Wall surface flatness quantification system and method based on contour curvature features

Abstract

The present invention discloses a wall flatness quantification system and method based on contour curvature features, belonging to the technical field of wall flatness quantification. The method specifically includes: scanning the wall to be measured to obtain a wall point cloud data set, preprocessing the obtained wall point cloud data set, generating a surface fitting model of the wall based on the preprocessed wall point cloud data set, extracting the curvature of the wall contour, quantifying based on the curvature deviation of the wall flatness, defining the flatness index of the local part of the wall, generating a flatness distribution map of the wall based on the local flatness index of the wall, and outputting a flatness report. By extracting the curvature features of the wall contour, the flatness of each detection point is quantified and the overall flatness index of the wall is automatically calculated, which can be used for accurate quality assessment and greatly reduces the error of manual intervention.

Description

Technical Field

[0001] The present invention belongs to the technical field of wall flatness quantification, and specifically relates to a wall flatness quantification system and method based on contour curvature features. Background Art

[0002] During the process of building construction and interior decoration, wall flatness is an important indicator for evaluating construction quality. Traditional wall flatness detection methods are mostly based on tools such as straightedges, feeler gauges, laser levels, or vertical lines. These tools are mainly used to detect flatness deviations in large areas and rely on manual operation, so they are limited by operation accuracy, detection range, and human error. Common traditional methods include: 1) Straightedge and feeler gauge measurement: During construction, operators often use a long straightedge or feeler gauge to measure the wall flatness. The number of measurement points is small and manual reading records are required. This method can usually only detect large concave and convex areas, cannot capture minute uneven details on the wall surface, and is easily affected by manual operation deviations. 2) Laser level: The laser level measures the deviation of the wall through a horizontal laser line, and has a high detection efficiency for large wall areas, but it is difficult to capture small fluctuations on the wall surface. 3) Photogrammetry and image processing: Photogrammetry technology can generate the wall surface contour through images, but due to limited accuracy, minute unevenness is difficult to accurately present.

[0003] The main deficiencies of traditional methods are low accuracy, inability to capture minute unevenness, and cumbersome and time-consuming detection processes. Modern building construction requires more refined detection methods to ensure high-quality decoration projects. The flatness quantification method based on wall contour curvature can accurately extract minute uneven features of the wall and generate a quantitative flatness evaluation model through data processing. This method is more suitable for high-precision detection and can achieve automated processing over a large area, reducing human error. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention proposes a wall flatness quantification system and method based on contour curvature features. By analyzing the wall contour curvature, a flatness evaluation model is constructed to obtain flatness information in a quantitative manner.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A wall flatness quantification method based on contour curvature features, comprising:

[0007] Scanning the wall to be measured to obtain a wall point cloud data set;

[0008] Preprocessing the obtained wall point cloud data set;

[0009] Generate a surface fitting model of the wall based on the pre - processed wall point cloud dataset;

[0010] Extract the curvature of the wall contour, quantify it based on the curvature deviation of the wall flatness, and define the flatness index of the local wall;

[0011] Generate a flatness distribution map of the wall based on the local flatness index of the wall and output a flatness report.

[0012] Specifically, the generation of the surface fitting model of the wall based on the pre - processed wall point cloud dataset includes:

[0013] Set the pre - processed wall point cloud dataset as P, P = {p s (x s ,y s ,z s )|s = 1,2,...,N}, where p s (x s ,y s ,z s ) represents the spatial coordinates of the s - th sampling point on the wall, and N represents the total number of sampling points;

[0014] Generate a preliminary surface of the wall based on the pre - processed wall point cloud dataset P. The specific formula is:

[0015] ,

[0016] where, represents the preliminary surface of the wall, u and v represent the parameters in the parameter domain, Pk i,j represents the control points in the control point matrix, and the control point Pk i,j is represented as Pk i,j (x i,j ,y i,j ,z i,j ), represents the weight of the control point, N i,p (u) represents the B - spline basis function in the u - direction, N j,q (v) represents the B - spline basis function in the v - direction, m and n represent the number of control points, and p and q respectively represent the order of the basis function in the u and v directions;

[0017] Optimize the point cloud data of the wall and calculate the chord length distance between adjacent sampling points. The specific formula is:

[0018] ,

[0019] where, d s represents the chord length distance of the s - th adjacent sampling point, (x s+1 ,y s+1, z s+1 ) represents the spatial coordinates of the sampling point p s+1 ;

[0020] Calculate the cumulative chord length parameterization value of the sampling point. The specific formula is:

[0021] ,

[0022] where t s represents the cumulative chord length parameterization value of the sampling point, and d k represents the chord length distance of the k-th adjacent sampling point.

[0023] Specifically, generating a surface fitting model of the wall based on the preprocessed wall point cloud dataset further includes:

[0024] Select the number of control points Pk i,j according to the complexity of the wall and adjust the weights to minimize the preliminary surface fitting error of the wall. The fitting error specific formula is:

[0025] ,

[0026] where E represents the preliminary surface fitting error of the wall, represents the coordinates of the fitting surface point calculated according to the parameter , represents the sampling point P s and the Euclidean distance between the fitting surface point ;

[0027] Score the accuracy of the preliminary surface fitting of the wall and determine whether re-fitting is required based on the accuracy score of the preliminary surface fitting of the wall. The specific formula for the fitting accuracy score is:

[0028] ,

[0029] where Q represents the accuracy score of the preliminary surface fitting of the wall, represents the average error based on all sampling points, represents the standard deviation of the error, D represents the total dispersion of the preliminary surface of the wall, , and represent the weight coefficients;

[0030] Set the accuracy threshold Yz nh of the preliminary surface fitting of the wall. When Q ≥ Yz nh , it is determined that the accuracy of the preliminary surface fitting of the wall meets the standard. When Q ≤ Yz nhWhen it is determined that the accuracy of the preliminary surface fitting of the wall surface does not meet the standard, control points are added and weights are adjusted, and fitting is performed again until Q≥Yz nh to obtain the final surface of the wall surface .

[0031] Specifically, the curvature of the wall surface contour is extracted and quantified based on the curvature deviation of the wall surface flatness, and the flatness index of the local wall surface is defined, including:

[0032] In the surface of the wall surface , any two sampling points are randomly selected as the sampling points for curvature calculation, and the principal curvatures of the sampling points selected in the surface of the wall surface are calculated. The specific formula is:

[0033] ,

[0034] ,

[0035] where and represent the principal curvatures of two sampling points selected in the surface of the wall surface, S u , S v represents the first-order partial derivative of the surface of the wall surface , S uu , S uv and S vv represent the second-order partial derivatives of the surface of the wall surface ;

[0036] The sum of the absolute values of the principal curvatures is used to represent the total curvature of the two sampling points in the surface of the wall surface. The specific formula is:

[0037] , where represents the total curvature of the two sampling points in the surface of the wall surface, represents the absolute value function;

[0038] Based on the total curvature of the two sampling points in the surface of the wall surface, the local flatness of the wall surface is evaluated, and the local flatness index of the wall surface is calculated. The specific formula is:

[0039] ,

[0040] where Pzd represents the local flatness index of the wall surface, and l mean represents the average curvature of all sampling points of the wall surface.

[0041] Specifically, based on the local flatness index of the wall surface, a flatness distribution map of the wall surface is generated and a flatness report is output, including:

[0042] Based on the local flatness index Pzd of the wall surface, visually present the wall surface flatness using color gradients;

[0043] The generated wall surface flatness distribution map is presented in the form of a 3D model, combined with the 3D contour of the actual wall surface, and the 3D contour of the wall surface is enhanced for display;

[0044] Generate a detailed flatness numerical report.

[0045] Specifically, the preprocessing in the preprocessing of the obtained wall surface point cloud dataset includes:

[0046] Denoising and data registration;

[0047] For the denoising, use Gaussian filtering or bilateral filtering to perform noise reduction processing on the obtained wall surface point cloud dataset to remove measurement errors;

[0048] For the data registration, adopt multi-view scanning and use the ICP algorithm for data registration.

[0049] A wall surface flatness quantification system based on contour curvature features, used to implement the wall surface flatness quantification method based on contour curvature features, includes: a data acquisition module, a data preprocessing module, a surface fitting module, an index definition module, and a result output module;

[0050] The data acquisition module is used to scan the wall surface to be measured to obtain a wall surface point cloud dataset;

[0051] The data preprocessing module is used to preprocess the obtained wall surface point cloud dataset;

[0052] The surface fitting module is used to generate a surface fitting model of the wall surface based on the preprocessed wall surface point cloud dataset;

[0053] The index definition module is used to extract the curvature of the wall surface contour, quantify based on the curvature deviation of the wall surface flatness, and define the local flatness index of the wall surface;

[0054] The result output module is used to generate a flatness distribution map of the wall surface based on the local flatness index of the wall surface and output a flatness report.

[0055] Specifically, the surface fitting module includes: a preliminary fitting unit, a fitting error calculation unit, a fitting accuracy scoring unit, and a re-fitting unit;

[0056] The preliminary fitting unit is used to generate a preliminary surface of the wall surface based on the preprocessed wall surface point cloud dataset;

[0057] The fitting error calculation unit is used to optimize the point cloud data of the wall surface and calculate the error of the wall surface surface fitting;

[0058] The fitting accuracy scoring unit is used to score the optimized fitting accuracy;

[0059] The refitting unit is used to set the accuracy threshold for the preliminary surface fitting of the wall surface. When the fitting accuracy does not meet the standard, control points are added and weights are adjusted to perform refitting again.

[0060] Specifically, the index definition module includes: a principal curvature calculation unit, a curvature representation unit, and a flatness index definition unit;

[0061] The principal curvature calculation unit is used to select two sampling points in the surface of the wall surface to calculate the principal curvature;

[0062] The curvature representation unit is used to represent the total curvature of two sampling points in the surface of the wall surface by the sum of the absolute values of the principal curvatures;

[0063] The flatness index definition unit is used to evaluate the local flatness of the wall surface based on the total curvature of two sampling points in the surface of the wall surface, and calculate the local flatness index of the wall surface.

[0064] Compared with the prior art, the beneficial effects of the present invention are:

[0065] 1. The present invention proposes a method for quantifying the flatness of a wall surface based on contour curvature features, which can be applied to wall surfaces of various complex structures, whether it is a flat wall, an arc wall, a three-dimensional concave-convex wall, or an arc, a curved surface, an inclined surface, etc., solving the problem that traditional detection tools are not suitable for complex surfaces.

[0066] 2. The present invention proposes a method for quantifying the flatness of a wall surface based on contour curvature features. By extracting the curvature features of the wall surface contour, the flatness of each detection point is quantified and the overall flatness index of the wall surface is automatically calculated, which can be used for accurate quality assessment and greatly reduces the error of manual intervention.

[0067] 3. The present invention proposes a method for quantifying the flatness of a wall surface based on contour curvature features. The generated three-dimensional flatness distribution map uses color mapping, which is convenient for users to intuitively judge the flatness of different regions of the wall surface, and provides detailed numerical and curvature fluctuation analysis, accurately marking the regions with higher unevenness, indicating the specific position and deviation size. Description of the Drawings

[0068] Figure 1 It is a flowchart of the method for quantifying the flatness of a wall surface based on contour curvature features provided by the present invention;

[0069] Figure 2 It is an architecture diagram of the system for quantifying the flatness of a wall surface based on contour curvature features provided by the present invention. Detailed implementation manners

[0070] In order to make the technical means, creative features, achieved purposes and functions of the present invention easy to understand, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, so it cannot be understood as a limitation of the present invention. In addition, the terms "No. 1", "No. 2", "No. 3" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. The present invention will be further described below in conjunction with specific implementation manners.

[0071] Embodiment 1

[0072] Please refer to Figure 1 , an embodiment provided by the present invention: a method for quantifying the flatness of a wall surface based on contour curvature features, including the following specific steps:

[0073] Step S1: Use a high-precision 3D scanner or laser rangefinder to scan the wall surface to be measured to obtain a wall surface point cloud data set;

[0074] Step S2: Preprocess the obtained wall surface point cloud data set;

[0075] The preprocessing includes: denoising and data registration;

[0076] Denoising: Use Gaussian filtering or bilateral filtering to perform noise reduction processing on the obtained wall surface point cloud data set to remove measurement errors;

[0077] Data registration: Adopt multi-view scanning and use the ICP (Iterative Closest Point) algorithm for data registration to ensure the overall continuity of data points.

[0078] Step S3: Generate a surface fitting model of the wall surface based on the preprocessed wall surface point cloud data set;

[0079] The specific steps of Step S3 are as follows:

[0080] Step S301: Set the preprocessed wall surface point cloud data set as P, P = {p s (x s ,y s ,z s )|s = 1, 2,..., N}, where p s (x s ,y s ,z s ) represents the spatial coordinates of the s-th sampling point on the wall surface, and N represents the total number of sampling points;

[0081] Step S302: Based on the preprocessed wall point cloud dataset P, generate a preliminary surface of the wall. The specific formula is as follows:

[0082] ,

[0083] where, represents the preliminary surface of the wall, u and v represent the parameters in the parameter domain, usually taking values from 0 to 1, and Pk i,j represents the control point in the control point matrix. The control point Pk i,j is represented as Pk i,j (x i,j ,y i,j ,z i,j ), which determines the shape of the preliminary surface of the wall. represents the weight of the control point, allowing greater influence to be exerted at certain points to better fit the data. N i,p (u) represents the B-spline basis function in the u direction, and N j,q (v) represents the B-spline basis function in the v direction. m and n represent the number of control points, and p and q represent the orders of the basis functions in the u and v directions respectively;

[0084] N i,p (u) and N j,q (v) are both B-spline basis functions, and the B-spline basis functions are provided in the prior art;

[0085] Step S303: Optimize the point cloud data of the wall, and calculate the chord length distance between adjacent sampling points. The specific formula is as follows:

[0086] ,

[0087] where, d s represents the chord length distance of the s-th adjacent sampling point, and (x s+1 ,y s+1 ,z s+1 ) represents the spatial coordinates of the sampling point p s+1 ;

[0088] Step S304: Calculate the cumulative chord length parameterization value of the sampling point. The specific formula is as follows:

[0089] ,

[0090] where, t s represents the cumulative chord length parameterization value of the sampling point, and d k represents the chord length distance of the k-th adjacent sampling point;

[0091] Calculating the cumulative chord length parameterization value of the sampling point is achieved by calculating from the first point to ps The ratio of the cumulative distance to the total distance maps each sampling point to the parameter interval [0, 1] to ensure a smooth transition between different distances, which helps to maintain the uniformity of the surface;

[0092] Step S305: Select control points Pk i,j in number and adjust the weights to minimize the initial surface fitting error of the wall. The specific formula for the fitting error is:

[0093] ,

[0094] where E represents the initial surface fitting error of the wall, represents the coordinates of the fitting surface point calculated according to the parameter , that is, the point on the surface S corresponding to the sampling point P s , represents the sampling point P s and the fitting surface point is the Euclidean distance;

[0095] In this embodiment, E also represents the total deviation of all sampling points to the fitting surface. This value is used to evaluate the quality of the fitting surface. The smaller the value, the better the fitting effect;

[0096] The initial number of control points is (m + 1) × (n + 1). For relatively flat walls, fewer control points can be used, while for complex structures, the number of control points needs to be increased to ensure the fitting accuracy;

[0097] Step S306: Score the accuracy of the initial surface fitting of the wall and determine whether re - fitting is required based on the accuracy score of the initial surface fitting of the wall. The specific formula for the fitting accuracy score is:

[0098] ,

[0099] where Q represents the accuracy score of the initial surface fitting of the wall, represents the average error based on all sampling points, represents the standard deviation of the error, D represents the total dispersion of the initial surface of the wall, , and represent weight coefficients, which are adjusted according to measurement requirements to balance the overall error, discreteness, and local smoothness of the fitting accuracy;

[0100] In this embodiment, the formula for the average error based on all sampling points is: , E sDenote the initial surface fitting error of the wall based on the s-th sampling point. The average error reflects the overall deviation of the fitted surface relative to the sampled point cloud data and is used to measure the overall situation of the fitting accuracy; the standard deviation of the error The formula is as follows: , the standard deviation of the error represents the degree of dispersion of the fitting error. A high degree of dispersion indicates that the fitting quality in some areas is uneven, suggesting that the control points or weights need to be optimized;

[0101] The surface dispersion is used to measure the smoothness and consistency of the surface. The calculation of the dispersion introduces quadratic difference error analysis. By calculating the error change between adjacent sampling points, the smoothness of the surface is evaluated. The specific formula for the total dispersion D of the initial surface of the wall is: , M represents the number of all adjacent sampling point pairs, sl represents the adjacent point of sampling point s, Indicates that sampling point s and sl belong to the neighborhood, that is, sampling point s and sl are adjacent;

[0102] Step S307: Set the accuracy threshold Yz for the initial surface fitting of the wall nh , when Q ≥ Yz nh , it is determined that the accuracy of the initial surface fitting of the wall meets the standard. When Q ≤ Yz nh , it is determined that the accuracy of the initial surface fitting of the wall does not meet the standard. Increase the control points and adjust the weights, and refit until Q ≥ Yz nh , to obtain the final surface of the wall .

[0103] Step S4: Extract the curvature of the wall contour, quantify it based on the curvature deviation of the wall flatness, and define the flatness index of the local wall;

[0104] The specific steps of Step S4 are as follows:

[0105] Step S401: In the surface of the wall, arbitrarily select two sampling points as the sampling points for curvature calculation, and calculate the principal curvatures of the selected sampling points in the surface of the wall. The specific formula is:

[0106] ,

[0107] ,

[0108] Among them, and represent the principal curvatures of the two selected sampling points in the surface of the wall, S u , S v represents the first-order partial derivative of the surface of the wall, S uu , S uv and S vvRepresents the curved surface of the wall The second-order partial derivative of;

[0109] In this embodiment, S u Is the first-order partial derivative with respect to u, and is the curved surface of the wall The rate of change in the u direction, S v Is the first-order partial derivative with respect to v, and is the curved surface of the wall The rate of change in the v direction, S uu Is the second-order partial derivative with respect to u, and is the curved surface of the wall The degree of curvature of the curved surface in the u direction, S vv Is the second-order partial derivative with respect to v, and is the curved surface of the wall The degree of curvature of the curved surface in the v direction, S uv Is the second-order partial derivative with respect to u and v, and is the curved surface of the wall The mutual influence in the u and v directions;

[0110] Step S402: Use the sum of the absolute values of the principal curvatures to represent the total curvature of two sampling points in the curved surface of the wall. The specific formula is:

[0111] , where, Represents the total curvature of two sampling points in the curved surface of the wall, Represents the absolute value function;

[0112] Step S403: Based on the total curvature of two sampling points in the curved surface of the wall , evaluate the local flatness of the wall, and calculate the local flatness index of the wall. The specific formula is:

[0113] ,

[0114] Among them, Pzd represents the local flatness index of the wall, and l mean Represents the average curvature of all sampling points on the wall.

[0115] In this embodiment, the flatness index formula reflects the relative flatness of the overall wall by calculating the average of the deviation values between the total curvature of each sampling point and the total average curvature. The smaller the index Pzd, the flatter the wall.

[0116] Step S5: Generate a flatness distribution map of the wall based on the local flatness index of the wall, and output a flatness report.

[0117] The specific steps of Step S5 are:

[0118] Step S501: Based on the local flatness index Pzd of the wall, visually present the wall flatness using a color gradient;

[0119] Specifically, it includes: 1) Color mapping: converting the wall surface flatness information into color distribution to visually observe the flatness of different areas of the wall. Specifically, areas with higher flatness deviation (i.e., higher unevenness) are represented in red, with obvious surface undulations, while areas with lower flatness deviation (i.e., better flatness) are represented in green, with a smoother surface;

[0120] 2) Gradient transition: The color transitions from green to yellow, orange, and then to red, indicating the gradual change from flat areas to areas with larger unevenness. Through the detailed differentiation of the gradient colors, users can judge the subtle local changes of the wall surface and intuitively understand the flatness distribution state of the wall surface;

[0121] Step S502: The generated wall surface flatness distribution map is presented in the form of a 3D model, combined with the 3D contour of the actual wall surface, and the 3D contour of the wall surface is enhanced for display;

[0122] Specifically, it includes: 1) Viewpoint rotation: Users can freely rotate and zoom the model to observe the flatness of the wall surface from different angles, so as to conduct a comprehensive inspection of the wall surface;

[0123] 2) Highlighting of depth and curvature features: According to the flatness deviation of different areas, a dynamic highlighting effect is set, and visual cues of depth or protrusion are added in the red areas to more intuitively present the position and shape of the problem areas;

[0124] Step S503: Generate a detailed flatness numerical report.

[0125] Embodiment 2

[0126] Please refer to Figure 2 , another embodiment provided by the present invention: a wall surface flatness quantification system based on contour curvature features, including: a data acquisition module, a data preprocessing module, a surface fitting module, an index definition module, and a result output module;

[0127] The data acquisition module is used to scan the wall surface to be measured to obtain a wall surface point cloud data set;

[0128] The data preprocessing module is used to preprocess the obtained wall surface point cloud data set;

[0129] The surface fitting module is used to generate a surface fitting model of the wall surface based on the preprocessed wall surface point cloud data set;

[0130] The index definition module is used to extract the curvature of the wall surface contour, quantify it based on the curvature deviation of the wall surface flatness, and define the flatness index of the local area of the wall surface;

[0131] The result output module is used to generate a flatness distribution map of the wall surface based on the local flatness index of the wall surface and output a flatness report.

[0132] The surface fitting module includes: a preliminary fitting unit, a fitting error calculation unit, a fitting accuracy scoring unit, and a re-fitting unit;

[0133] The preliminary fitting unit is used to generate a preliminary surface of the wall surface based on the pre-processed wall surface point cloud data set;

[0134] The fitting error calculation unit is used to optimize the point cloud data of the wall surface and calculate the error of the wall surface fitting;

[0135] The fitting accuracy scoring unit is used to score the optimized fitting accuracy;

[0136] The re-fitting unit is used to set the accuracy threshold for the preliminary surface fitting of the wall surface. When the fitting accuracy does not meet the standard, control points are added and weights are adjusted to re-perform the fitting.

[0137] The index definition module includes: a principal curvature calculation unit, a curvature representation unit, and a flatness index definition unit;

[0138] The principal curvature calculation unit is used to calculate the principal curvature for two sampling points selected in the surface of the wall surface;

[0139] The curvature representation unit is used to represent the total curvature of two sampling points in the surface of the wall surface by the sum of the absolute values of the principal curvatures;

[0140] The flatness index definition unit is used to evaluate the local flatness of the wall surface based on the total curvature of two sampling points in the surface of the wall surface and calculate the local flatness index of the wall surface.

[0141] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are the same as the corresponding technical solutions in the prior art in terms of implementation principles are not described in detail to avoid excessive elaboration.

[0142] As described above in the specific embodiments, the purpose, technical solutions, and beneficial effects of the present invention have been further described in detail. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A wall surface flatness quantification method based on contour curvature features, characterized in that: include: Scan the wall to be measured to obtain a wall point cloud data set; Preprocess the acquired wall point cloud data set; Generate a surface fitting model of the wall based on the preprocessed wall point cloud dataset; Extract the curvature of the wall profile, quantify the curvature deviation of the wall flatness, and define the local flatness index of the wall; Based on the local flatness index of the wall, generate the flatness distribution map of the wall and output the flatness report; The method of generating a surface fitting model of the wall based on the preprocessed wall point cloud data set includes: Set the preprocessed wall point cloud dataset as P, P={p s (x s ,y s ,z s )|s=1,2,...,N}, where p s (x s ,y s ,z s ) represents the spatial coordinates of the sth sampling point on the wall, and N represents the total number of sampling points; Based on the preprocessed wall point cloud dataset P, the preliminary surface of the wall is generated. The specific formula is: ， in, represents the preliminary surface of the wall, u and v represent parameters in the parameter domain, Pk i,j Represents the control point in the control point matrix, control point Pk i,j Denoted as Pk i,j (x i,j ,y i,j ,z i,j ), represents the weight of the control point, N i,p (u) represents the B-spline basis function in the u direction, N j,q (v) represents the B-spline basis function in the v direction, m and n represent the number of control points, and p and q represent the order of the basis function in the u and v directions, respectively; The point cloud data of the wall is optimized and the chord length distance between adjacent sampling points is calculated. The specific formula is: ， Among them, d s represents the chord length distance between the sth adjacent sampling points, (x s+1 ,y s+1 ,z s+1 ) represents the sampling point p s+1 The spatial coordinates of Calculate the parameterized value of the cumulative chord length of the sampling point. The specific formula is: ， Among them, t s Represents the parameterized value of the cumulative chord length at the sampling point, d k Represents the chord length distance between the kth adjacent sampling points; Select control point Pk according to the complexity of the wall i,j The number of and adjust the weight To minimize the initial surface fitting error of the wall, the specific formula of the fitting error is: ， Where E represents the initial surface fitting error of the wall, Indicates that according to the parameters The calculated coordinates of the fitting surface points, Represents the sampling point P s and fitting surface points The Euclidean distance of The accuracy of the preliminary surface fitting of the wall is scored, and whether refitting is needed is determined based on the accuracy score of the preliminary surface fitting of the wall. The specific formula for the fitting accuracy score is: ， Where Q represents the accuracy score of the preliminary surface fitting of the wall, represents the average error based on all sampling points, represents the standard deviation of the error, D represents the total dispersion of the preliminary surface of the wall, , and represents the weight coefficient; Set the accuracy threshold Yz of the preliminary surface fitting of the wall nh , when Q ≥ Yz nh When Q≤Yz nh When the wall surface is judged to have a low accuracy, the control points are increased and the weights are adjusted, and the fitting is repeated until Q ≥ Yz nh , and get the final wall surface .

2. The wall surface flatness quantification method based on contour curvature features according to claim 1, characterized in that: The curvature of the wall contour is extracted, and the curvature deviation of the wall flatness is quantified to define the flatness index of the local wall surface, including: Curved surface on the wall In , two sampling points are randomly selected as the sampling points for curvature calculation, and the principal curvature of the selected sampling points in the wall surface is calculated. The specific formula is: ， ， in, and Represents the principal curvature of the two sampling points selected from the wall surface, S u , S v Surface representing the wall The first-order partial derivative of S uu , S uv and S vv Surface representing the wall The second-order partial derivative of ; The sum of the absolute values ​​of the principal curvatures is used to represent the sum of the curvatures of two sampling points on the wall surface. The specific formula is: ,in, Represents the sum of the curvatures of the two sampling points in the wall surface, It represents the absolute value function; The sum of the curvatures of two sampling points in the wall-based surface , evaluate the local flatness of the wall and calculate the local flatness index of the wall. The specific formula is: ， Among them, Pzd represents the local flatness index of the wall, l mean Represents the average curvature of all sampling points on the wall.

3. The wall surface flatness quantification method based on contour curvature features according to claim 1, characterized in that: The method generates a flatness distribution map of the wall based on the local flatness index of the wall, and outputs a flatness report, including: Based on the local flatness index Pzd of the wall, the wall flatness is intuitively presented using color gradient; The generated wall flatness distribution map is presented in the form of a 3D model, combined with the actual 3D contour of the wall, and the 3D contour of the wall is enhanced and displayed; Generates detailed flatness value reports.

4. The wall surface flatness quantification method based on contour curvature features according to claim 1, characterized in that: The preprocessing of the acquired wall point cloud data set includes: denoising and data registration; the denoising uses Gaussian filtering or bilateral filtering to perform noise reduction processing on the acquired wall point cloud data set to remove measurement errors; The data registration adopts multi-view scanning and uses the ICP algorithm to perform data registration.

5. A wall surface flatness quantification system based on contour curvature features, used to implement the wall surface flatness quantification method based on contour curvature features as claimed in any one of claims 1 to 4, characterized in that: include: Data acquisition module, data preprocessing module, surface fitting module, indicator definition module and result output module; The data acquisition module is used to scan the wall surface to be measured and obtain a wall surface point cloud data set; The data preprocessing module is used to preprocess the acquired wall point cloud data set; The surface fitting module is used to generate a surface fitting model of the wall based on the preprocessed wall point cloud data set; The index definition module is used to extract the curvature of the wall profile, quantify the curvature deviation of the wall flatness, and define the flatness index of the local wall surface; The result output module is used to generate a flatness distribution map of the wall surface based on the local flatness index of the wall surface, and output a flatness report.

6. The wall surface flatness quantification system based on profile curvature characteristics according to claim 5, characterized in that: The surface fitting module includes: a preliminary fitting unit, a fitting error calculation unit, a fitting accuracy scoring unit and a refitting unit; The preliminary fitting unit is used to generate a preliminary curved surface of the wall based on the preprocessed wall point cloud data set; The fitting error calculation unit is used to optimize the point cloud data of the wall surface and calculate the error of the wall surface fitting; The fitting accuracy scoring unit is used to score the optimized fitting accuracy; The refitting unit is used to set a precision threshold for the preliminary surface fitting of the wall surface. When the fitting precision does not meet the standard, control points are added and weights are adjusted to perform the fitting again.

7. The wall surface flatness quantification system based on profile curvature characteristics according to claim 6, characterized in that: The index definition module includes: a principal curvature calculation unit, a curvature representation unit, and a flatness index definition unit; The principal curvature calculation unit is used to calculate the principal curvature at two sampling points selected from the curved surface of the wall; The curvature representation unit is used to represent the sum of curvatures of two sampling points in the curved surface of the wall by using the sum of the absolute values ​​of the principal curvatures; The flatness index definition unit is used to evaluate the local flatness of the wall based on the sum of curvatures of two sampling points in the curved surface of the wall, and calculate the local flatness index of the wall.

Images