Three-dimensional measurement method and device for gear based on composite scanning

Abstract

This invention provides a method and apparatus for three-dimensional gear measurement based on composite scanning, belonging to the field of precision measurement technology. The method includes: acquiring the theoretical parameters of the target gear; performing a macroscopic scan of the target gear based on the target global trajectory corresponding to the theoretical parameters to obtain a macroscopic point cloud; performing a local scan of the target region of the target gear by superimposing a high-frequency micropath on the target global trajectory to obtain a microscopic point cloud; wherein the target region includes curvature abrupt change regions and defect regions; and fusing the macroscopic point cloud and the microscopic point cloud through a unified coordinate transformation to obtain a three-dimensional model of the target gear. This invention optimizes scanning efficiency through macroscopic scanning of the target gear, improves measurement accuracy by refining the trajectory of complex local regions through superimposing high-frequency micropaths, and achieves efficient and high-precision measurement of complex gear surfaces through point cloud fusion via a unified coordinate transformation.

Description

Technical Field

[0001] This invention relates to the field of precision measurement technology, and in particular to a method and apparatus for three-dimensional gear measurement based on composite scanning. Background Technology

[0002] In the field of high-end equipment manufacturing, gears, as key transmission components in core equipment such as aerospace, new energy vehicles, and precision machine tools, play a decisive role in the transmission efficiency, operating noise, and fatigue life of the equipment due to their tooth surface accuracy. Taking new energy vehicles as an example, the tooth surface error of reducer gears must be controlled within the micrometer level to effectively reduce transmission losses and vibration noise. Meanwhile, the complex curved surface gears in aero-engine gearboxes must meet stringent reliability requirements under high-speed and heavy-load conditions to ensure stable operation of the equipment under extreme conditions.

[0003] As the manufacturing industry moves towards higher precision and complexity, modern gear tooth surfaces exhibit complex geometric features such as asymmetric helical surfaces, modified tooth profiles, and tooth root transition curves. However, traditional measurement methods, such as contact coordinate measuring machines (CMMs) and two-dimensional grid scanning, reveal numerous irreconcilable contradictions when dealing with these complex tooth surfaces. On the one hand, while dense single-point measurement can acquire detailed data, it is extremely time-consuming, severely impacting production efficiency. On the other hand, sparse sampling makes it difficult to accurately capture the subtle features of local curvature abrupt changes, such as stress concentration at the tooth root, resulting in measurement results that do not accurately reflect the actual precision of the gear. Furthermore, different types of gears (such as spur gears, helical gears, and hypoid gears) have significant structural differences, and traditional measurement trajectories require independent programming, making it difficult to adapt to the needs of rapid measurement switching in multi-variety, small-batch production scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a three-dimensional measurement method and device for gears based on composite scanning, so as to achieve efficient and high-precision measurement of complex curved surfaces of gears.

[0005] In a first aspect, the present invention provides a three-dimensional gear measurement method based on composite scanning, comprising: Obtain the theoretical parameters of the target gear to be measured; Based on the target global trajectory corresponding to the gear theoretical parameters, a macroscopic scan of the target gear is performed to obtain a macroscopic point cloud. By superimposing high-frequency micropaths on the global trajectory of the target, a local scan of the target region of the target gear is performed to obtain a micro-view cloud; the target region includes curvature change region and defect region; By using a unified coordinate transformation, the macroscopic point cloud and the microscopic point cloud are fused to obtain a three-dimensional model of the target gear.

[0006] In an optional implementation, based on the target global trajectory corresponding to the gear's theoretical parameters, a macroscopic scan of the target gear is performed to obtain a macroscopic point cloud, including: An initial global trajectory is generated based on the gear theory parameters, which include gear type, module, pressure angle, and helix angle. The initial global trajectory is smoothed to obtain the target global trajectory; A macroscopic scan of the target gear is performed based on the target global trajectory to obtain a macroscopic point cloud. During the macroscopic scan, the probe posture of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the density of measurement points is dynamically adjusted based on the tooth surface curvature distribution obtained from the scan.

[0007] In an optional implementation, the probe attitude of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the density of measurement points is dynamically adjusted based on the tooth surface curvature distribution obtained by scanning, including: Based on the Frenet coordinate system, the normal vector of each point on the global trajectory of the target is calculated, and the probe posture is adjusted in real time by controlling the motion axis of the measuring equipment to make the probe direction consistent with the normal vector of the current measuring point. Based on the position information of each point on the tooth surface obtained from the current scan, the curvature of each point on the tooth surface is calculated in real time, and the sampling interval is dynamically adjusted according to the relationship between the curvature of each point and the preset curvature threshold.

[0008] In an optional implementation, a micro-viewpoint cloud is obtained by superimposing a high-frequency micropath onto the target global trajectory to perform a local scan of the target region of the target gear, including: Using curvature-based region segmentation and edge detection algorithms, region identification is performed on the digital data corresponding to the target gear to obtain the target region; wherein, the digital data includes the CAD model of the target gear or the point cloud data obtained by global scanning of the target gear; High-frequency micropaths are superimposed on the local trajectory corresponding to the target region in the global trajectory of the target to obtain the micro trajectory; The target region of the target gear is scanned microscopically based on the microscopic trajectory to obtain an initial microscopic point cloud; during the microscopic scanning process, the micropath density is adjusted in real time based on the real-time collected curvature change data. After binarizing the initial micro-viewpoint cloud, edge point correction and singular point correction are performed to obtain the micro-viewpoint cloud.

[0009] In an optional implementation, after binarizing the initial micro-viewpoint cloud, edge point correction and singular point correction are performed to obtain the micro-viewpoint cloud, including: The initial micro-viewpoint cloud is binarized to obtain a binarized micro-viewpoint cloud; By using a crawler method to track the tooth profile edge, and combining stepped pixel boundary feature analysis and adjacent curvature compensation, edge point correction is performed on the binarized micro-view cloud to obtain the corrected micro-view cloud. By using local coordinate system transformation and pixel interpolation techniques, singular point correction is performed on the corrected micro-viewpoint cloud to obtain the micro-viewpoint cloud.

[0010] In an optional implementation, a three-dimensional model of the target gear is obtained by fusing the macroscopic point cloud and the microscopic point cloud through a unified coordinate transformation, including: Using the coordinate system of the macro point cloud as the global unified coordinate system, the macro point cloud and the micro point cloud are registered to obtain the registration data; Based on the registration data, a weighted average method is used to fuse the micro point cloud into the macro point cloud to obtain a 3D model.

[0011] In an optional implementation, the coordinate system of the macro point cloud is used as a global unified coordinate system, and the macro point cloud and micro point cloud are registered to obtain registration data, including: The macroscopic point cloud is used as the target point cloud, the microscopic point cloud is used as the source point cloud, and an initial transformation matrix is ​​set. The kd-tree is used to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, thus obtaining the corresponding point pair data. Based on the corresponding point pair data, the transformation matrix is ​​updated using the least squares method to minimize the sum of squared distances between corresponding point pairs; Determine whether the preset iteration stop condition is met; If not, re-execute the step of using the kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix to obtain the corresponding point pair data; If so, the current transformation matrix is ​​determined as the registration data.

[0012] Secondly, the present invention provides a gear three-dimensional measurement device based on composite scanning, comprising: The data acquisition module is used to acquire the theoretical parameters of the target gear to be measured; The macroscopic scanning module is used to perform macroscopic scanning of the target gear based on the target global trajectory corresponding to the gear's theoretical parameters, thereby obtaining a macroscopic point cloud. The microscopic scanning module is used to perform local scanning of the target region of the target gear by superimposing high-frequency micropaths on the global trajectory of the target, thereby obtaining a microscopic view cloud; the target region includes curvature change region and defect region; The point cloud fusion module is used to fuse macroscopic point clouds and microscopic point clouds through a unified coordinate transformation to obtain a 3D model of the target gear.

[0013] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement any of the methods described in the foregoing embodiments.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, performs the method of any of the foregoing embodiments.

[0015] The present invention provides a method and apparatus for three-dimensional gear measurement based on composite scanning. This method acquires the theoretical parameters of the target gear; performs a macroscopic scan of the target gear based on the target global trajectory corresponding to the theoretical parameters, obtaining a macroscopic point cloud; and performs a local scan of the target region of the target gear by superimposing a high-frequency micropath on the target global trajectory, obtaining a microscopic point cloud. The target region includes curvature abrupt change regions and defect regions. Through unified coordinate transformation, the macroscopic point cloud and the microscopic point cloud are fused to obtain a three-dimensional model of the target gear. This approach optimizes scanning efficiency through macroscopic scanning of the target gear, improves measurement accuracy through trajectory refinement of complex local regions by superimposing high-frequency micropaths, and achieves efficient and high-precision measurement of complex gear surfaces through unified coordinate transformation and point cloud fusion. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 A schematic flowchart of a three-dimensional gear measurement method based on composite scanning provided in an embodiment of the present invention; Figure 2 A flowchart illustrating another gear three-dimensional measurement method based on composite scanning provided in an embodiment of the present invention; Figure 3 A schematic diagram of a gear three-dimensional measuring device based on composite scanning provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] In fields such as optical imaging and MEMS (Micro-Electro-Mechanical Systems), macro-micro composite drive technology, with its unique cross-scale measurement advantages, has become an effective means of solving complex measurement problems. For example, the coarse-fine adjustment platform of scanning electron microscopes significantly improves measurement efficiency and accuracy by combining rapid macroscopic positioning with fine microscopic measurement. However, in the crucial field of gear measurement, the application of macro-micro composite drive technology remains unexplored. With the rapid development of Industry 4.0 and intelligent manufacturing, gear manufacturing is shifting from the traditional "offline sampling inspection" mode to an "online full inspection" mode. This places higher demands on the adaptability, efficiency, and accuracy of measurement technology, urgently requiring innovative trajectory planning methods to meet the production needs of multi-variety, high-precision, and high-efficiency manufacturing.

[0020] To meet the demands of high-precision gear inspection, various gear measurement technologies have emerged, including traditional measurement algorithms based on equidistant and equiangular sampling, and trajectory generation methods based on CAD (Computer-Aided Design) models. However, traditional measurement algorithms do not fully consider the non-uniformity of tooth surface curvature distribution, resulting in a large number of redundant measurement points in flat areas, leading to resource waste. In areas with drastic curvature changes, the density of measurement points is relatively insufficient, failing to accurately reflect the actual tooth surface morphology. Trajectory planning methods based on CAD models struggle to overcome deviations between the actual tooth surface and the theoretical model (such as machining errors and heat treatment deformation), resulting in a mismatch between the measurement trajectory and the actual tooth surface. Furthermore, the data generated by macroscopic scanning (millimeter-level resolution) and microscopic scanning (micrometer-level resolution) lack a unified coordinate transformation framework, leading to significant splicing errors during data fusion. This fails to meet the stringent accuracy assessment requirements of standards such as ISO 1328 gear accuracy, hindering the improvement of gear manufacturing quality.

[0021] To address the challenges of high precision requirements for gear tooth surfaces in high-end equipment manufacturing, the difficulty of balancing efficiency and accuracy with traditional measurement methods, and the complexity of programming measurement trajectories for different types of gears, this invention provides a three-dimensional gear measurement method and device based on composite scanning. It employs a macro-micro composite scanning approach for complex gear surfaces and utilizes a "layered optimization-dynamic adaptation-data fusion" technical route for scanning trajectory planning, achieving efficient and high-precision measurement. Layered optimization refers to optimizing the trajectory at both the macro and micro levels; dynamic adaptation refers to dynamically adjusting the probe posture, measurement point density, and micropath density during the scanning process; and data fusion refers to the fusion of macro point clouds and micro point clouds.

[0022] To facilitate understanding of this embodiment, a detailed description of a gear three-dimensional measurement method based on composite scanning disclosed in this embodiment of the invention will be provided first.

[0023] This invention provides a method for three-dimensional gear measurement based on composite scanning. This method can be executed by an electronic device with data processing capabilities, located within a measuring device used to measure the gear and obtain a three-dimensional model of it. Exemplarily, the measuring device employs a Coordinate Measuring Machine (CMM), a device used for measuring geometric dimensions and tolerances, capable of accurately measuring the physical geometric characteristics of an object. The CMM mainly consists of a mechanical body, a probe system, a control system, and a software system. The mechanical body may include three mutually perpendicular motion axes (X, Y, Z), allowing the probe to move in three-dimensional space. The probe system can be sensors mounted on the CMM, which can be contact or non-contact, used to acquire data points on the surface of the object being measured. The control system manages the precise positioning and motion trajectory of the probe and each motion axis. The software system is used to program and control the measurement process, analyze data, and generate reports. Therefore, this method can be applied to the control and software systems of the measuring device, both of which are mounted on an electronic device.

[0024] See Figure 1 The diagram shows a flowchart of a three-dimensional gear measurement method based on composite scanning. The method mainly includes the following steps S110 to S140: Step S110: Obtain the theoretical gear parameters of the target gear to be measured.

[0025] The target gear can be a gear with a complex curved surface. The purpose of this embodiment is to use a measuring device to scan the target gear in a specific way to obtain a three-dimensional model of the target gear, so as to meet the production needs of multiple varieties, high precision and high efficiency.

[0026] The aforementioned theoretical gear parameters can be obtained by measuring the target gear using high-precision measuring instruments, or they can be obtained from the target gear's design documents. These theoretical parameters include gear type, module, pressure angle, and helix angle. Gear type can be categorized into involute cylindrical gears, hypoid bevel gears, etc., and the module... m The value range can be 1-10 mm, pressure angle α It can be 20°, etc., helix angle β The common range for helical gears is 8°-25°.

[0027] Step S120: Based on the target global trajectory corresponding to the gear theoretical parameters, perform a macroscopic scan of the target gear to obtain a macroscopic point cloud.

[0028] In some possible embodiments, step S120 above may include: generating an initial global trajectory based on gear theoretical parameters; wherein the gear theoretical parameters include gear type, module, pressure angle, and helix angle; smoothing the initial global trajectory to obtain a target global trajectory; performing a macroscopic scan of the target gear based on the target global trajectory to obtain a macroscopic point cloud; wherein, during the macroscopic scan, the probe posture of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the density of measurement points is dynamically adjusted based on the tooth surface curvature distribution obtained by the scan.

[0029] The smoothing process described above can, but is not limited to, using cubic spline interpolation. Using cubic spline interpolation to smooth the initial global trajectory not only effectively improves the smoothness and accuracy of the trajectory but also preserves important geometric features, while offering high computational efficiency and flexibility. Specifically, cubic spline interpolation ensures that the interpolation results between known data points have continuous first and second derivatives. This means it can provide a smoother transition while maintaining the original data characteristics, making it particularly suitable for applications requiring high precision. Since cubic splines are composed of piecewise cubic polynomials, and these polynomials and their first and second derivatives are continuous across the entire interval, they can effectively eliminate noise and reduce irregular fluctuations in the data, providing a smooth path for the scanning trajectory. An important characteristic of cubic splines is their locality; each spline segment is only affected by its adjacent data points. This ensures that adjustments to a single data point will not have a significant impact on the entire curve, which is beneficial for local optimization and smoothing. The solution algorithm for cubic spline interpolation is highly efficient and can quickly obtain the interpolation results. Cubic spline interpolation can preserve the trend and shape of the original data and avoid unexpected shape changes, which is crucial to ensuring that the processed data still accurately reflects the actual situation.

[0030] The Frenet coordinate system can be used when adjusting head posture, which not only improves the accuracy and efficiency of the measurement process but also simplifies the design and implementation of related algorithms. Specifically, the Frenet coordinate system is based on the local geometric properties of curves (such as tangents, normals, and binormals), thus providing highly accurate position and orientation information when describing motion along a trajectory. Using the Frenet framework simplifies the mathematical calculations related to path following and attitude control. The Frenet coordinate system naturally adapts to the shapes of curves and surfaces, making it easier for objects moving along complex paths (such as sensors or probes) to adjust their posture to maintain optimal contact with the measured surface. This contributes to improved measurement data quality and reliability.

[0031] Based on this, in one possible implementation, the steps of adjusting the probe posture of the measuring device in real time based on the normal vectors of each point on the target global trajectory, and dynamically adjusting the density of measurement points based on the tooth surface curvature distribution obtained by scanning, may include: calculating the normal vectors of each point on the target global trajectory based on the Frenet coordinate system, and adjusting the probe posture in real time by controlling the motion axis of the measuring device to make the probe direction consistent with the normal vector of the current measurement point; calculating the curvature of each point on the tooth surface in real time according to the position information of each point obtained by the current scanning, and dynamically adjusting the sampling interval according to the relationship between the curvature of each point and the preset curvature threshold.

[0032] In practice, the macroscopic layer can generate a global coverage trajectory (i.e., the initial global trajectory) based on the gear theoretical parameters. After smoothing by cubic spline interpolation, the probe posture is optimized using the Frenet coordinate system, and the density of measurement points is dynamically adjusted according to the tooth surface curvature, which can reduce redundant measurement points by more than 30%.

[0033] Step S130: By superimposing a high-frequency micropath on the target global trajectory, a local scan of the target region of the target gear is performed to obtain a micro-view cloud; wherein, the target region includes curvature change region and defect region.

[0034] In some possible embodiments, step S130 may include: using a curvature-based region segmentation algorithm and an edge detection algorithm to identify the target region from the digital data corresponding to the target gear; wherein the digital data includes a CAD model of the target gear or point cloud data obtained by global scanning of the target gear; superimposing high-frequency micropaths on the local trajectory corresponding to the target region in the global trajectory of the target to obtain a micro trajectory; performing a microscopic scan of the target region of the target gear based on the micro trajectory to obtain an initial microscopic view cloud; wherein, during the microscopic scan, the micropath density is adjusted in real time based on the real-time collected curvature change data; after binarizing the initial microscopic view cloud, edge point correction and singular point correction are performed to obtain the microscopic view cloud.

[0035] The point cloud data mentioned above can be the aforementioned macroscopic point cloud, or point cloud data acquired through other scanning methods such as laser scanning. The target region can include curvature abrupt change regions where the rate of curvature change between adjacent points exceeds a set curvature change threshold, or it can include potential defect regions (such as wear spots). The rate of curvature change between adjacent points can be denoted as... ,in, K i Indicates the first i Curvature at each point; set the rate of change of curvature threshold ∆ K th It can be set according to actual needs; there are no restrictions here. For example, ∆ K th =0.05mm -2 .

[0036] Based on the global trajectory of the target obtained at the macroscopic level, high-frequency micropaths can be superimposed on the identified target region. High-frequency micropaths can include S-shaped micropaths or spiral micropaths. Taking an S-shaped micropath as an example, in the local coordinate system, the micropath equation can be expressed as: ; in, The starting point on the target's global trajectory. a and b They are respectively x The amplitude and frequency parameters of the direction. a Used to control micropaths x The directional fluctuation range (generally 0.1-0.5mm). b Used to control the fluctuation frequency (value range 10-50). c and d They are respectively y By adjusting the amplitude and frequency parameters of the direction, local scanning with micrometer-level resolution can be achieved.

[0037] In one possible implementation, after binarizing the initial micro-viewpoint cloud, edge point correction and singular point correction are performed to obtain the micro-viewpoint cloud. This can include: binarizing the initial micro-viewpoint cloud to obtain a binarized micro-viewpoint cloud; using a crawler method to track the tooth profile edges, and combining stepped pixel boundary feature analysis and adjacent curvature compensation, edge point correction is performed on the binarized micro-viewpoint cloud to obtain a corrected micro-viewpoint cloud; and singular point correction is performed on the corrected micro-viewpoint cloud through local coordinate system transformation and pixel interpolation techniques to obtain the micro-viewpoint cloud.

[0038] In practice, the micro-layer can preprocess the gear CAD model or point cloud data, identify curvature change areas and defect areas, and overlay high-frequency micropaths. It can also use crawling methods to correct tooth surface deviations and dynamically adjust the micropath density in real time.

[0039] Step S140: Through unified coordinate transformation, the macro point cloud and micro point cloud are fused to obtain a three-dimensional model of the target gear.

[0040] In some possible embodiments, step S140 above may include: using the coordinate system where the macro point cloud is located as a global unified coordinate system, registering the macro point cloud and the micro point cloud to obtain registration data; and using a weighted average method to fuse the micro point cloud into the macro point cloud based on the registration data to obtain a three-dimensional model.

[0041] This embodiment provides an improved ICP (Iterative Closest Point) algorithm for registering macroscopic point clouds and microscopic point clouds. Based on this, the coordinate system of the macroscopic point cloud is used as a global unified coordinate system. The registration of the macroscopic and microscopic point clouds to obtain the registration data includes: using the macroscopic point cloud as the target point cloud and the microscopic point cloud as the source point cloud, and setting an initial transformation matrix; using a kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, obtaining corresponding point pairs; updating the transformation matrix using the least squares method based on the corresponding point pairs to minimize the sum of squared distances between corresponding point pairs; determining whether a preset iteration stopping condition is met; if not, re-executing the step of using a kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, obtaining corresponding point pairs; if yes, determining the current transformation matrix as the registration data. The iteration stopping condition can be set according to actual conditions and is not limited here. For example, the iteration stopping condition can be that the change in the mean square error between two iterations is less than a set threshold.

[0042] In practice, a globally unified coordinate system is established, and the improved ICP algorithm is used to register macroscopic point clouds and microscopic point clouds. The weighted average method is used to fuse overlapping area data to generate a full-tooth surface 3D model with an error of ≤±5μm, which meets the production needs of multiple varieties, high precision and high efficiency.

[0043] The three-dimensional gear measurement method based on composite scanning provided in this invention optimizes scanning efficiency through macroscopic scanning of the target gear, improves measurement accuracy by superimposing high-frequency micropaths for trajectory refinement in local complex areas, and achieves efficient and high-precision measurement of complex gear surfaces through unified coordinate transformation and point cloud fusion.

[0044] In one possible implementation, generating the initial global trajectory based on gear theoretical parameters can include: For involute cylindrical gears, the coordinates of each point can be calculated at certain angular intervals (e.g., 0.1°) using a programming algorithm based on the involute parametric equations, generating a basic trajectory along the tooth profile direction; then, combined with the tooth width direction, it is extended along the z-axis with a fixed step size (e.g., 0.5mm) to construct a three-dimensional basic trajectory covering the entire tooth surface (i.e., the initial global trajectory of the involute cylindrical gear). For helical gears, a helical equation is introduced based on the basic trajectory along the tooth profile direction of the involute cylindrical gear, extending the tooth profile curve along the helical direction to generate a helical trajectory along the tooth width direction (i.e., the initial global trajectory of the helical gear). The involute parametric equations can be as follows: ; in, r b The radius of the base circle can be determined using the formula... r b = mz cos α / 2( m For modulus, z This refers to the number of teeth on the gear. α It is calculated from the pressure angle; θ For the exhibition angle.

[0045] In one possible implementation, the process of smoothing the initial global trajectory using cubic spline interpolation may include: (1) For the n three-dimensional data points that have been acquired These three-dimensional data points are decomposed into three one-dimensional sequences, namely , , .

[0046] (2) Apply cubic spline interpolation. For each one-dimensional sequence (i.e. X , Y , Z The coefficients are calculated using one-dimensional cubic spline interpolation, which includes: determining the cubic polynomial expression for each interval; establishing a system of equations based on node conditions (function value continuity, first derivative continuity, and second derivative continuity); and for each dimension, introducing the second derivative value of the variable at each node. M i The above system of equations can be transformed into a tridiagonal linear system of equations; solving the tridiagonal linear system of equations will yield the second derivative value for each dimension. M i ; Use the obtained M i The coefficients of the cubic spline function for each interval are solved by back substitution.

[0047] (3) Construct a three-dimensional spline curve. Once obtained... X , Y , Z Cubic spline interpolation functions in each direction can be combined to represent spline curves in the entire three-dimensional space. For parameters... t (Typically representing the path length or time along the curve), curves can be described using parametric forms: ,in, S x ( t ), S y ( t ), S z ( t ) are respectively corresponding to X , Y , Z The cubic spline interpolation function for the axis.

[0048] In this way, a smooth spline curve can be created for a 3D data point set, which satisfies the properties of cubic spline interpolation in each dimension. This method not only preserves the positional information of the original data points but also ensures consistency and smooth transitions across all dimensions, thereby effectively eliminating abrupt velocity changes at path corners.

[0049] In one possible implementation, when calculating the normal vector of each point on the target global trajectory based on the Frenet coordinate system, the tangent vector of each point on the target global trajectory can be calculated using the following formula. Normal vector and binormal vector : , , ; in, The trajectory parametric equations corresponding to the global trajectory of the target are given. These equations consist of three independent functions, each describing the trajectory of the object along its path. x , y and z The position of the axis depends on a parameter (such as time). t The change of the normal vector. As a basis for adjusting the probe posture, the probe posture is adjusted in real time by controlling the motion axis of the measuring equipment to ensure that the probe direction is aligned with the normal vector. Consistent, enabling precise measurements perpendicular to the tooth surface.

[0050] Optionally, during the aforementioned macroscopic scanning process, when dynamically adjusting the density of measurement points based on the tooth surface curvature distribution, the curvature of each point on the tooth surface may include Gaussian curvature and average curvature. K and mean curvature H The formula for calculating can be as follows: , ; in, E , F , G The coefficients are the first fundamental form coefficients. L , M , N For the second fundamental form coefficient. When the calculated curvature K and H When all values ​​are less than a preset threshold, the region where the point is located is determined to be a region with gentle curvature. In this case, the sampling interval can be increased to twice the initial sampling interval. Conversely, in regions with complex curvature, the initial sampling interval is maintained to reduce redundant measurement points and improve measurement efficiency. It should be noted that in some embodiments, both regions with gentle curvature and regions with complex curvature can be further divided into two or more categories. The sampling intervals corresponding to different categories of regions with gentle curvature are different multiples of the initial sampling interval, such as 2 times, 3 times, or more. The sampling intervals corresponding to different categories of regions with complex curvature can be the initial sampling interval, 1 / 2 of the initial sampling interval, 1 / 3 of the initial sampling interval, etc.

[0051] The first basic form mentioned above concerns the metric properties of the surface itself, describing the variation in distance between two points on the surface. For a parameterized surface r( u , v The coefficients of the first basic form are: , , Here, r u and r v These are the position vector r paired with parameters. u and v The partial derivatives of . Where, u It can vary along the tooth width direction. v It can vary along the direction of tooth height or tooth profile.

[0052] The second fundamental form mentioned above relates to how a surface is embedded in three-dimensional space; it describes the variation along the surface normal direction. Its coefficients are: , , Here, r uu r uv and r vv These are the position vector r paired with parameters. u and vThe partial derivative of , where n is the unit normal vector at a point on the surface.

[0053] In one possible implementation, the above-mentioned use of curvature-based region segmentation and edge detection algorithms to identify regions in the digitized data corresponding to the target gear, thereby obtaining the target region, may include: using a curvature-based region segmentation algorithm to set a curvature change rate threshold ∆. K th =0.05mm -2 When the rate of change of curvature of adjacent points >∆ K th At this time, the region is marked as a curvature abrupt change region. Simultaneously, edge detection algorithms in image processing (such as the Canny operator) are used to identify potential defect regions (such as wear spots).

[0054] In one possible implementation, adjusting the micropath density in real time based on the real-time acquired curvature change data may include: real-time acquisition of the probe's curvature change rate ∆ K When ∆ K > ∆ K th When the micropath sampling interval is reduced to 0.5 times its original value; when ∆ K < ∆ K th When the sampling interval is increased to 1.5 times its original value, the micropath sampling interval remains unchanged in other cases. This achieves dynamic adjustment of the micropath density, ensuring measurement accuracy.

[0055] In one possible implementation, binarizing the initial micro-viewpoint cloud may include: setting a grayscale threshold T=128 (8-bit grayscale image), marking points in the initial micro-viewpoint cloud with grayscale values ​​greater than T as 1, and points with grayscale values ​​less than T as 0.

[0056] In one possible implementation, using a crawler to track the tooth profile edge, combined with stepped pixel boundary feature analysis and adjacent curvature compensation, to correct the edge points of the binarized micro-view cloud can include: starting from a certain point on the tooth profile edge of the binarized micro-view cloud using a crawler, tracking the edge point according to the pixel value change law, and combining stepped pixel boundary feature analysis and adjacent curvature compensation to correct the deviation between the theoretical tooth surface and the actual tooth surface.

[0057] In one possible implementation, the singularity correction of the modified micro-view cloud can be achieved through local coordinate system transformation and pixel interpolation techniques. This can include: for singularities, establishing a local coordinate system, transforming the singularity coordinates to the new coordinate system, and using trilinear interpolation techniques to calculate the missing coordinate values, effectively solving the problem of numerical instability.

[0058] The specific process of the above singularity correction can be summarized as follows: Step 1: Define a local coordinate system for each singular point.

[0059] A local coordinate system can be determined in the following ways: (1) Select reference points: Select several neighboring points around the singular point as reference points. These points should be as close to the singular point as possible and evenly distributed. (2) Construct a local coordinate system: The best-fit plane can be found using methods such as principal component analysis, and then a local coordinate system can be constructed based on this plane. Alternatively, a local coordinate system can be defined simply using three non-collinear points.

[0060] Step 2: Transform the coordinates of the singular point and its neighboring reference points to the newly established local coordinate system.

[0061] Step 3: In the new local coordinate system, calculate the coordinates of the singular points using trilinear interpolation.

[0062] Step 4: After completing the interpolation, transform the obtained coordinate values ​​back to the original coordinate system.

[0063] In one possible implementation, for the fusion of macroscopic point clouds and microscopic point clouds, a globally unified coordinate system is established, with the central axis of the gear as the coordinate system. z The origin is set as the center of the tooth tip circle. The specific registration steps are as follows: First, initialize by selecting the macro point cloud as the target point cloud. P Microview Cloud as the source cloud Q And set the initial transformation matrix. T 0. Then perform a corresponding point search, using a kd-tree data structure to accelerate the calculation. Q Each point in T Under 0 transformation and P The correspondence between the nearest points is determined. Then, the optimal transformation matrix is ​​calculated using the least squares method. T 1. Minimize the sum of the squares of the distances between corresponding pairs of points. Let T 0= T 1. Repeat the corresponding point search and transformation parameter calculation steps until the change in mean square error between two iterations is less than the set threshold ε=10. -6 mm 2 .

[0064] The weighted average method is used to fuse data in overlapping areas. For points in the overlapping areas... p i and q i The merged point coordinates for: ; in, w1 and w 2 represents the weights for macroscopic and microscopic point clouds, respectively, which can be set according to the measurement accuracy. For example, the weights for macroscopic point clouds... w 1 = 0.3, the weight of Microview Cloud w 2 = 0.7.

[0065] For ease of understanding, this embodiment also provides a more detailed overall process of the above method.

[0066] See Figure 2 The diagram illustrates another method for 3D gear measurement based on composite scanning. After obtaining the theoretical parameters of the gear, macroscopic trajectory planning is performed. Macroscopic trajectory planning includes: generating a basic trajectory for the entire tooth surface, smoothing the trajectory through cubic spline interpolation, optimizing the probe posture using the Frenet coordinate system, and dynamically adjusting the density of measurement points based on curvature to obtain a gear CAD model (i.e., macroscopic point cloud).

[0067] After macroscopic trajectory planning is completed, microscopic trajectory planning is performed. Microscopic trajectory planning includes: identifying regions of abrupt curvature change, superimposing high-frequency micropaths, binarizing point clouds, correcting tooth surface deviations using a crawler method, and dynamically adjusting micropath density.

[0068] After the micro-level trajectory planning is completed, macro-micro data fusion is performed. Macro-micro data fusion includes: establishing a globally unified coordinate system, registering macro and micro point cloud data (i.e., macro point cloud and micro point cloud) using an improved ICP algorithm, and fusing overlapping region data using a weighted average method.

[0069] Using the above methods, a full-tooth surface 3D model with an error ≤ ±5μm can be generated, enabling efficient and high-precision measurement of complex gear surfaces.

[0070] For ease of understanding, the following is about Figure 2 The content will be further introduced.

[0071] The method proposed in this invention achieves efficient and high-precision measurement of complex gear surfaces through a technical approach of "layered optimization-dynamic adaptation-data fusion". Specifically, it includes: generating a global coverage path based on a gear theoretical model to optimize scanning efficiency; refining the trajectory for local complex areas to improve measurement accuracy; and unifying coordinate transformation and point cloud fusion to generate a complete 3D model.

[0072] In macroscopic trajectory planning, a globally covered trajectory is constructed based on gear theoretical parameters. According to the theoretical parameters such as module, pressure angle, and helix angle of the gear type (involute cylindrical gear, quasi-hyperboloid bevel gear, etc.), parametric equations are used to generate a basic trajectory covering the entire tooth surface (i.e., the initial global trajectory). For helical gears, a helical trajectory is generated along the tooth width direction to ensure coverage of the tooth tip, tooth root, and working tooth surface. To optimize trajectory quality, cubic spline interpolation is used to smoothly fit the initial global trajectory, eliminating sudden velocity changes at path corners. Using the Frenet coordinate system, the trajectory is decomposed into tangential, normal, and sub-normal components. The probe posture is dynamically adjusted based on the normal vector to ensure the probe is always perpendicular to the tooth surface, reducing measurement errors. Simultaneously, the measurement point density is intelligently adjusted according to the tooth surface curvature distribution, reducing the number of sampling points in areas with gentle curvature and maintaining a basic density in complex areas, thereby reducing redundant measurement points by over 30% and significantly improving scanning efficiency.

[0073] The micro-level focuses on precise measurements of complex areas. By preprocessing gear CAD models or initial scan point clouds, it accurately identifies regions with abrupt curvature changes (tooth root fillets, modified tooth profiles, etc.) and potential defect areas (wear spots). High-frequency micropaths, such as S-shaped or spiral micropaths, are superimposed on the macro-trajectory (i.e., the target global trajectory) to achieve micrometer-level resolution local scanning. After binarizing the scan point cloud, a crawler method is used to track the tooth profile edges. Combined with stepped pixel boundary feature analysis and adjacent curvature compensation, the deviation between theoretical and actual tooth surfaces is corrected. For singular points, local coordinate system transformation and pixel interpolation techniques are used to address numerical instability. Furthermore, data such as curvature changes are acquired in real time. When the rate of curvature change exceeds a threshold, the micropath density is automatically increased; when the rate of curvature change is less than half of the threshold, the micropath density is automatically decreased, improving measurement efficiency while maintaining accuracy.

[0074] In the macro-point cloud-micro-viewpoint cloud fusion process, a globally unified coordinate system is first established, and an improved ICP algorithm is used to register the macro-point cloud and micro-viewpoint cloud, eliminating stitching errors from scanning at different scales. Then, a weighted average method is used to fuse the overlapping area data, ultimately generating a full-tooth surface 3D model with an error ≤ ±5μm.

[0075] In the above process, a continuous and differentiable scanning trajectory is generated by cubic spline interpolation, which is then converted into probe posture parameters through the Frenet coordinate system to ensure that the probe is in contact with the tooth surface normal. After binarizing the scanned point cloud, a boundary tracking algorithm is used to extract the tooth profile. Edge points are compensated based on the curvature gradient of adjacent pixels, which improves the tooth profile edge extraction accuracy to ±2μm, providing a guarantee for the construction of a high-quality 3D model.

[0076] Corresponding to the aforementioned gear three-dimensional measurement method based on composite scanning, this embodiment of the invention also provides a gear three-dimensional measurement device based on composite scanning. See also... Figure 3The diagram shows a structural schematic of a gear three-dimensional measurement device based on composite scanning. The device includes: The data acquisition module 301 is used to acquire the theoretical parameters of the target gear to be measured; The macroscopic scanning module 302 is used to perform macroscopic scanning on the target gear according to the target global trajectory corresponding to the gear theoretical parameters, and obtain macroscopic point cloud; The microscopic scanning module 303 is used to perform local scanning of the target region of the target gear by superimposing a high-frequency micropath on the global trajectory of the target, thereby obtaining a microscopic view cloud; wherein, the target region includes curvature change region and defect region; The point cloud fusion module 304 is used to fuse macroscopic point clouds and microscopic point clouds through a unified coordinate transformation to obtain a three-dimensional model of the target gear.

[0077] The gear 3D measurement device based on composite scanning provided in this invention optimizes scanning efficiency through macroscopic scanning of the target gear, improves measurement accuracy by superimposing high-frequency micropaths for trajectory refinement in local complex areas, and achieves efficient and high-precision measurement of complex gear surfaces through point cloud fusion via unified coordinate transformation.

[0078] Furthermore, the aforementioned macroscopic scanning module 302 is specifically used for: generating an initial global trajectory based on gear theoretical parameters; wherein the gear theoretical parameters include gear type, module, pressure angle, and helix angle; smoothing the initial global trajectory to obtain a target global trajectory; performing a macroscopic scan of the target gear based on the target global trajectory to obtain a macroscopic point cloud; wherein, during the macroscopic scanning process, the probe posture of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the density of measurement points is dynamically adjusted based on the tooth surface curvature distribution obtained by scanning.

[0079] Furthermore, the aforementioned macroscopic scanning module 302 is also used to: calculate the normal vector of each point on the global trajectory of the target based on the Frenet coordinate system, and adjust the probe posture in real time by controlling the motion axis of the measuring device so that the probe direction is consistent with the normal vector of the current measurement point; calculate the curvature of each point on the tooth surface in real time according to the position information of each point obtained by the current scan, and dynamically adjust the sampling interval according to the relationship between the curvature of each point and the preset curvature threshold.

[0080] Furthermore, the aforementioned microscopic scanning module 303 is specifically used for: using a curvature-based region segmentation algorithm and an edge detection algorithm to identify the target region from the digital data corresponding to the target gear; wherein, the digital data includes the CAD model of the target gear or point cloud data obtained by global scanning of the target gear; superimposing high-frequency micropaths on the local trajectory corresponding to the target region in the global trajectory of the target to obtain a microscopic trajectory; performing a microscopic scan of the target region of the target gear based on the microscopic trajectory to obtain an initial microscopic view cloud; wherein, during the microscopic scanning process, the micropath density is adjusted in real time based on the curvature change data collected in real time; after binarizing the initial microscopic view cloud, edge point correction and singular point correction are performed to obtain the microscopic view cloud.

[0081] Furthermore, the aforementioned microscopic scanning module 303 is also used to: binarize the initial microscopic viewpoint cloud to obtain a binarized microscopic viewpoint cloud; track the tooth profile edge using a crawler method, and combine stepped pixel boundary feature analysis and adjacent curvature compensation to correct the edge points of the binarized microscopic viewpoint cloud to obtain a corrected microscopic viewpoint cloud; and correct the singular points of the corrected microscopic viewpoint cloud using local coordinate system transformation and pixel interpolation techniques to obtain a microscopic viewpoint cloud.

[0082] Furthermore, the point cloud fusion module 304 is specifically used to: use the coordinate system of the macro point cloud as the global unified coordinate system, register the macro point cloud and the micro point cloud to obtain registration data; and use the weighted average method to fuse the micro point cloud into the macro point cloud to obtain a three-dimensional model based on the registration data.

[0083] Furthermore, the point cloud fusion module 304 is also used to: take the macroscopic point cloud as the target point cloud, take the microscopic point cloud as the source point cloud, and set an initial transformation matrix; use a kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, and obtain corresponding point pair data; update the transformation matrix according to the corresponding point pair data using the least squares method to minimize the sum of squared distances between corresponding point pairs; determine whether the preset iteration stopping condition is met; if not, re-execute the step of using a kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, and obtain corresponding point pair data; if yes, determine the current transformation matrix as the registration data.

[0084] The device provided in this embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0085] like Figure 4As shown, an electronic device 400 provided in this embodiment of the invention includes: a processor 401, a memory 402 and a bus. The memory 402 stores a computer program that can run on the processor 401. When the electronic device 400 is running, the processor 401 and the memory 402 communicate through the bus. The processor 401 executes the computer program to realize the above-mentioned gear three-dimensional measurement method based on composite scanning.

[0086] Specifically, the memory 402 and processor 401 mentioned above can be general-purpose memory and processor, without any specific limitations here.

[0087] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, this computer program performs the three-dimensional gear measurement method based on composite scanning described in the preceding method embodiments. The computer-readable storage medium includes various media capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), RAM, magnetic disk, or optical disk.

[0088] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0089] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limitations; therefore, other examples of exemplary embodiments may have different values.

[0090] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0091] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0092] The modules described as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0093] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0094] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for three-dimensional measurement of gears based on composite scanning, characterized in that, include: Obtain the theoretical parameters of the target gear to be measured; Based on the target global trajectory corresponding to the gear's theoretical parameters, a macroscopic scan of the target gear is performed to obtain a macroscopic point cloud. This includes: generating an initial global trajectory based on the gear's theoretical parameters, wherein the gear's theoretical parameters include gear type, module, pressure angle, and helix angle; smoothing the initial global trajectory to obtain a target global trajectory; and performing a macroscopic scan of the target gear based on the target global trajectory to obtain a macroscopic point cloud. During the macroscopic scan, the probe posture of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the density of measurement points is dynamically adjusted based on the tooth surface curvature distribution obtained from the scan. The method of dynamically adjusting the measurement point density based on the tooth surface curvature distribution obtained by scanning includes: calculating the Gaussian curvature and average curvature of each point on the tooth surface in real time according to the position information of each point on the tooth surface obtained by the current scanning; when the calculated Gaussian curvature and average curvature are both less than the corresponding preset threshold, the area where the point is located is determined to be a region with gentle curvature, and the sampling interval of the macroscopic scan is increased to twice the initial sampling interval; otherwise, the area where the point is located is determined to be a region with complex curvature, and the initial sampling interval is maintained. By superimposing high-frequency micropaths on the target global trajectory, a local scan of the target region of the target gear is performed to obtain a micro-viewpoint cloud. This includes: using a curvature-based region segmentation algorithm and an edge detection algorithm to identify the target region from the digitized data corresponding to the target gear; superimposing high-frequency micropaths on the local trajectory corresponding to the target region in the target global trajectory to obtain a micro-trajectory; performing a micro-scan of the target region of the target gear based on the micro-trajectory to obtain an initial micro-viewpoint cloud; binarizing the initial micro-viewpoint cloud and then performing edge point correction and singular point correction to obtain the final micro-viewpoint cloud; wherein the target region includes curvature abrupt change regions and defect regions; the digitized data includes a CAD model of the target gear or point cloud data obtained from a global scan of the target gear; during the micro-scanning process, the micropath density is adjusted in real-time based on real-time collected curvature change data; The step of adjusting the micropath density in real time based on the real-time collected curvature change data includes: real-time acquisition of the probe's curvature change rate. K ,when K > K th When the micropath sampling interval is reduced to 0.5 times its original value, the sampling interval is reduced to 0.5 times its original value. K < K th When the micropath sampling interval is increased to 1.5 times its original value, the micropath sampling interval remains unchanged in other cases. K th The set threshold for the rate of change of curvature; By using a unified coordinate transformation, the macroscopic point cloud and the microscopic point cloud are fused to obtain a three-dimensional model of the target gear.

2. The composite scan based gear three-dimensional measurement method according to claim 1, characterized in that, The real-time adjustment of the probe attitude of the measuring device based on the normal vectors of each point on the target global trajectory includes: Based on the Frenet coordinate system, the normal vector of each point on the global trajectory of the target is calculated, and the probe posture is adjusted in real time by controlling the motion axis of the measuring device so that the probe direction is consistent with the normal vector of the current measuring point.

3. The composite scan based gear three-dimensional measurement method according to claim 1, wherein, After binarizing the initial micro-viewpoint cloud, edge point correction and singular point correction are performed to obtain the micro-viewpoint cloud, including: The initial micro-viewpoint cloud is binarized to obtain a binarized micro-viewpoint cloud; By using a crawler method to track the tooth profile edge, and combining stepped pixel boundary feature analysis and adjacent curvature compensation, the edge point correction is performed on the binarized micro-view cloud to obtain the corrected micro-view cloud. The singularity of the modified micro-viewpoint cloud is corrected by local coordinate system transformation and pixel interpolation techniques to obtain the micro-viewpoint cloud.

4. The composite scan based gear three-dimensional measurement method according to claim 1, wherein, The process of fusing the macroscopic point cloud and the microscopic point cloud through a unified coordinate transformation to obtain a three-dimensional model of the target gear includes: Using the coordinate system of the macro point cloud as the global unified coordinate system, the macro point cloud and the micro point cloud are registered to obtain registration data. Based on the registration data, the micro-point cloud is fused into the macro-point cloud using a weighted average method to obtain the three-dimensional model.

5. The composite scan based gear three-dimensional measurement method according to claim 4, characterized in that, The process of using the coordinate system of the macro point cloud as a global unified coordinate system to register the macro point cloud and the micro point cloud to obtain registration data includes: The macroscopic point cloud is used as the target point cloud, the microscopic point cloud is used as the source point cloud, and an initial transformation matrix is ​​set. The kd-tree is used to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix, thus obtaining the corresponding point pair data. Based on the corresponding point pair data, the transformation matrix is ​​updated using the least squares method to minimize the sum of squared distances between corresponding point pairs; Determine whether the preset iteration stopping condition is met; If not, repeat the step of using the kd-tree to calculate the correspondence between each point in the source point cloud and the nearest point in the target point cloud under the current transformation matrix to obtain the corresponding point pair data; If so, the current transformation matrix is ​​determined as the registration data.

6. A gear three-dimensional measuring device based on compound scanning, characterized in that, include: The data acquisition module is used to acquire the theoretical parameters of the target gear to be measured; The macroscopic scanning module is used to perform a macroscopic scan of the target gear based on the target global trajectory corresponding to the gear theoretical parameters, and obtain a macroscopic point cloud. This includes: generating an initial global trajectory based on the gear theoretical parameters, wherein the gear theoretical parameters include gear type, module, pressure angle, and helix angle; smoothing the initial global trajectory to obtain a target global trajectory; and performing a macroscopic scan of the target gear based on the target global trajectory to obtain a macroscopic point cloud. During the macroscopic scanning process, the probe posture of the measuring device is adjusted in real time based on the normal vector of each point on the target global trajectory, and the measurement point density is dynamically adjusted based on the tooth surface curvature distribution obtained from the scan. The method of dynamically adjusting the measurement point density based on the tooth surface curvature distribution obtained by scanning includes: calculating the Gaussian curvature and average curvature of each point on the tooth surface in real time according to the position information of each point on the tooth surface obtained by the current scanning; when the calculated Gaussian curvature and average curvature are both less than the corresponding preset threshold, the area where the point is located is determined to be a region with gentle curvature, and the sampling interval of the macroscopic scan is increased to twice the initial sampling interval; otherwise, the area where the point is located is determined to be a region with complex curvature, and the initial sampling interval is maintained. A microscopic scanning module is used to perform a local scan of the target region of the target gear by superimposing high-frequency micropaths on the target global trajectory to obtain a microscopic point cloud. This includes: using a curvature-based region segmentation algorithm and an edge detection algorithm to identify the target region from the digitized data corresponding to the target gear; superimposing high-frequency micropaths on the local trajectory corresponding to the target region in the target global trajectory to obtain a microscopic trajectory; performing a microscopic scan of the target region of the target gear based on the microscopic trajectory to obtain an initial microscopic point cloud; binarizing the initial microscopic point cloud and then performing edge point correction and singular point correction to obtain the final microscopic point cloud; wherein the target region includes curvature abrupt change regions and defect regions; the digitized data includes a CAD model of the target gear or point cloud data obtained from a global scan of the target gear; during the microscopic scanning process, the micropath density is adjusted in real time based on real-time collected curvature change data; The step of adjusting the micropath density in real time based on the real-time collected curvature change data includes: real-time acquisition of the probe's curvature change rate. K ,when K > K th When the micropath sampling interval is reduced to 0.5 times its original value, the sampling interval is reduced to 0.5 times its original value. K < K th When the micropath sampling interval is increased to 1.5 times its original value, the micropath sampling interval remains unchanged in other cases. K th The set threshold for the rate of change of curvature; The point cloud fusion module is used to fuse the macroscopic point cloud and the microscopic point cloud through a unified coordinate transformation to obtain a three-dimensional model of the target gear.

7. An electronic device comprising a memory, a processor, the memory having stored therein a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the gear three-dimensional measurement method based on composite scanning as described in any one of claims 1-5.

8. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program, when executed by the processor, performs the gear three-dimensional measurement method based on composite scanning as described in any one of claims 1-5.

Images