building construction. The flowchart for determining the verticality offset of a construction wall is shown below. Figure 2 As shown.

[0076] Based on the same inventive concept as the above methods, this application also provides a construction wall tilt detection system suitable for building engineering, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods for construction wall tilt detection suitable for building engineering.

[0077] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0078] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0079] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A method for detecting the inclination of a construction wall for building works, characterized in that, The method comprises the following steps: The method comprises the following steps: Obtain a three-dimensional point cloud model of a construction project, and identify each structural surface in the three-dimensional point cloud model; Determine a ground surface structural surface among all structural surfaces based on the numerical distribution of the z coordinate values of all points in each structural surface, and fit a ground reference surface; Object segmentation is performed on the three-dimensional point cloud model, the wall body confidence of each object is determined by the numerical size of the in-point rate of all structural surfaces in each object, the proportion of the number of points belonging to the structural surface in each object, and the fluctuation of the proportion of the number of points belonging to the structural surface in each object and the connected object of each object; Threshold segmentation is performed on the wall body confidence of all objects to obtain a construction wall body, the concave-convex coefficient of each point in each construction wall body is determined based on the fluctuation degree of the curvature of each point and its neighboring points in each construction wall body, and the local in-point rate of each point obtained by plane fitting of each point and its neighboring points in each construction wall body, to obtain a concave-convex point in each construction wall body; The concave-convex points in each construction wall body are removed to determine the reference surface of each construction wall body, the normal vector included angle between the reference surface of each construction wall body and the ground reference surface is obtained, and the actual height of the construction wall body is combined to obtain the verticality offset of each construction wall body, and the construction wall body is detected for inclination; The determination of the wall body confidence comprises:

2. The construction wall inclination detection method for construction work according to claim 1, wherein The mean value of the in-point rate of all structural surfaces in each object is calculated, denoted as a first mean value, the dispersion degree of the proportion of the number of points belonging to the structural surface in each object and the connected object of each object is calculated, denoted as a first dispersion degree, the product of the first mean value and the proportion is calculated, the sum of the first dispersion degree and a preset value greater than 0 is calculated, and the ratio of the product to the sum is determined as the wall body confidence. The determination of the ground surface structural surface comprises:

3. The construction wall inclination detection method for construction work according to claim 2, wherein The mean value and the dispersion degree of the z coordinate values of all points in each structural surface are calculated, and the structural surface with the smallest fusion result of the mean value and the dispersion degree is determined as the ground surface structural surface. The fitting of the ground reference surface comprises:

4. The construction wall inclination detection method for construction work according to claim 1, wherein The plane fitting algorithm is used to perform plane fitting on the x coordinate and y coordinate of all points in the ground surface structural surface, the mean value of the ground surface structural surface is taken as the unified height value of the fitted plane, and the ground reference surface is obtained.

5. The construction wall inclination detection method for construction work according to Claim 1, wherein The construction wall body is an object with a wall body confidence greater than or equal to a segmentation threshold. The determination of the concave-convex coefficient of each point in each construction wall body comprises:

6. The construction wall inclination detection method for construction work according to Claim 1, wherein The dispersion degree of the curvature of each point and its neighboring points in each construction wall body is calculated, denoted as a second dispersion degree, and the concave-convex coefficient of each point in each construction wall body is the ratio of the second dispersion degree to the local in-point rate.

7. The method for detecting inclination of a construction wall for building work according to Claim 6, wherein The concave-convex points in each construction wall body are points with a concave-convex coefficient greater than or equal to a segmentation threshold obtained by threshold segmentation on the concave-convex coefficient of all points in each construction wall body, and the reference surface of each construction wall body is obtained by point cloud filling on the construction wall body after removing the concave-convex points, and the two largest areas in the three-dimensional minimum circumscribed rectangle of the point cloud filled construction wall body are taken as the reference surface of the construction wall body. The acquisition process of the verticality offset comprises: The normal vector included angle degrees between the two reference surfaces of each construction wall body and the ground reference surface are calculated respectively, the absolute value of the difference between the mean value of the two included angle degrees of each construction wall body and 90° is calculated, and the verticality of each construction wall body is denoted as the verticality offset. The product of the sine function value of the verticality of each construction wall and the actual height of each construction wall is taken as the verticality offset of each construction wall.

8. A system for detecting the inclination of a construction wall for building works, comprising a memory, a processor and a computer program stored in said memory and running on said processor, characterized in that, The processor implements the steps of the method of any one of claims 1-7 when executing the computer program.

Images