Gear parameter measurement method based on point cloud technology

Through the gear parameter measurement method based on point cloud technology, the problems of low degree of automation and lack of accurate measurement in the existing technology are solved, efficient and accurate measurement of gear parameters are achieved, suitable for complex-shaped gears, and the impact of human error is reduced.

CN119090797APending Publication Date: 2024-12-06CHANGCHUN UNIV OF TECH

Patent Information

Application Number
CN202410742931.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-11
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing two-dimensional gear image measurement methods are low in automation and lack accurate three-dimensional point cloud measurement methods for gears, resulting in low measurement efficiency, susceptible to human errors, and the inability to accurately obtain the depth information of complex-shaped gears.

Method used

The gear parameter measurement method based on point cloud technology is adopted to achieve accurate measurement of gear parameters through point cloud preprocessing, local public area splicing, gear point cloud inner hole axis extraction and establishment of mathematical models. Specific steps include Gaussian filtering to remove outliers, point cloud splicing, inner hole axis extraction and parameter calculation, and optimize point cloud data processing using PCA and RANSAC methods.

Benefits of technology

It improves the accuracy and efficiency of gear parameter measurement, adapts to gear measurement in complex shapes, reduces human error, provides more comprehensive data support, and provides technical support for industrial automation and intelligent manufacturing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure SMS_1
    Figure SMS_1
  • Figure SMS_3
    Figure SMS_3
  • Figure SMS_4
    Figure SMS_4
Patent Text Reader

Abstract

The invention discloses a gear parameter measurement method based on a point cloud technology, and belongs to the technical field of three-dimensional point cloud measurement. Firstly, local point cloud data of a gear is acquired by using a three-dimensional scanner, and the data is preprocessed and corrected; and then, finding a corresponding distance by using a one-to-one correspondence relationship of the feature points of the public area, and determining corresponding points, thereby splicing to form a measurable point cloud picture. Thirdly, extracting the central axis of the gear by utilizing the incomplete point cloud data of the inner hole wall of the gear; and finally, the contour of the gear is extracted, a measurement model is established, and corresponding basic parameters are calculated. According to the method, the point cloud data of the gear can be automatically extracted, matching of wrong points is avoided, and the size parameters of the gear are conveniently and accurately measured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of three-dimensional point cloud measurement, and in particular to a gear parameter measurement method based on point cloud technology. Background Art

[0002] In industrial production, in order to ensure the comprehensive quality of products, the inspection link of output parts is very important. The measurement methods of parts are divided into contact measurement and non-contact measurement. Contact measurement is divided into manual measurement (operating plug gauges or inner hole detection equipment) and measuring instrument measurement. This type of contact measurement method is easy to damage the hole wall and has low efficiency; and the common non-contact measurement method is mainly based on the two-dimensional image processing technology of machine vision theory, but the two-dimensional image measurement method has poor anti-interference ability and is limited to measuring plane information; with the deepening of my country's industrial intelligence, the increasing maturity of laser scanning technology, relying on the good real-time and stability of three-dimensional sensors and the advantages of three-dimensional data containing depth information, the three-dimensional point cloud measurement method can accurately describe the geometric characteristics and positional relationships of the three-dimensional model, and has unique advantages in the structural size measurement of parts.

[0003] Gears are important transmission parts widely used in modern mechanical equipment, and their geometric accuracy directly affects the working performance and service life of mechanical equipment. The gear tooth shape is usually an involute surface with complex geometric shapes and multiple parameters. Therefore, the measurement of gear parameters has always been a research hotspot and difficulty in the field of measurement technology. At present, most gear parameter measurements still use dedicated precision measuring instruments and use contact mechanical measurement methods for measurement and detection. However, these methods have problems such as complex operation, single parameter measurement, large workload, long time consumption and low measurement efficiency. At the same time, these methods are easily affected by subjective factors of personnel, resulting in human errors. In addition, the 2D image method can only provide the projection information of the gear surface on the plane, and cannot obtain the depth information of the gear. For gears with complex shapes or uneven surfaces, the 2D image method may not be able to accurately measure their shape and size; structured light measurement The structured light measurement method measures the surface shape of the gear by projecting special structured light stripes. However, due to reasons such as diffraction and interference of light, the structured light stripes may be deformed, distorted or blurred, resulting in a decrease in the accuracy of the measurement results. Therefore, it is of great significance to study the three-dimensional gear measurement method based on the point cloud mode to reduce these effects.

[0004] The gear measurement method based on 3D point cloud can realize the accurate measurement of gear parameters through the steps of point cloud processing, feature extraction and parameter calculation. This method not only has the advantages of high precision and high efficiency, but also can adapt to various complex gear shapes and structures. In addition, the gear measurement method based on 3D point cloud can also realize the 3D reconstruction and shape analysis of gears, providing more comprehensive data support for the design, manufacturing and performance evaluation of gears. Therefore, it is of great theoretical and practical significance to study the gear measurement method based on 3D point cloud. By adopting advanced point cloud processing and computational analysis technology, accurate measurement and evaluation of gear parameters can be realized, providing technical support for improving the efficiency and reliability of gear transmission. With the advancement of industrial automation and intelligent manufacturing, the gear measurement method based on 3D point cloud will play a more important role in industrial production. Summary of the invention

[0005] The purpose of the present invention is to provide a gear parameter measurement method based on point cloud technology to address the problems of low automation in existing gear two-dimensional image measurement methods and lack of accurate gear three-dimensional point cloud measurement methods. The method can extract point cloud data of gears, and focuses on research in detailed areas such as local common area splicing, outer contour extraction, and parameter measurement by establishing mathematical models. The relevant parameters of the gear can be conveniently and accurately measured based on the established gear measurement model.

[0006] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0007] Step 1, preprocessing the original point cloud data;

[0008] In step 1, the point cloud is first collected and input, and then the outliers in the original point cloud are removed using Gaussian filtering to obtain the preprocessed point cloud set, which is denoted as Q1; Q2.

[0009] Step 2, local common area stitching; stitch the preprocessed point cloud data to generate a local gear point cloud; first obtain the covariance of the two point clouds, calculate the eigenvector corresponding to the eigenvalue, and use the eigenvector corresponding to the maximum eigenvalue as the coordinate axis to establish a three-dimensional coordinate system in the two point clouds. The two coordinate systems are adjusted to align the common areas of the two point clouds, and then the feature points of the common areas are determined from the two point clouds Q1 and Q2. The feature points of the common areas in the two aligned point clouds correspond one to one and the corresponding Euclidean distances are equal, so point cloud feature matching can be used to achieve stitching.

[0010] In step 2, a three-dimensional coordinate system is established in the two point clouds, and the two coordinate systems are adjusted so that the common areas of the two point clouds are aligned.

[0011] In the above formulas, Q is a point cloud dataset in the format of a 3×N dimensional matrix, C is the covariance matrix of the Q matrix, V is a matrix composed of eigenvectors v1, v2, and v3, and D is a diagonal matrix composed of eigenvalues ​​d1, d2, and d3. The eigenvalues ​​are sorted from large to small, and the center point is taken as the origin. The eigenvectors corresponding to the first three eigenvalues ​​are taken as the coordinate axes in the three directions to establish a plane; then the two coordinate origins are moved to the same position, and R is the coordinate system rotation matrix. Respectively represent the angles between the three coordinate axes, t i To correspond to the translation distance, two coordinate systems are then obtained that correspond to each other after being aligned.

[0012] In step 2, feature points in the common area of ​​the two point clouds are extracted, and the corresponding Euclidean distance is equal to achieve stitching.

[0013] In the above formula, D is the Euclidean distance between the corresponding points in the two common point clouds, V is the distance between all corresponding feature points put into a matrix, the rows and columns in the matrix correspond to the distances between the feature points in the two point clouds, and the voting principle is used to find the distance d with the largest number of appearances in each row in the matrix V, and extract its corresponding row and column, which are the corresponding points with the correct relationship; finally, the optimal solution of the rotation matrix R and the translation vector T is used on the point cloud Q1, so that Q1Q2 is spliced.

[0014] Step 3: Extract the inner hole axis of the gear point cloud.

[0015] Equation F1 is established for the central circular hole on the gear point cloud surface, and it is used as the complete inner circular hole of the gear to obtain the radius r and center coordinates (a, b) of the circle. Each point is brought into the equation to determine whether it meets the equation F1. Starting from the selected seed point, its neighboring points are found and it is determined whether these neighboring points meet the characteristic function of the target surface. If they meet the characteristic function, they are added to the current cluster block and marked as new seed points. Then this process is repeated, and the outward search is performed until all points that meet the specific feature type and growth parameters are clustered to obtain the gear inner hole point cloud set Q. Since only part of the point cloud on one side of the gear inner hole can be photographed, it is impossible to directly use it to fit the circular hole to obtain the center of the circle. Therefore, the gear inner hole point cloud Q is first layered according to the Z axis. Due to the characteristics of cylindrical spur gears, the radius r of each layer of the gear circular hole is the same. Therefore, the radius r obtained by the complete circle of the tooth surface in function F1 is used as the mark radius. For each layer of the circular hole, any two points are selected to establish the characteristic function F2 of the circle. The center coordinate O (a i ,b i ), record the coordinate set E of all center points; select a point O(a i ,b i) and the center point N(a, b) of the gear end face use the straight line two-point method to construct the axis L, n is the direction vector, calculate the angle θ between the axis L and the normal vector of the gear end face, set the threshold T2, when the vector angle θ is less than the set threshold, the axis is considered to be the gear axis, if it is greater, re-select two points and repeat the iteration until the condition is met

[0016] In the above formulas, F1 is the gear end face inner hole equation established to determine the inner hole radius r and the end face center point (a, b), Q is the set of gear inner hole points, t z is the layering accuracy, F2 is the equation established by the inner hole points of each layer after the inner hole is layered, O(ai,bi) is the center of the circle equation established in each layer, E is the set of center points, L is the fitting line n(u,v,w) is the direction vector, θ is the angle between the two lines, one of which is the line L constructed by the center point of one layer and the center point of the gear end face, and the other is a line with the normal vector of the gear end face as the direction.

[0017] Step 4: Establish a gear measurement model and calculate relevant parameters.

[0018] First, the boundary contour of the gear point cloud is extracted, and then plane F1 is established in the direction of the axis L to project the gear point cloud contour onto a gear point cloud, and a rotation plane F2 is established with the axis L as the rotation axis to project the gear point cloud onto the plane to form a rectangular point cloud. Curve fitting is performed on the gear contour point cloud data on the F1 plane to establish a polar coordinate equation for the angle θ and distance H from the center point to the gear boundary. The fitted boundary points are substituted into the polar coordinate equation to calculate the unknown parameters. From the known trigonometric function equation, the maximum value Ra is used as the predetermined tooth top circle radius value and the minimum value R is used as the minimum value. f As the predetermined tooth root circle radius, calculate the angles θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number Establish a rotation plane F2 around the gear with the axis as the edge, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i =Ra±w, then it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter.

[0019] In the above formula, H is the distance from the center point on the F1 plane to the gear point cloud contour, θ is the corresponding angle, N and M are the parameters to be determined, is the angle difference between two adjacent maximum values ​​in the trigonometric function, B is the tooth width, Da is the tooth top circle diameter, b i is the length of the rectangle width on all planes F2, hi is the length of the rectangle on all planes F2; ​​m is the module, z is the number of teeth, The normal tooth system is 1 for the tooth addendum coefficient, s for the tooth thickness, P for the tooth pitch, and d for the pitch circle diameter.

[0020] Furthermore, the splicing method in step 2 is specifically as follows:

[0021] Step 2-1, use the KD tree to establish the topological relationship of the point cloud, calculate the covariance matrix C to find its eigenvalues ​​and sort them, extract the first three principal components of the two point clouds as coordinate axes, and establish the rotation angle θ through the corresponding coordinate axis angle. Straighten it to ensure that the Euclidean distances of the corresponding feature points in the two common areas correspond one to one.

[0022] Step 2-2, calculate the normal vector n(a, b, c) and curvature S of each point, set the curvature threshold T1, and find the corresponding feature points of the two point clouds. Establish matrix U, with the corresponding feature points selected by point cloud Q1 as the rows of matrix U, and the corresponding feature points selected by point cloud as the columns of matrix U. Matrix U corresponds to the distance between feature points Q1 and Q2. Use the voting principle to find the value with the most distance occurrences in matrix U, and its row and column are the correct correspondence relationship between feature points Q1 and Q2.

[0023] Furthermore, in step 3, in view of the fact that the inner hole of the gear cannot be completely photographed, a method for extracting the gear axis based on the residual defect cloud of the arc is proposed. The specific contents are as follows:

[0024] Step 3-1, after the boundary extraction in the previous step, use point cloud segmentation to separate the point cloud of the gear end surface, select any three points to establish the equation of the circle, calculate multiple groups of circle centers and radii multiple times, and take the average as the standard parameters of the gear inner hole.

[0025] Step 3-2: For the interior that cannot be photographed, divide the Z axis and sort the points in the point cloud according to the Z axis coordinate size, find the maximum value Zmax and the minimum value Zmin, and calculate the layering accuracy T z , establish the plane according to the layering accuracy, and classify the points with the same Z-axis coordinate value into the same plane point set.

[0026] Step 3-3, perform circle fitting on each layer, take the radius of the circular hole on the end face as a parameter, select any two points to establish the function of the circle of the layer, and calculate the center of the layer.

[0027] Step 3-4, collect the center points of each layer into a set and construct the axis with the center point of the end face using the two-point method, and directly calculate the angle between the axis and the normal vector direction of the gear end face. If the angle is less than the set threshold, it can be used as the gear axis. Otherwise, take another point from the center point set to recalculate the axis and angle for iteration.

[0028] Furthermore, in step 4, the specific gear size measurement method is as follows:

[0029] Step 4-1, establish a plane in the direction of the gear axis, project the gear end face point cloud, establish the polar coordinate equation for the angle θ and distance H from the center point to the gear boundary, bring the fitted boundary points into the polar coordinate equation to calculate the unknown parameters, and use the maximum value Ra as the predetermined tooth top circle radius and the minimum value R as the trigonometric function equation. f As the predetermined tooth root circle radius, calculate the angles θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number

[0030] Step 4-2, use the axis as the rotation axis to establish a plane, project a rectangular point cloud, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i= Ra±w, it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter.

[0031] Compared with the prior art, the present invention has the following advantages: 1. In view of the inability to collect complete gear data, it is necessary to integrate and splice gear data from different perspectives. A new registration method is proposed to reduce the matching between incorrect corresponding points, shorten the pairing time, and improve the pairing accuracy. 2. Aiming at the situation that the inner hole of the gear cannot be completely photographed and its axis is difficult to find, a method of extracting the gear axis based on the residual defect cloud of the arc is proposed. 3. In view of the spatiality and discreteness of point cloud data, direct gear measurement will result in large errors. A gear measurement model based on three-dimensional point cloud is proposed, and gear parameter measurement is performed. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is the overall flow chart of the method of the present invention.

[0033] Figure 2 It is a point cloud stitching flow chart of the method of the present invention.

[0034] Figure 3 This is a flow chart of point cloud axis extraction of the method of the present invention.

[0035] Figure 4 It is a schematic diagram of obtaining basic parameters of gear point cloud according to the method of the present invention.

[0036] Figure 5 It is a schematic diagram of residual defect cloud splicing of the method of the present invention.

[0037] Figure 6 It is a schematic diagram of layered slicing of the method of the present invention.

[0038] Figure 7 It is a schematic diagram of the gear profile after projection in the axial direction according to the method of the present invention.

[0039] Figure 8 It is a schematic diagram of a rectangular outline after projection with the axis as the rotation axis of the method of the present invention. DETAILED DESCRIPTION

[0040] The present invention is further described in detail below with reference to the accompanying drawings.

[0041] like Figure 1 As shown, the method of the present invention comprises the following steps:

[0042] Step 1, preprocessing the original point cloud data;

[0043] Step 2, stitching the point cloud;

[0044] Step 3, extracting the inner hole axis of the gear point cloud;

[0045] Step 4: Establish a gear measurement model and calculate relevant parameters.

[0046] In step 1, the point cloud is first collected and input, and then the outliers in the original point cloud are removed using Gaussian filtering to obtain the preprocessed point cloud set, which is denoted as Q1; Q2;

[0047] Step 2: local public area splicing Figure 2 As shown: the pre-processed point cloud data is spliced ​​to generate a complete gear model. First, the PCA principle is used to extract the first three eigenvalues ​​as the corresponding XYZ axes, and a three-dimensional coordinate system is established in the two point clouds. The two coordinate systems are adjusted to align the common areas of the two point clouds. Then, the feature points of the common areas are determined from the two point clouds Q1 and Q2. Based on the principle that the feature points of the common areas of the two point clouds after alignment correspond one to one and the corresponding Euclidean distances are equal, point cloud feature matching can be used to achieve splicing.

[0048] Step 2-1 uses the PCA principle to establish a three-dimensional coordinate system in the two point clouds, and adjusts the two coordinate systems to align the common areas of the two point clouds.

[0049] In the above formulas, Q is a point cloud dataset in the format of a 3×N dimensional matrix, C is the covariance matrix of the Q matrix, V is a matrix composed of eigenvectors v1, v2, and v3, and D is a diagonal matrix composed of eigenvalues ​​d1, d2, and d3. The eigenvalues ​​are sorted from large to small, and the center point is taken as the origin. The eigenvectors corresponding to the first three eigenvalues ​​are taken as the coordinate axes in the three directions to establish a plane; then the two coordinate origins are moved to the same position, and R is the coordinate system rotation matrix. Respectively represent the angles between the three coordinate axes, t i To correspond to the translation distance, two coordinate systems are then obtained that correspond to each other after being aligned.

[0050] Step 2-2 extracts the feature points in the common area of ​​the two point clouds, and uses the corresponding relationship and the equal Euclidean distance to achieve stitching.

[0051] In the above formula, D is the Euclidean distance between the corresponding points in the two common point clouds, V is the distance between all corresponding feature points put into a matrix, the rows and columns in the matrix correspond to the distances between the feature points in the two point clouds, and the voting principle is used to find the distance d with the largest number of appearances in each row in the matrix V, and extract its corresponding row and column, which are the corresponding points with the correct relationship; finally, the optimal solution of the rotation matrix R and the translation vector T is used on the point cloud Q1, so that Q1Q2 is spliced.

[0052] Step 3: Extract the inner axis of the gear point cloud. Figure 3 As shown:

[0053] Step 3, establish equation F1 for the central circular hole on the gear point cloud surface, take it as the complete inner circular hole of the gear, and obtain the radius r and center coordinates (a, b) of the circle. Substitute each point into equation F1 to determine whether it meets the requirement. Starting from the selected seed point, find its neighboring points and determine whether these neighboring points meet the characteristic function of the target surface. If they meet the requirement, add them to the current cluster block and mark them as new seed points. Then repeat this process, search outward until all points that meet the specific feature type and growth parameters are clustered, and obtain the gear inner hole point cloud set Q. Since only part of the point cloud on one side of the gear inner hole can be photographed, it is impossible to directly use it to fit the circular hole to obtain the center of the circle. Therefore, the gear inner hole point cloud Q is first layered according to the Z axis. Due to the characteristics of cylindrical spur gears, the radius r of each layer of the gear circular hole is the same. Therefore, the radius r obtained by the complete circle of the tooth surface in function F1 is used as the mark radius. For each layer of the circular hole, any two points are selected to establish the characteristic function F2 of the circle. The center coordinate O (a) of the circle of this layer can be obtained by using the incomplete arc of this layer. i ,b i ), record the coordinate set E of all center points; select a point O (a i ,b i ) and the center point N(a, b) of the gear end face use the straight line two-point method to construct the axis L, n is the direction vector, calculate the angle θ between the axis L and the normal vector of the gear end face, set the threshold T2, when the vector angle θ is less than the set threshold, the axis is considered to be the gear axis, if it is greater, re-select any two points and repeat the iteration until the conditions are met.

[0054] In the above formulas, F1 is the gear end face inner hole equation established to determine the inner hole radius r and the end face center point (a, b), Q is the set of gear inner hole points, t z is the layering accuracy, F2 is the equation established by the inner hole points of each layer after the inner hole is layered, O(ai,bi) is the center of the circle equation established in each layer, E is the set of center points, L is the fitting line n(u,v,w) is the direction vector, θ is the angle between the two lines, one of which is the line L constructed by the center point of one layer and the center point of the gear end face, and the other is a line with the normal vector of the gear end face as the direction.

[0055] Step 3-1, after the boundary extraction in the previous step, use point cloud segmentation to separate the point cloud of the gear end surface, select any three points to establish the equation of the circle, calculate multiple groups of circle centers and radii multiple times, and take the average as the standard parameters of the gear inner hole.

[0056] Step 3-2: For the interior that cannot be photographed, divide the Z axis and sort the points in the point cloud according to the Z axis coordinate size, find the maximum value Zmax and the minimum value Zmin, and calculate the layering accuracy T z, establish the plane according to the layering accuracy, and classify the points with the same Z-axis coordinate value into the same plane point set.

[0057] Step 3-3, perform circle fitting on each layer, take the radius of the circular hole on the end face as a parameter, select any two points to establish the function of the circle of the layer, and calculate the center of the layer.

[0058] Step 4: Establish a gear measurement model and calculate relevant parameters such as Figure 4 As shown:

[0059] Step 4, first extract the boundary contour of the gear point cloud, then establish plane F1 in the direction of axis L to project the gear point cloud contour onto it to form a gear point cloud, and establish rotation plane F2 with axis L as the rotation axis to project the gear point cloud onto the plane to form a rectangular point cloud; perform curve fitting on the gear contour point cloud data on plane F1, establish polar coordinate equations for the angle θ and distance H from the center point to the gear boundary, bring the fitted boundary points into the polar coordinate equation to calculate the unknown parameters, and use the maximum value Ra as the predetermined tooth top circle radius value and the minimum value R as the minimum value from the trigonometric function equation. f As the predetermined tooth root circle radius, calculate the angles θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number Establish a rotation plane F2 around the gear with the axis as the edge, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i= Ra±w, it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter.

[0060] In the above formula, H is the distance from the center point on the F1 plane to the gear point cloud contour, θ is the corresponding angle, N and M are the parameters to be determined, is the angle difference between two adjacent maximum values ​​in the trigonometric function, B is the tooth width, Da is the diameter of the tooth tip circle, bi is the length of the rectangle width on all planes F2, hi is the length of the rectangle on all planes F2; ​​m is the module, z is the number of teeth, The normal tooth system is 1 for the tooth addendum coefficient, s for the tooth thickness, P for the tooth pitch, and d for the pitch circle diameter.

[0061] Step 4-1, establish a plane in the direction of the gear axis, project the gear end face point cloud, establish the polar coordinate equation for the angle θ and distance H from the center point to the gear boundary, bring the fitted boundary points into the polar coordinate equation to calculate the unknown parameters, and use the maximum value Ra as the predetermined tooth top circle radius and the minimum value R as the trigonometric function equation. f As the predetermined tooth root circle radius, calculate the angles θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number

[0062] Step 4-2, use the axis as the rotation axis to establish a plane, project a rectangular point cloud, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i= Ra±w, it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter.

Claims

1. A gear parameter measurement method based on point cloud technology is characterized by: The method comprises the following steps: Step 1, preprocessing the original point cloud data; First, the point cloud is collected and input, and then the outliers in the original point cloud are removed by using Gaussian filtering, so as to obtain the point cloud Q1 containing only the tooth surface to be spliced ​​after preprocessing, and the point cloud Q2 containing part of the tooth surface and tooth width; Step 2, local public area splicing; The preprocessed point cloud data are spliced ​​to generate a local gear point cloud; first, the covariance of the two point clouds is calculated, and the eigenvector corresponding to the eigenvalue is calculated. The eigenvector corresponding to the maximum eigenvalue is used as the coordinate axis to establish a three-dimensional coordinate system in the two point clouds. The two coordinate systems are adjusted to align the common areas of the two point clouds. Then, the feature points of the common areas are determined from the two point clouds Q1 and Q2. The feature points of the common areas in the two point clouds after alignment correspond one to one and the corresponding Euclidean distances are equal, so point cloud feature matching can be used to achieve splicing. In step 2, a three-dimensional coordinate system is established in the two point clouds, and the two coordinate systems are adjusted so that the common areas of the two point clouds are aligned. In the above formulas, Q is a point cloud dataset in the format of a 3×N dimensional matrix, C is the covariance matrix of the Q matrix, V is a matrix composed of eigenvectors v1, v2, and v3, and D is a diagonal matrix composed of eigenvalues ​​d1, d2, and d3. The eigenvalues ​​are sorted from large to small, and the center point is taken as the origin. The eigenvectors corresponding to the first three eigenvalues ​​are taken as the coordinate axes in the three directions to establish a plane; then the two coordinate origins are moved to the same position, and R is the coordinate system rotation matrix. Respectively represent the angles between the three coordinate axes, t i To correspond to the translation distance, two coordinate systems are then obtained that correspond to each other after being aligned. In step 2, feature points in the common area of ​​the two point clouds are extracted, and the corresponding Euclidean distance is equal to achieve stitching. In the above formula, D is the Euclidean distance between the corresponding points in the two common point clouds, V is the distance between all corresponding feature points put into a matrix, the rows and columns in the matrix correspond to the distances between the feature points in the two point clouds, and the voting principle is used to find the distance d with the largest number of appearances in each row in the matrix V, and extract its corresponding row and column, which are the corresponding points with the correct relationship; finally, the optimal solution of the rotation matrix R and the translation vector T is used on the point cloud Q1, so that Q1Q2 is spliced. Step 3: Extract the inner hole axis of the gear point cloud. Equation F1 is established for the central circular hole on the gear point cloud surface, and it is used as the complete inner circular hole of the gear to obtain the radius r and center coordinates (a, b) of the circle. Each point is brought into the equation to determine whether it meets the equation F1. Starting from the selected seed point, its neighboring points are found and it is determined whether these neighboring points meet the characteristic function of the target surface. If they meet the characteristic function, they are added to the current clustering block and marked as new seed points. Then this process is repeated, and the outward search is performed until all points that meet the specific feature type and growth parameters are clustered to obtain the gear inner hole point cloud set Q. Since only part of the point cloud on one side of the gear inner hole can be collected, it is impossible to directly use it to fit the circular hole to obtain the center of the circle. Therefore, the gear inner hole point cloud Q is first layered according to the Z axis. Due to the characteristics of cylindrical spur gears, the radius r of each layer of the gear circular hole is the same. Therefore, the radius r obtained by the complete circle of the tooth surface in function F1 is used as the mark radius. For each layer of the circular hole, any two points are selected to establish the characteristic function F2 of the circle. The center coordinate O (a) of the circle of this layer can be obtained by using the incomplete arc of this layer. i , b i ), record the coordinate set E of all center points; select a point O(a i , b i ) and the center point N(a, b) of the gear end face use the straight line two-point method to construct the axis L, n is the direction vector, calculate the angle θ between the axis L and the normal vector of the gear end face, set the threshold T2, when the vector angle θ is less than the set threshold, the axis is considered to be the gear axis, if it is greater than, select any two points and repeat the iteration until the conditions are met. In the above formulas, F1 is the gear end face inner hole equation established to determine the inner hole radius r and the end face center point (a, b), Q is the set of gear inner hole points, t z is the layering accuracy, F2 is the equation established by the inner hole points of each layer after the inner hole is layered, O(ai,bi) is the center of the circle equation established in each layer, E is the set of center points, L is the fitting line n(u,v,w) is the direction vector, θ is the angle between the two lines, one of which is the line L constructed by the center point of one layer and the center point of the gear end face, and the other is a line with the normal vector of the gear end face as the direction. Step 4: Establish a gear measurement model and calculate relevant parameters. First, the boundary contour of the gear point cloud is extracted, and then plane F1 is established in the direction of the axis L to project the gear point cloud contour onto a gear point cloud. A rotation plane F2 is established with the axis L as the rotation axis to project the gear point cloud onto the plane to form a rectangular point cloud to establish a gear measurement model; curve fitting is performed on the gear contour point cloud data on the F1 plane to establish a polar coordinate equation for the angle θ and distance H from the center point to the gear boundary, and the fitted boundary points are substituted into the polar coordinate equation to calculate the unknown parameters. From the known trigonometric function equation, the maximum value Ra is used as the predetermined tooth top circle radius value and the minimum value R is used as the minimum value R f As the predetermined tooth root circle radius, calculate the angle θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number Establish a rotation plane F2 around the gear with the axis as the edge, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i =Ra±w, then it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter. In the above formula, H is the distance from the center point on the F1 plane to the gear point cloud contour, θ is the corresponding angle, N and M are the parameters to be determined, is the angle difference between two adjacent maximum values ​​in the trigonometric function, B is the tooth width, Da is the diameter of the tooth tip circle, bi is the length of the rectangle width on all planes F2, hi is the length of the rectangle on all planes F2; ​​m is the module, z is the number of teeth, The normal tooth system is 1 for the tooth addendum coefficient, s for the tooth thickness, P for the tooth pitch, and d for the pitch circle diameter.

2. The gear parameter measurement method based on point cloud technology according to claim 1 is characterized in that The splicing method in step 2 is as follows: Step 2-1, use the KD tree to establish the topological relationship of the point cloud, calculate the covariance matrix C to find its eigenvalues ​​and sort them, extract the first three principal components of the two point clouds as coordinate axes, and establish the rotation angle θ through the corresponding coordinate axis angle. Straighten it to ensure that the Euclidean distances of the corresponding feature points in the two common areas correspond one to one. Step 2-2, calculate the normal vector n(a, b, c) and curvature S of each point, set the curvature threshold T1, and find the corresponding feature points of the two point clouds. Establish matrix U, with the corresponding feature points selected by point cloud Q1 as the rows of matrix U, and the corresponding feature points selected by point cloud Q2 as the columns of matrix U. Matrix U corresponds to the distance between the feature points of Q1 and Q2. Use the voting principle to find the value with the most distance occurrences in matrix U, and its row and column are the correct correspondence between the feature points of Q1 and Q2.

3. The gear parameter measurement method based on point cloud technology according to claim 1 is characterized in that In step 3, in view of the fact that the inner hole of the gear cannot be completely photographed and it is difficult to obtain the axis, a method for extracting the gear axis based on the residual defect cloud of the arc is proposed. The specific contents are as follows: Step 3-1, after the boundary extraction in the previous step, use point cloud segmentation to separate the circular hole point cloud of the gear end face, select any three points to establish the equation of the circle, calculate multiple groups of circle centers and radii multiple times, and take the average as the standard parameters of the gear inner hole. Step 3-2: For the interior that cannot be photographed, divide the Z axis and sort the points in the point cloud according to the Z axis coordinate size, find the maximum value Zmax and the minimum value Zmin, and calculate the layering accuracy T z , establish the plane according to the layering accuracy, and classify the points with the same Z-axis coordinate value into the same plane point set. Step 3-3, perform circle fitting on each layer, take the radius of the circular hole on the end face as a parameter, select any two points to establish the function of the circle of the layer, and calculate the center of the layer. Step 3-4, collect the center points of each layer into a set and construct the axis with the center point of the end face using the two-point method, and directly calculate the angle between the axis and the normal vector direction of the gear end face. If the angle is less than the set threshold, it can be used as the gear axis. Otherwise, take another point from the center point set to recalculate the axis and angle for iteration.

4. The gear parameter measurement method based on point cloud technology according to claim 1 is characterized in that In step 4, the specific gear size measurement method is as follows: Step 4-1, establish a plane in the direction of the gear axis, project the gear end face point cloud, establish the polar coordinate equation for the angle θ and distance H from the center point to the gear boundary, substitute the fitted boundary points into the polar coordinate equation to calculate the unknown parameters, and use the maximum value Ra as the predetermined tooth top circle radius and the minimum value Rf as the predetermined tooth root circle radius from the known trigonometric function equation, calculate the angles θ1θ2 corresponding to the adjacent maximum or minimum values, and calculate their difference As the angle between two teeth of the gear, the last tooth number Step 4-2, use the axis as the rotation axis to establish a plane, project a rectangular point cloud, record the projection onto the plane, use the RANSAC method to fit the four boundaries into four straight lines, and record the lengths h of multiple sets of rectangles i and width b i , set the threshold w, if the length of the matrix h i =Ra±w, then it is proved that this rectangular projection corresponds to the semicircle of the gear, and the length of the matrix corresponds to the radius Ra of the tooth top circle; if the length of the matrix h i =2Ra±w, it is proved that this rectangular projection corresponds to the entire circle of the gear, and the length of the matrix corresponds to the tooth top circle diameter Da. After measuring multiple sets of data, the average value is taken as the accurate value of the final tooth top circle diameter.

Citation Information

Patent Citations

  • Gear scanning algorithm for template matching and point cloud comparison

    CN110599449A

  • Robot gear chamfering machining method based on parallel binocular stereoscopic vision

    CN113902810A

  • Point cloud registration method and device for part surface quality detection

    CN114663373A

  • Gear cross-rod distance measurement system and method based on non-contact 3D vision

    CN117029705A

Cited By

  • Hardware size error detection method and system based on point cloud processing

    CN120339369A

  • Gear three-dimensional structure construction method, system, equipment and medium

    CN120339505A

  • A gear three-dimensional structure construction method, system, device and medium

    CN120339505B

  • Platform and method for measuring tooth surface deviation of small-modulus gear

    CN122107995A