Three-dimensional point cloud data polynomial fitting method, system and device and storage medium
The method addresses the challenge of simultaneous accuracy and continuity in 3D point cloud data fitting by employing grid partitioning and dynamic weighting, enhancing the precision and smoothness of surface representation.
Patent Information
- Application Number
- CN202411014086.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-07-26
AI Technical Summary
Existing 3D point cloud data polynomial fitting technologies struggle to simultaneously ensure accuracy, speed, and continuity, failing to meet the requirements of applications like surface reconstruction, CAD/CAM manufacturing, and remote sensing data analysis.
A method involving grid partitioning and dynamic weighting of polynomials based on inverse distance principles, using least squares methods to fit 3D point cloud data, ensuring accurate and continuous surface representation.
Enhances fitting accuracy by utilizing each data point effectively and maintaining continuity through weighted polynomial connections, improving surface reconstruction and data analysis in CAD/CAM and remote sensing.
Smart Images

Figure CN119091088A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of mathematics and computer image processing, and in particular, relates to a three-dimensional point cloud data polynomial fitting method, system, device and storage medium. Background Art
[0002] In the field of mathematics and computer image processing, 3D point cloud data fitting is a commonly used data analysis and image generation method. As one of the methods, 3D point cloud data polynomial fitting has broad application prospects.
[0003] First of all, polynomial fitting of 3D point cloud data is to find a polynomial function to approximate the surface defined by a set of 3D point cloud data. This polynomial function can be a quadratic, cubic or higher polynomial. By fitting discrete data points, a continuous surface model can be obtained, which can be used for data analysis, visualization, modeling and other aspects.
[0004] Secondly, the algorithm basis of polynomial fitting of 3D point cloud data is the least squares method. Based on the least squares method, the fitting polynomial function with the smallest sum of square errors is calculated to fit the surface data points as much as possible. In the fitting process, polynomial functions of different degrees can be selected, and better fitting effects can be achieved by appropriately adjusting parameters.
[0005] Finally, in practical applications, polynomial fitting of 3D point cloud data is widely used in the fields of surface reconstruction, CAD / CAM manufacturing, remote sensing data analysis, etc. For example, in surface reconstruction, we can fit a surface based on point cloud data to restore the shape of the original object. In CAD / CAM manufacturing, we can use 3D point cloud data polynomial fitting technology to generate part models, which is convenient for subsequent processing and manufacturing. In remote sensing analysis, we can extract the ground shape through 3D point cloud data fitting for use in geographic information systems, etc.
[0006] In general, 3D point cloud polynomial fitting is an important data analysis and image processing method with wide application value. When performing 3D point cloud polynomial fitting, the existing technology is difficult to simultaneously ensure the polynomial fitting accuracy, fitting speed and fitting continuity of 3D point cloud data, and cannot meet the accuracy requirements of various fields such as surface reconstruction, CAD / CAM manufacturing, and remote sensing data analysis. Summary of the invention
[0007] The technical problem to be solved by the present invention is to provide a three-dimensional point cloud data polynomial fitting method, system, device and storage medium, aiming to establish a polynomial that can accurately reflect the characteristics of three-dimensional point cloud data.
[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is: a three-dimensional point cloud data polynomial fitting method, comprising the following steps: Step 1: Determine the variables of the three-dimensional point cloud data: determine the independent variables and dependent variables of the three-dimensional point cloud data polynomial to be fitted respectively; Step 2: Determine the range of the independent variable of the three-dimensional point cloud data: project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the range of the independent variable; The fitting grid is started from the first grid close to the origin, and x Axis direction and y Get every other grid in the axis direction; Said x The mesh connection along the axis direction is x Axis direction, the grid in the middle of the two fitted grids; Said y The mesh connection along the axis direction is y Axis direction, the grid in the middle of the two fitted grids; The circumferentially connected grid is the grid at the center of the four fitting grids; Step 4: Determine the fitting range of the grid polynomial; Step 5: Calculate the fitted grid polynomial: according to the fitting range of the fitted grid polynomial, use the least square method to solve each fitted grid polynomial; Step 6: x Calculation of mesh polynomials in the axial direction: x The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. x The weight is determined according to the principle that the distance between the two fitting grids is inversely proportional to the axis direction. x The principle of inversely proportional distance in the axial direction: the greater the distance, the smaller the weight; Step 7: y Calculation of mesh polynomials in the axial direction: y The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. y The weight is determined according to the principle that the distance between the two fitting grids is inversely proportional to the axis direction. y The principle of inversely proportional distance in the axial direction, the greater the distance, the smaller the weight; Step 8: Calculation of circumferential mesh polynomial: The circumferential mesh polynomial is generated by dynamically weighting the four adjacent fitting mesh polynomials. The weights are calculated based on the distance from the four fitting meshes. x axis, y The weight is determined according to the principle of inversely proportional to the distance in the axial direction, and the weight is determined according to the distance of the four fitting grids. x axis,y The principle is inversely proportional to the distance in the axial direction. The larger the distance, the smaller the weight.
[0009] In a preferred embodiment, the number of independent variables is 2 and the number of dependent variables is 1.
[0010] In the preferred scheme, in step four, when the number of three-dimensional point cloud data falling into the fitting grid is less than the number of fitting polynomial variables, it is necessary to extend the fitting range based on the fitting grid; when the number of three-dimensional point cloud data falling into the fitting grid is greater than the number of fitting polynomial variables, the fitting grid range is directly selected as the fitting range.
[0011] In the preferred solution, in step 5, the least square method is used to solve the calculation expression of each fitting grid polynomial as follows: ; in, The coordinates are ( m , n )’s fitting grid polynomial corresponding to the fitting grid; B is the basis function vector; C is the coefficient vector; m for x Axis direction index; n for y Axis direction index.
[0012] In the preferred embodiment, in step 6, x The calculation formula of the axial connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get: ; Substituting the corresponding fitting grid polynomial into the above formula, we can get x The mesh polynomial in the axial direction is: ; in, The coordinates are ( m , n )of x Axis direction connection grid polynomial; x is the horizontal axis value; For coordinates The fitting grid polynomial corresponding to the fitting grid of ; In the preferred embodiment, in step seven, y The calculation formula of the axial connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get:
[0013] Substituting the corresponding fitting grid polynomial into the above formula, we can get y The mesh polynomial in the axial direction is:
[0014] in, for y The mesh polynomial in the axial direction; y is the vertical coordinate value; The coordinates are The fitting mesh polynomial corresponding to the fitting mesh of .
[0015] In a preferred solution, in step eight, the calculation formula of the circumferential connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get: ; Will y Substitute the axial direction fitting grid polynomial into the above formula to obtain: ; Substituting the fitting grid polynomial into the above formula, the circumferential connection grid polynomial is obtained as follows: ; in, is the circumferential mesh polynomial.
[0016] The present invention also provides a three-dimensional point cloud data polynomial fitting system, comprising: 3D point cloud data variable determination module: used to respectively determine the independent variables and dependent variables of the 3D point cloud data polynomial to be fitted; Independent variable range determination module: used to project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the independent variable range; Data grid partitioning module: It is used to partition the two-dimensional plane formed by the independent variable into grids according to the range and polynomial fitting accuracy to obtain the fitting grid, x Axis direction connection grid, y Axial connection grid and circumferential connection grid; A fitting grid polynomial fitting range determination module: used for determining the fitting grid polynomial fitting range; Polynomial calculation module: used to fit grid polynomials, x Axis direction connection grid polynomial, yCalculation of axial mesh polynomials and circumferential mesh polynomials.
[0017] The present invention also provides a three-dimensional point cloud data polynomial fitting device, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the above-mentioned three-dimensional point cloud data polynomial fitting method when executing the computer program.
[0018] The present invention also provides a storage medium, which can store a three-dimensional point cloud data polynomial fitting program, and when the program is read and executed by a computer, the above-mentioned three-dimensional point cloud data polynomial fitting method can be implemented.
[0019] The present invention provides a three-dimensional point cloud data polynomial fitting method, system, device and storage medium, which have the following beneficial effects: 1. In step three of the present invention, when dividing the grid, the number of data points falling within each fitting grid is taken into consideration, thereby fully utilizing each data point in the three-dimensional point cloud and maximizing the fitting accuracy of each fitting grid. By fitting the three-dimensional point cloud data in blocks, the local characteristics of the three-dimensional point cloud polynomial are ensured.
[0020] 2. In step 4 of the present invention, when determining the fitting range, the situation where the data points in the fitting grid are used to fit the abnormality is taken into consideration, and the data points around the fitting grid play an important supporting role in the fitting polynomial.
[0021] 3. In step 5 of the present invention, when fitting each fitting grid, the least square method is used to ensure the fitting accuracy of each fitting polynomial from the mathematical principle.
[0022] 4. Steps 6, 7 and 8 of the present invention use the inverse distance weighted method to perform weighted summation on the fitting polynomials of each part, thereby ensuring the overall continuity of the three-dimensional point cloud data polynomial and playing a role in smoothly connecting the fitting surface. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 is a flow chart of a polynomial fitting method for three-dimensional point cloud data provided by an embodiment of the present invention; Figure 2 is the independent variable range of the three-dimensional point cloud data provided by the embodiment of the present invention; Figure 3 is a schematic diagram of grid division in a polynomial fitting method for three-dimensional point cloud data provided by an embodiment of the present invention; Figure 4 This is a polynomial fitting result diagram of three-dimensional point cloud data provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0024] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0025] Combination Figure 1~Figure 4 The specific embodiments of the present invention are described in further detail.
[0026] like Figure 1 As shown, a polynomial fitting method for three-dimensional point cloud data includes the following steps: Step 1: Determine the variables of the three-dimensional point cloud data: determine the independent variables and dependent variables of the three-dimensional point cloud data polynomial to be fitted, respectively, the number of independent variables is 2, and the number of dependent variables is 1; Step 2: Determine the range of the independent variable of the three-dimensional point cloud data: project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the range of the independent variable; Step 3: Grid division of 3D point cloud data: Grid the 2D plane formed by the independent variable according to the range and polynomial fitting accuracy to obtain the fitting grid, x Axis direction connection grid, y Axial connection grid and circumferential connection grid; The fitting grid is started from the first grid close to the origin, and x Axis direction and y The data is obtained at every grid in the axial direction. When the required polynomial fitting accuracy is relatively high, the fitting grid can be divided as densely as possible, and the data points falling in each fitting grid are guaranteed to meet the data volume requirements of the least squares fitting. That is, when the fitting polynomial has n variables, n+1 data points are usually required for fitting. Said x The mesh is connected along the axis direction x Axis direction, the grid in the middle of the two fitted grids; Said y The mesh is connected along the axis direction y Axis direction, the grid in the middle of the two fitted grids; The circumferentially connected grid is the grid at the center of the four fitting grids; Step 4. Determine the fitting range of the fitting grid polynomial: By default, the fitting range of the fitting grid polynomial is the fitting grid range. However, when the fitting results of the data points in the fitting grid are abnormal, extend the boundary of the fitting grid outward until the fitting effect meets the requirements.
[0027] Step 4. Determine the fitting range of the fitting grid polynomial: When the number of 3D point cloud data falling into the fitting grid is less than the number of fitting polynomial variables, the fitting range needs to be extended on the basis of the fitting grid to meet the data volume requirements of the least squares fitting; when the number of 3D point cloud data falling into the fitting grid is greater than the number of fitting polynomial variables, directly select the fitting grid range as the fitting range.
[0028] Step 5: Calculate the fitted grid polynomial: according to the fitting range of the fitted grid polynomial, use the least square method to solve each fitted grid polynomial; Step 6: x Calculation of mesh polynomials in the axial direction: x The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. x It is determined based on the principle that the distance in the axial direction is inversely proportional to the distance in the axial direction. Step 7: y Calculation of mesh polynomials in the axial direction: y The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. y It is determined based on the principle that the distance in the axial direction is inversely proportional to the distance in the axial direction. Step 8: Calculation of circumferential mesh polynomial: The circumferential mesh polynomial is generated by dynamically weighting the four adjacent fitting mesh polynomials. The weights are calculated based on the distance from the four fitting meshes. x axis, y It is determined based on the principle that the distance in the axial direction is inversely proportional.
[0029] Example: This embodiment solves the polynomial fitting problem of turbine flow characteristics, and the specific operation steps are: (1) Determination of 3D point cloud data variables In the polynomial fitting problem of turbine flow characteristics, the independent variable is the unit speed and guide vane opening y , the dependent variable is the unit flow .
[0030] (2) Determination of the range of independent variables of 3D point cloud data Project the three-dimensional point cloud data of the turbine flow characteristics onto the two-dimensional plane formed by the unit speed and the guide vane opening, such as Figure 2 As shown, the independent variable The range of is [63.9,83.1], the independent variable y The range is [9.95,42.04].
[0031] (3) 3D point cloud data mesh division like Figure 3As shown, the independent variable The interval is divided into 3 equal parts, with coordinates of [63.9, 70.3, 76.7, 83.1], and the independent variable y The interval is divided into 7 equal parts, with coordinates of [9.95, 14.5, 19, 23.5, 28.5, 33, 37.5, 42.04], and the independent variable and y The coordinates of the two-dimensional plane are divided into 21 parts, and the grid coordinates are ( m , n ), , . m and n When both are odd numbers, the grid is a fitted grid. m is an even number, n The grid when is an odd number is x The mesh is connected in the axis direction. m is an odd number, n The grid when is an even number is y The mesh is connected in the axis direction. m and n When they are even numbers, the grids are circumferentially connected grids.
[0032] (4) Determination of the fitting range of the fitting grid polynomial Since the number of 3D point cloud data that falls into each fitting grid is small, select The fitting range is , y The fitting range is .
[0033] (5) Fitting grid polynomial calculation Select basis functions according to the fitting grid polynomial fitting range , the fitting grid polynomial coefficient matrix A1 is calculated using the least squares method as shown in Table 1.
[0034] Table 1 Fitting grid polynomial coefficient matrix A1
[0035] (6) x Axial mesh polynomial calculation According to the fitted grid polynomial, the inverse distance weighted method can be used to calculate x The coefficient matrices of the axial connection grid polynomials are divided into A2 and A3, as shown in Tables 2 and 3.
[0036] Table 2 x Coefficient matrix A2 of the axial mesh polynomial A2 A2{1,:} A2{:,1} [2527.32;-62.59;27.01;0.16;0.38;-0.07] A2{:,2} [3300.65;-75.68;4.12;1.07;0.33;-1.16] A2{:,3} [964.51;-6.95;0.32;0.82;-0.14;-0.56] A2{:,4} [-2419.60;33.26;106.79;-1.36;0.13;0.12] Table 3x Coefficient matrix A3 of the axial mesh polynomial A3 A3{1,:} A3{:,1} [-29.31;0.70;0.42;-0.01;0;0.01] A3{:,2} [-38.02;0.84;0.63;-0.01;0;0.010] A3{:,3} [-4.37;-0.11;0.56;-0.01;0;0] A3{:,4} [25.20;-0.44;-0.42;0.02;0;-0.01]
[0037] (7) y Axial mesh polynomial calculation According to the fitted grid polynomial, the inverse distance weighted method can be used to calculate y The coefficient matrices of the axial connection grid polynomials are divided into A4 and A5, as shown in Tables 4 and 5.
[0038] Table 4 y Coefficient matrix A4 of the axial mesh polynomial A4 A4{1,:} A4{2,:} A4{:,1} [717.89;-16.72;31.70;1.21;-0.03;-2.61] [1137.98;-25.10;23.34;1.44;-0.01;-2.78] A4{:,2} [1895.79;-40.02;-2.95;1.67;0;-1.79] [810.20;-7.60;-4.31;1.49;-0.21;-1.43] A4{:,3} [-8487.85;107.27;300.84;-0.99;-0.52;-2.85] [-10006.27;125.15;348.36;-2.33;-0.30;-2.20] Table 5 y Coefficient matrix A5 of the axial mesh polynomial A5 A5{1,:} A5{2,:} A5{:,1} [-22.94;0.39;1.42;-0.09;0.01;0.15] [-35.10;0.60;1.72;-0.10;0.01;0.16] A5{:,2} [-53.43;1.03;1.96;-0.06;0;0.05] [-6.45;-0.31;1.86;-0.05;0.01;0.03] A5{:,3} [243.41;-3.28;-6.87;0.03;0.01;0.06] [284.70;-3.74;-8.24;0.07;0.01;0.04]
[0039] (8) Circumferential mesh polynomial calculation according to x The axial connection grid polynomials, according to the inverse distance weighted method, the coefficient matrices of the circumferential connection grid polynomials are A6, A7, A8 and A9, as shown in Tables 6 to 9. The fitting results of the three-dimensional point cloud data are shown in Figure 4 .
[0040] Table 6 Coefficient matrix A6 of the circumferential mesh polynomial A6 A6{1,:} A6{:,1} [5752.45;-117.19;-68.44;3.96;0.18;-4.61] A6{:,2} [-11114.37;348.43;-19.34;-0.47;-2.58;2.57] A6{:,3} [-26685.23;321.53;870.28;-17.02;2.09;4.94] Table 7 Coefficient matrix A7 of circumferential mesh polynomial A7 A7{1,:} A7{:,1} [-65.64;1.31;1.31;-0.04;0;0.03] A7{:,2} [169.62;-5.06;0.21;0.03;0.03;-0.06] A7{:,3} [237.25;-2.79;-7.42;0.21;-0.03;-0.10] Table 8 Coefficient matrix A8 of circumferential mesh polynomial A8 A8{1,:} A8{:,1} [-168.69;2.86;4.99;-0.20;0.01;0.24] A8{:,2} [509.60;-14.99;0.83;0.05;0.10;-0.13] A8{:,3} [738.20;-8.77;-23.23;0.48;-0.06;-0.15] Table 9 Coefficient matrix A9 of circumferential mesh polynomial A9 A9{1,:} A9{:,1} [1.90;-0.03;-0.05;0;0;0] A9{:,2} [-7.34;0.21;0.01;0;0;0] A9{:,3} [-6.45;0.07;0.21;-0.01;0;0]
[0041] A three-dimensional point cloud data polynomial fitting system provided by an embodiment of the present invention comprises: a three-dimensional point cloud data variable determination module, a three-dimensional point cloud data independent variable range determination module, a three-dimensional point cloud data grid division module, a fitting grid polynomial fitting range determination module, a fitting grid polynomial calculation module, x Axis direction connection grid polynomial calculation module, y Axial connection mesh polynomial calculation module, circumferential connection mesh polynomial calculation module.
[0042] 3D point cloud data variable determination module: used to respectively determine the independent variables and dependent variables of the 3D point cloud data polynomial to be fitted, wherein there are 2 independent variables and 1 dependent variable.
[0043] Independent variable range determination module: used to project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the independent variable range.
[0044] Data grid partitioning module: It is used to partition the two-dimensional plane formed by the independent variable into grids according to the range and polynomial fitting accuracy to obtain the fitting grid, x Axis direction connection grid, y Axial connection grid and circumferential connection grid.
[0045] Fitting grid polynomial fitting range determination module: when the 3D point cloud data falls into the fitting grid less, the fitting range needs to be expanded on the basis of the fitting grid to ensure sufficient fitting data; when the 3D point cloud data falls into the fitting grid more, the fitting grid range can be directly selected as the fitting range.
[0046] Polynomial calculation module: used to fit grid polynomials, x Axis direction connection grid polynomial, y Calculation of axial mesh polynomials and circumferential mesh polynomials.
[0047] The present invention also provides a three-dimensional point cloud data polynomial fitting device, comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the above-mentioned three-dimensional point cloud data polynomial fitting method when executing the computer program.
[0048] The present invention also provides a storage medium, which can store a three-dimensional point cloud data polynomial fitting program, and when the program is read and executed by a computer, the above-mentioned three-dimensional point cloud data polynomial fitting method can be implemented.
[0049] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A polynomial fitting method for three-dimensional point cloud data, characterized in that: The following steps are involved: Step 1: Determine the variables of the three-dimensional point cloud data: determine the independent variables and dependent variables of the three-dimensional point cloud data polynomial to be fitted respectively; Step 2: Determine the range of the independent variable of the three-dimensional point cloud data: project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the range of the independent variable; Step 3: Grid division of 3D point cloud data: Grid the 2D plane formed by the independent variable to obtain the fitted grid, x Axis direction connection grid, y Axial connection grid and circumferential connection grid; The fitting grid is started from the first grid close to the origin, and x Axis direction and y Get every other grid in the axis direction; Said x The mesh connection along the axis direction is x Axis direction, the grid in the middle of the two fitted grids; Said y The mesh connection along the axis direction is y Axis direction, the grid in the middle of the two fitted grids; The circumferentially connected grid is the grid at the center of the four fitting grids; Step 4: Determine the fitting range of the grid polynomial; Step 5: Calculate the fitted grid polynomial: according to the fitting range of the fitted grid polynomial, use the least square method to solve each fitted grid polynomial; Step 6: x Calculation of mesh polynomials in the axial direction: x The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. x It is determined based on the principle that the distance in the axial direction is inversely proportional to the distance in the axial direction. Step 7: y Calculation of mesh polynomials in the axial direction: y The axial connection grid polynomial is generated by the dynamic weighting of two adjacent fitting grid polynomials, and the weight is based on the distance between the two fitting grids. y It is determined based on the principle that the distance in the axial direction is inversely proportional to the distance in the axial direction. Step 8: Calculation of circumferential mesh polynomial: The circumferential mesh polynomial is generated by dynamically weighting the four adjacent fitting mesh polynomials. The weights are calculated based on the distance from the four fitting meshes. x axis, y It is determined based on the principle that the distance in the axial direction is inversely proportional.
2. A three-dimensional point cloud data polynomial fitting method according to claim 1, characterized in that: The number of independent variables is 2, and the number of dependent variables is 1.
3. The polynomial fitting method for three-dimensional point cloud data according to claim 1, characterized in that: In step 4, when the number of three-dimensional point cloud data falling into the fitting grid is less than the number of fitting polynomial variables, the fitting range needs to be extended on the basis of the fitting grid; when the number of three-dimensional point cloud data falling into the fitting grid is greater than the number of fitting polynomial variables, the fitting grid range is directly selected as the fitting range.
4. The polynomial fitting method for three-dimensional point cloud data according to claim 1, characterized in that: In step 5, the least square method is used to solve the calculation expression of each fitting grid polynomial as follows: ; in, The coordinates are ( m , n )’s fitting grid polynomial corresponding to the fitting grid; B is the basis function vector; C is the coefficient vector; m for x Axis direction index; n for y Axis direction index.
5. The polynomial fitting method for three-dimensional point cloud data according to claim 1, characterized in that: In the step six, x The calculation formula of the axial connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get: ; Substituting the corresponding fitting grid polynomial into the above formula, we can get x The mesh polynomial in the axial direction is: ; in, The coordinates are ( m , n )of x Axis direction connection grid polynomial; x is the horizontal axis variable; and Respectively indicate the number and The grid abscissa of For coordinates The fitting mesh polynomial corresponding to the fitting mesh of .
6. The polynomial fitting method for three-dimensional point cloud data according to claim 1, characterized in that: In the step seven, y The calculation formula of the axial connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get: Substituting the corresponding fitting grid polynomial into the above formula, we can get y The mesh polynomial in the axial direction is: in, for y The mesh polynomial in the axial direction; y is the vertical axis variable; and Respectively indicate the number and The grid ordinate of The coordinates are The fitting mesh polynomial corresponding to the fitting mesh of .
7. The polynomial fitting method for three-dimensional point cloud data according to claim 1, characterized in that: In the step eight, the calculation formula of the circumferential connection grid polynomial is: ; After splitting the above formula and combining similar terms, we get: ; Will y Substitute the axial direction fitting grid polynomial into the above formula to obtain: ; Substituting the fitting grid polynomial into the above formula, the circumferential connection grid polynomial is obtained as follows: ; in, is the circumferential mesh polynomial.
8. A three-dimensional point cloud data polynomial fitting system, characterized in that: include: 3D point cloud data variable determination module: used to respectively determine the independent variables and dependent variables of the 3D point cloud data polynomial to be fitted; Independent variable range determination module: used to project the three-dimensional point cloud data to be fitted onto the two-dimensional plane formed by the independent variable, and connect the projection boundary points to obtain the independent variable range; Data grid partitioning module: It is used to partition the two-dimensional plane formed by the independent variable into grids according to the range and polynomial fitting accuracy to obtain the fitting grid, x Axis direction connection grid, y Axial connection grid and circumferential connection grid; A fitting grid polynomial fitting range determination module: used for determining the fitting grid polynomial fitting range; Polynomial calculation module: used to fit grid polynomials, x Axis direction connection grid polynomial, y Calculation of axial mesh polynomials and circumferential mesh polynomials.
9. A three-dimensional point cloud data polynomial fitting device, characterized in that: It comprises a memory and a processor; the memory is used to store a computer program; the processor is used to implement the three-dimensional point cloud data polynomial fitting method as described in any one of claims 1 to 7 when executing the computer program.
10. A storage medium, characterized in that: The storage medium can store a three-dimensional point cloud data polynomial fitting program, and when the program is read and executed by a computer, the three-dimensional point cloud data polynomial fitting method as described in any one of claims 1 to 7 can be implemented.
Citation Information
Patent Citations
Grid association-based quadtree index point cloud sorting method
CN106780748A
Rapid reconstruction method for laser radar point cloud data
CN108648277A
Positioning and navigation method for automatic inspection of unmanned aerial vehicle in water diversion pipeline of hydropower station
CN115290093A
Three-dimensional geographic point cloud adaptive slice thinning method and system
CN116129071A
Point cloud denoising systems and methods
US20200327646A1
