Automated measurement method and system based on the size of an automobile wheel hub

Abstract

The present application relates to the technical field of optical size measurement, in particular to a size automatic measurement method and system based on automobile wheel hub, comprising the following steps: controlling the industrial computer to set light source and exposure and record angle timing, collecting images according to angle, extracting and recombining edge points, comparing distribution, eliminating jitter interference and screening continuous sequence, aligning edge points and extracting boundary line, and generating wheel hub contour graph set. In the present application, by constructing image acquisition timing based on angle trigger, the accurate correspondence of edge points and angle information is realized, the stable correlation of edge point sequence in axial direction is ensured, the jitter interference is eliminated by means of change detection, the continuous edge data is screened, the boundary elements are extracted by setting equal interval reference position in image, the line frame graph set with clear structure is generated, the complete contour graph is constructed by cross section graph superposition, the precision and consistency of contour reconstruction are improved, the recognition ability for complex structure is enhanced, and the stability of measurement data and the anti-interference performance of image processing are improved.

Description

Technical Field

[0001] This invention relates to the field of optical dimension measurement technology, and in particular to an automated method and system for measuring the dimensions of automobile wheel hubs. Background Technology

[0002] The field of optical dimension measurement technology involves non-contact detection and evaluation of the geometric dimensions of objects using optical principles. Core aspects include the precise measurement of linear or curved dimensions such as length, diameter, contour, displacement, and thickness based on laser triangulation, structured light, interferometry, and optical imaging. This technology is widely used in industrial manufacturing, quality control, and automated inspection, characterized by high precision and efficiency. It is often combined with image processing and CNC technology to achieve automated inspection and feedback control. Traditional automated dimension measurement methods involve acquiring images or laser reflection signals of key parts of workpieces such as automobile wheel hubs during production or assembly using industrial cameras and laser rangefinders. Then, image analysis algorithms or laser contour analysis are used to extract contour edges and measure dimensional parameters such as diameter, depth, and thickness. These are then compared with manually set benchmark values ​​to complete dimension identification and tolerance judgment. Typically, two-dimensional image recognition or single-point laser scanning methods are used. However, these methods are limited by equipment layout and workpiece movement trajectory, resulting in problems such as fluctuating measurement accuracy, weak resistance to interference, and inability to adapt to complex structures.

[0003] Existing technologies lack precise control over the hub rotation angle during image acquisition, and the lack of a unified angular benchmark between image frames leads to discontinuities and misalignments in edge point distribution. This makes it impossible to construct a stable image sequence for subsequent analysis. In some areas, due to insufficient angular coverage or image interference, false edges are easily generated, interfering with the accuracy of size determination. Edge extraction and image comparison results are highly sensitive to changes in lighting and workpiece surface, and fluctuation errors are difficult to identify and eliminate, causing radial dimension measurement fluctuations. This affects the overall detection stability and image restoration integrity, making it difficult to support fine analysis and high consistency determination of contour structures. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: an automated measurement method for the dimensions of automobile wheel hubs, comprising the following steps:

[0005] S1: Control the industrial computer to set the light source status and laser exposure cycle, read the reference trigger signal and the equal angle trigger signal, record the response time of the angle phase, arrange the response times in angle order, and construct an optical trigger timing set;

[0006] S2: Call the optical trigger timing set to control image acquisition, extract edge points in the axial and radial directions of the image frame, classify the edge points according to the angle phase number, reorganize the edge points of the same angle phase in axial order, and generate an initial sequence set of image edge points.

[0007] S3: Call the initial sequence set of image edge points, compare the distribution of edge points in the radial direction of adjacent angular phases, identify and remove jittery interference point sequences, and filter the continuous image edge point sequence set;

[0008] S4: Call the continuous image edge point sequence set, align the edge point column based on the angle phase number, set the axial reference position, extract the boundary points and draw the boundary lines, and combine them to generate a set of contour section wireframe diagrams;

[0009] S5: Call the set of contour section wireframe diagrams, extract the radial distance between boundary lines, draw the diameter line and outer boundary line graphics in the order of angle and phase, and combine them to generate a set of automotive wheel hub measurement contour graphics.

[0010] As a further aspect of the present invention, the optical trigger timing set includes a reference trigger signal time, an equal-angle trigger signal time, an angle phase number, and a trigger time arrangement sequence; the initial sequence set of image edge points includes an axial edge point set, a radial edge point set, a phase number correspondence, and edge point classification results; the continuous image edge point sequence set includes radial distribution change points, jump interference marker points, and a continuous filtering edge point sequence; the contour cross-section wireframe diagram set includes equally spaced reference points, boundary line node structures, and angle phase corresponding wireframe diagrams; and the automotive wheel hub measurement contour graphic set includes a diameter distribution line diagram, a contour outer boundary line diagram, and an angle sequence combination graphic structure.

[0011] As a further aspect of the present invention, the step of identifying and removing jittery interference points refers to comparing the distribution of edge points in the radial direction of adjacent angular phases, identifying the changing point sequence, and removing it from the edge point sequence.

[0012] As a further aspect of the present invention, the step of classifying edge points by angle and phase number refers to classifying and organizing the extracted edge points according to their corresponding angle and phase numbers.

[0013] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0014] S101: Based on the industrial control computer control signal, the exposure period of the light source control signal and the laser profile acquisition is configured with parameters. The exposure duration and emission period parameters in the laser profiler and the light source controller are called to establish the time synchronization relationship between the light source emission and the laser acquisition, and an exposure period synchronization configuration group is generated.

[0015] S102: Based on the exposure cycle synchronization configuration group, collect the reference trigger signal and the constant angle trigger signal output by the rotating fixture, extract the time index representing the wheel hub angle, and normalize and rearrange all constant angle trigger times to generate the wheel hub constant angle response time sequence.

[0016] S103: Call the wheel hub angle response time series, sort them sequentially according to the angle position within the rotation cycle based on the time index, perform linear fitting processing, obtain the time mapping point corresponding to the angle, and establish an optical trigger timing set.

[0017] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0018] S201: Call the optical trigger timing set, control the line laser profilometer to acquire contour image frames according to the trigger time, synchronously start image acquisition for each trigger time point, and number and store the acquired image frames according to the trigger sequence number to establish a contour image frame sequence set;

[0019] S202: Based on the contour image frame sequence set, detect the gray-level gradient changes in the axial and radial regions of each frame image, extract image coordinate points whose edge intensity values ​​are greater than the preset edge extraction threshold, perform structural splitting according to the direction of the coordinates, and obtain a set of direction-separated edge points.

[0020] S203: Call the direction-separated edge point set, classify the edge point data frames according to the angle phase number corresponding to the image, and reorganize the edge point sequence according to the axial position index within each angle phase to establish an initial sequence set of image edge points.

[0021] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0022] S301: Call the initial sequence set of image edge points, perform continuity detection on the image edge point sequences corresponding to all angle phases, construct adjacent phase pairs according to the phase number order, and calculate the position difference of the edge point coordinates of adjacent phases in the radial direction to generate a phase radial displacement sequence.

[0023] S302: Based on the phase radial displacement sequence, compare the radial displacement of the phase pair with the preset radial runout judgment threshold, mark the phase index that exceeds the preset radial runout judgment threshold, and determine the edge point sequence within the corresponding phase as an abnormal point sequence. Perform logical splitting of abnormal phases and normal phases to obtain a runout interference phase identifier set.

[0024] S303: Call the jitter interference phase identifier set, perform a removal operation on the edge point sequence corresponding to the abnormal phase in the initial sequence set of image edge points, and re-aggregate the edge point sequences in the remaining phase according to the phase order to form a data structure consisting only of continuous phase edge points, and filter and establish a continuous image edge point sequence set.

[0025] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0026] S401: Call the continuous image edge point sequence set, compare the coordinate offsets of adjacent edge point columns in the axial and radial directions according to the angle phase number, and adjust the position parameters of the edge point columns in the axial coordinate system according to the comparison results, so that all edge point sequences are aligned with the reference reference in the axial direction, and establish an aligned edge point sequence set.

[0027] S402: Based on the set of aligned edge point sequences, set equally spaced reference positions in the axial direction in each edge point column, extract the coordinates of radial boundary points on both sides of the reference positions, and perform a linear connection operation on the coordinates of adjacent boundary points to construct boundary line segments at the corresponding angles and obtain a set of multi-phase boundary line segments.

[0028] S403: Call the multi-phase boundary line segment set, perform structural combination of boundary line segment data frames under all angle phases in phase number order, and perform coordinate transformation on the combined boundary line segments to construct a contour section wireframe set.

[0029] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0030] S501: Call the set of outline cross-section wireframe diagrams, extract the distance values ​​of coordinate points between boundary line pairs in the radial direction in each cross-section diagram, calculate the coordinate difference values ​​of the upper and lower boundaries at each axial position, and organize all distance data according to the axial sequence to establish a cross-section radial distance sequence set;

[0031] S502: Based on the radial distance sequence set of the cross section, in the boundary wireframe structure corresponding to the angle phase, interpolation reference points are set at uniform intervals in the axial direction, a diameter line is constructed according to the corresponding radial distance value, and the outer boundary line shape graphic is superimposed and drawn in each cross section to obtain the angle phase contour graphic sequence.

[0032] S503: Call the angle phase contour graphic sequence, aggregate all graphic structures according to the phase number order, and perform coordinate transformation and structural reorganization on all angle bitmaps to construct a set of automotive wheel hub measurement contour graphics.

[0033] An automated measurement system for automotive wheel hub dimensions includes:

[0034] The trigger timing generation module is used to implement S1: control the industrial computer to set the light source state and laser exposure cycle, read the reference trigger signal and the equal angle trigger signal, record the response time of the angle phase, arrange the response times in angle order, and construct an optical trigger timing set;

[0035] The edge point extraction module is used to implement S2: calling the optical trigger timing set to control image acquisition, extracting edge points in the axial and radial directions of the image frame, classifying edge points by angle phase number, reorganizing edge points of the same angle phase in axial order, and generating an initial sequence set of image edge points;

[0036] The interference removal module is used to implement S3: call the initial sequence set of image edge points, compare the distribution of edge points in the radial direction of adjacent angular phases, identify and remove jumping interference point sequences, and filter the continuous image edge point sequence set;

[0037] The contour line construction module is used to implement S4: calling the continuous image edge point sequence set, aligning the edge point column based on the angle phase number, setting the axial reference position, extracting boundary points and drawing boundary lines, and combining them to generate a contour section wireframe set;

[0038] The measurement diagram generation module is used to implement S5: call the set of contour section wireframe diagrams, extract the radial distance between boundary lines, draw the diameter line and outer boundary line graphics in the order of angle and phase, and combine them to generate a set of automobile wheel hub measurement contour graphics.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In this invention, by constructing an angle-triggered image acquisition timing sequence, the precise correspondence between edge points and angle information is achieved, ensuring that the edge point sequence has a stable correlation in the axial direction. By using change detection to eliminate jitter interference and filtering continuous edge data, boundary elements are extracted by setting equally spaced reference positions in the image, generating a set of wireframe diagrams with clear structures. By superimposing cross-sectional diagrams, a complete contour graphic is constructed, improving the accuracy and consistency of contour reconstruction, enhancing the ability to recognize complex structures, and improving the stability of measurement data and the anti-interference performance of image processing. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a schematic diagram of the steps of the present invention;

[0043] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0044] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0045] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0046] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0047] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0048] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0049] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0050] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0051] Please see Figure 1 This invention provides an automated method for measuring the dimensions of automotive wheel hubs, comprising the following steps:

[0052] S1: In the automotive wheel hub inspection station, construct the acquisition link for the optical dimension acquisition task. The industrial control computer controls the light source emission state and the exposure cycle of the laser contour acquisition. During one rotation of the wheel hub, read the reference trigger signal and the equal angle trigger signal output by the rotating fixture, record the response time of the wheel hub at each angle phase, arrange all the times in the angle order, and construct the optical trigger timing set.

[0053] S2: Call the optical trigger timing set, control the line laser contour acquisition to acquire contour image frames according to the trigger time, extract edge points in the axial and radial directions in each image frame, classify the edge points by the angle phase number corresponding to the image, and reassemble the edge point sequence of the same phase according to the axial order to generate the initial sequence set of image edge points.

[0054] S3: Call the initial set of image edge points, perform continuity detection on the image edge point sequences corresponding to all angle phases, analyze the distribution of changing positions by comparing the edge point distribution of adjacent phases in the radial direction, identify image point sequences including jitter interference and perform removal, and filter the set of continuous image edge point sequences;

[0055] S4: Call the continuous image edge point sequence set, perform image contour alignment processing based on the angle phase number, set equally spaced reference positions for each group of edge point columns according to the axial direction, extract the boundary points on both sides at each reference position and draw the boundary line structure, combine the boundary line sets generated under all angle phases, and construct the contour section wireframe set.

[0056] S5: Call the contour section wireframe diagram set, extract the radial distance information between the boundary lines in each section diagram, draw the diameter line and the outer boundary line of the contour in sequence according to the angle phase, combine the graphic structures of all angle positions, and construct the automotive wheel hub measurement contour graphic set.

[0057] The optical trigger timing set includes the reference trigger signal time, the equal angle trigger signal time, the angle phase number, and the trigger time arrangement sequence. The image edge point initial sequence set includes the axial edge point set, the radial edge point set, the phase number correspondence, and the edge point classification results. The continuous image edge point sequence set includes radial distribution change points, jump interference marker points, and continuous filtering edge point sequences. The contour section wireframe diagram set includes equally spaced reference points, boundary line node structures, and angle phase correspondence wireframe diagrams. The automobile wheel hub measurement contour graphic set includes diameter distribution line diagrams, contour outer boundary line diagrams, and angle sequence combination graphic structures.

[0058] Please see Figure 2 The specific steps of S1 are as follows:

[0059] S101: Based on the industrial control computer control signal, the exposure period of the light source control signal and the laser profile acquisition is configured with parameters. The exposure duration and emission period parameters in the laser profiler and the light source controller are called to establish the time synchronization relationship between the light source emission and the laser acquisition, and an exposure period synchronization configuration group is generated.

[0060] In the initialization phase of the industrial automation testing process, the industrial control computer (ICC) establishes a bidirectional communication connection with the field-programmable gate array (FPGA) controller via a gigabit Ethernet interface. The ICC sends instructions to the controller to configure the light source control signals and the acquisition parameters of the laser profilometer. This process first defines the strobe mode of the light source and sets the operating mode of the light source controller to external trigger pulse following mode. The ICC then calls a pre-stored configuration file, reading the exposure duration and emission period parameters. For example, the sensor exposure time of the laser profilometer is set to 200 microseconds, and the emission pulse width of the light source controller is set to 250 microseconds to ensure that the light source brightness covers the sensor's integration time. Subsequently, the time synchronization logic is executed, using the crystal oscillator clock inside the controller as the master clock source, sending a synchronization reset signal to the laser profilometer to clear the internal counters of both. Based on this, the minimum trigger interval is calculated according to the preset rotational speed of the rotating fixture, for example, 60 revolutions per minute. If the rotational speed is 60 revolutions per minute, or 1 revolution per second, the corresponding single-cycle period is 1,000,000 microseconds. If the number of sampling points per cycle is 1000, the theoretical trigger interval is 1000 microseconds. The calculated exposure time of 200 microseconds, emission period of 1000 microseconds, and pulse width of 250 microseconds are encapsulated into a configuration data packet. This data packet is written to the underlying register via a communication protocol, completing the generation of the exposure period synchronization configuration group. This process ensures that the light source and camera can synchronize with microsecond-level precision each time a physical trigger signal is generated, eliminating the phenomenon of missed shots or uneven brightness caused by communication delays.

[0061] As shown in Table 1, detailed synchronization configuration parameters are recorded, which serve as baseline data for subsequent steps.

[0062] Table 1. Exposure Synchronization Configuration Parameters

[0063]

[0064] Table 1 lists the key parameter settings for synchronization control in the embodiments. For the above logic, to ensure the light source brightness fully covers the exposure cycle of the image sensor, the light source activation time needs to be set earlier than the image exposure start time (exposure pulse width minus exposure duration plus synchronization compensation time). Taking an exposure duration of 200 microseconds, a light source pulse width of 250 microseconds, and synchronization compensation of 5 microseconds as an example, the calculated advance of the light source activation is (250−200)+5 = 55 microseconds, meaning the light source should be activated 55 microseconds before the start of exposure to ensure continuous illumination throughout the sensor's exposure period. Subsequently, the ratio of the light source pulse width of 250 microseconds to the minimum trigger cycle of 1000 microseconds is calculated, yielding a duty cycle parameter of 0.25 for the light source. The advantage of this calculation logic is that by accurately calculating the duty cycle and advance, it effectively avoids the ghosting problem of the stroboscopic light source in high-speed rotation detection, while also reducing the thermal load of the LED light source and extending the equipment's lifespan.

[0065] S102: Based on the exposure cycle synchronization configuration group, the reference trigger signal and the constant angle trigger signal output by the rotating fixture are collected, the time index representing the hub angle is extracted, and all constant angle trigger times are normalized and rearranged to generate the hub constant angle response time series.

[0066] The motion control module of the rotary fixture is activated, and the pulse signals fed back from the encoder are acquired in real time through a high-speed input / output interface. The rotary fixture is equipped with a high-resolution incremental encoder, which outputs a reference trigger signal (Z-phase pulse) and several equal-angle trigger signals (A-phase pulses) per revolution. An internal high-speed counter records the arrival time of each pulse in microseconds. When the rising edge of the reference trigger signal is detected, the current timestamp is marked as the zero-point index, and timing capture of subsequent equal-angle trigger signals is initiated. Assuming the encoder resolution is 3600 pulses per revolution, 3600 timestamp data will be continuously captured within one rotation cycle. These time data are extracted and stored in a double-precision floating-point array to form the original time series. Next, a normalization rearrangement operation is performed to eliminate nonlinear errors caused by motor speed fluctuations. This process first obtains the total duration of the rotation cycle, i.e., the time difference between two adjacent reference trigger signals, for example, 1000500 microseconds. Then, the relative timestamp of each equal-angle trigger signal is divided by this total duration to obtain a normalized time coefficient between 0 and 1. Based on the magnitude of these coefficients, all acquired trigger time points are reordered to ensure that the time index is strictly monotonically increasing, thereby generating a time series of the wheel hub's angular response. This series accurately reflects the absolute time corresponding to every minute angular position of the wheel hub during rotation, providing a high-precision temporal reference for subsequent image acquisition.

[0067] For the normalization calculation logic, the relative timestamp value of the 500th trigger signal (138,900 microseconds) and the total duration of the current rotation cycle (1,000,000 microseconds) are obtained. A division operation is performed, dividing 138,900 by 1,000,000 to obtain a normalized time coefficient of 0.1389. This coefficient is then multiplied by 360 degrees to obtain the corresponding physical angle of 50.004 degrees. The advantage of this calculation logic is that the normalization process automatically compensates for the minute speed fluctuations of the motor during a single rotation, ensuring the accuracy of the angle calculation.

[0068] S103: Call the time series of angle response of the wheel hub, sort it in order according to the angle position of the time index within the rotation cycle, perform linear fitting processing, obtain the time mapping point corresponding to the angle, and establish an optical trigger timing set.

[0069] The system retrieves the hub angle response time series from memory, which contains normalized time point data. Based on the time index, these time points are mapped to an angle space of 0 to 360 degrees within the rotation period. Due to backlash and elastic deformation in mechanical transmission, directly mapped time points may exhibit local nonlinearity. Therefore, linear fitting is performed to optimize this mapping relationship. Ten consecutive time points are selected as a sliding window, and the least squares method is used to perform regression analysis on the time-angle relationship within the window. An objective function is constructed, which is the sum of squared distances from each data point to the fitted line. The slope and intercept parameters that minimize the objective function are calculated by differentiation. Using the fitted linear equation, the theoretically precise time point corresponding to each angular position is recalculated; these points are the time mapping points corresponding to the angle. These corrected time points are arranged sequentially to establish an optical triggering time series set. This set defines the exact time when the laser profilometer should acquire data at each predetermined angle, eliminating the influence of mechanical jitter on the imaging geometry.

[0070] In a specific example of linear fitting, a set of data points is selected, where the time variables are 10,000 microseconds, 20,000 microseconds, and 30,000 microseconds, and the corresponding angle variables are 3.6 degrees, 7.2 degrees, and 10.8 degrees, respectively. First, the average value of the time variable (20,000 microseconds) and the average value of the angle variable (7.2 degrees) are calculated. Next, the covariance of the time and angle variables, as well as the variance of the time variable, are calculated. By dividing the covariance by the variance, the slope parameter of the fitted line is obtained as 0.00036 degrees per microsecond. Subsequently, the intercept parameter is calculated using the average value and the slope: 7.2 − 0.00036 × 20,000 = 0. Finally, the angle corresponding to the time point 40,000 microseconds is predicted using this linear model: 0.00036 × 40,000 + 0 = 14.4 degrees. The advantage of this operational logic is that it smooths high-frequency noise through local linearization, making the triggering timing more uniform and reliable.

[0071] Please see Figure 3 The specific steps of S2 are as follows:

[0072] S201: Call the optical trigger timing set, control the line laser profilometer to acquire contour image frames according to the trigger time, synchronously start image acquisition for each trigger time point, and number and store the acquired image frames according to the trigger sequence number to establish a contour image frame sequence set;

[0073] A precise optical trigger timing sequence set contains a series of trigger timestamps accurate to the microsecond level. Based on these timestamps, the controller sends trigger pulses to the line laser profilometer via hardware interrupts. Whenever the clock matches the trigger moment, the laser profilometer immediately activates the CMOS sensor for exposure. For example, at the first trigger moment, the sensor turns on, integrates for 200 microseconds, and then turns off, converting the optical signal into an electrical signal and reading it out as a single digital image frame. This image captures the cross-sectional shape of the laser line projected onto the wheel hub surface. This process is repeated sequentially for each trigger time point, for example, acquiring 3600 images during one rotation. Each acquired image frame is assigned a unique trigger sequence number, ranging from 1 to 3600, and the image data is stored in a high-speed cache as a two-dimensional array. These image frames are strictly arranged according to their trigger sequence numbers, collectively forming a sequence set of profilometer image frames. Each image frame not only contains height information of the profilometer but also implicitly contains reflectivity information at that moment, providing raw material for subsequent feature extraction.

[0074] To verify the effectiveness of the acquisition logic, the trigger signal and image transmission status are monitored. Assuming the current trigger sequence number is 500, the time the trigger signal is emitted is recorded as a timestamp of 500500 microseconds, and the time the image acquisition is completed and transmitted to memory is 502000 microseconds. The time difference between the two is calculated as 1500 microseconds and compared with a preset timeout threshold of 2000 microseconds. Since 1500 is less than 2000, the frame acquisition is considered valid. If the difference exceeds the threshold, the frame is marked as a lost frame, and an interpolation compensation procedure is initiated.

[0075] S202: Based on the contour image frame sequence set, detect the gray-level gradient changes in the axial and radial regions of each frame image, extract image coordinate points with edge intensity values ​​greater than the preset edge extraction threshold, perform structural decomposition according to the direction of the coordinates, and obtain a set of direction-separated edge points.

[0076] Parallel processing is performed on each frame of the contour image frame sequence. For a single frame, a region of interest is defined, covering the entire effective range in both the axial and radial directions. The Sobel or Canny operator is applied to perform convolution operations on the image, calculating the gray-level gradient magnitude of each pixel. Specifically, the gray-level differences in the horizontal and vertical directions are calculated separately, and the square root of the sum of their squares is used to obtain the gradient intensity. A preset edge extraction threshold, such as a gray-level value of 50, is set. The entire image is traversed, and pixel coordinates with gradient intensities greater than or equal to 50 are marked as potential edge points. Subsequently, structural decomposition is performed based on the geometric location of the coordinates. For example, points with Y coordinates less than half the image height are classified as outer edges, and points greater than half are classified as inner edges, thus obtaining a set of directionally separated edge points. This step effectively separates the complex light stripes formed by the laser line on the wheel hub surface into independent geometric feature lines.

[0077] As shown in Table 2, edge extraction parameters for different material surfaces were set to adapt to different reflection characteristics.

[0078] Table 2 Edge Detection Parameter Configuration Table

[0079]

[0080] Table 2 provides the specific threshold settings for edge detection. In the specific calculation, the horizontal gradient value of 30 and the vertical gradient value of 40 for a given pixel are read. A square root operation is performed, i.e., the square of 30 plus the square of 40 equals 2500, and the square root yields a gradient magnitude of 50. The calculated result 50 is compared with the edge extraction threshold 50 to determine if the point is a valid edge point, and its coordinates are recorded. The advantage of this calculation logic is that it can robustly identify the laser centerline under weak contrast by utilizing the gradient magnitude, overcoming ambient light interference.

[0081] S203: Call the direction separation edge point set, classify the edge point data frame according to the angle phase number corresponding to the image, and reorganize the edge point sequence according to the axial position index in each angle phase to establish the initial sequence set of image edge points;

[0082] The system calls upon a split set of directional edge points, containing tens of thousands of discrete coordinate points. To construct an ordered 3D model, these points need to be structurally reorganized. First, based on the trigger number corresponding to each frame of the image, the point is mapped back to its corresponding angle phase number; for example, frame 1 corresponds to 0.1 degrees, frame 2 to 0.2 degrees. Within each specific angle phase, the radial edge point data (column coordinates) is further sorted in ascending order according to the axial position index (row coordinates). If multiple edge points exist at the same angle and axial position, a median filtering strategy is used to retain the most reliable point. Through this dual indexing mechanism, a three-dimensional array structure is established, with dimensions of angle phase, axial position, and radial position, which constitutes the initial sequence set of image edge points. This set transforms the chaotic pixels into regular grid data, laying the topological foundation for subsequent contour analysis.

[0083] In the reassembly logic, the data frame with angle phase number 100 is processed. This frame contains two edge points with coordinates (row 200, column 500) and (row 201, column 502). The axial position indices 200 and 201 are checked to confirm their continuity. If an axial index is missing, for example, only 200 and 202 are found, a linear interpolation operation is performed. The column coordinate 500 of row 200 and the column coordinate 504 of row 202 are read, their average 502 is calculated, and (row 201, column 502) is inserted into the sequence. The advantage of this operation logic is that it fills in local data gaps caused by surface high reflectivity or occlusion, ensuring the continuity of the contour curve.

[0084] Please see Figure 4 The specific steps of S3 are as follows:

[0085] S301: Call the initial sequence set of image edge points, perform continuity detection on the image edge point sequence corresponding to all angle phases, construct adjacent phase pairs according to the phase number order, and calculate the position difference of the edge point coordinates of adjacent phases in the radial direction to generate a phase radial displacement sequence.

[0086] Continuity detection is performed on an initial set of image edge points to identify abrupt noise caused by mechanical vibration or electrical interference. Adjacent phase pairs are constructed sequentially according to phase numbering, for example, from phase 1 to phase 3600, i.e., phase k and phase (k+1). For each phase pair, the radial coordinates of its edge points at the same axial position are extracted. The algebraic difference between these two coordinates is calculated to obtain the radial displacement. For example, if the radial coordinate of phase k at axial position 100 is 50.5 mm, and the coordinate of phase (k+1) at the same position is 50.6 mm, then the displacement is 0.1 mm. This calculation is repeated for all axial positions to generate a sequence of radial displacements corresponding to each phase pair. Finally, the calculation results for all phase pairs are summarized to generate a phase radial displacement sequence covering the entire rotation cycle. This sequence visually reflects the microscopic undulations and macroscopic vibrations of the hub surface in the circumferential direction.

[0087] In a specific example of displacement calculation, adjacent phase pairs A and B are selected. The radial coordinate value of phase A at a certain axial point is 150.25 mm, and the coordinate value of phase B at the same point is 150.35 mm. A subtraction operation is performed, calculating 150.35 minus 150.25, yielding a displacement of ±0.10 mm. Simultaneously, the rate of change at this point is calculated; assuming the time interval between the two phases is 1 millisecond, the rate of change is 0.10 mm per millisecond. The advantage of this operational logic is that it sensitively captures high-frequency abrupt changes in surface signals through differential operations.

[0088] S302: Based on the phase radial displacement sequence, compare the radial displacement of the phase pair with the preset radial runout judgment threshold, mark the phase index that exceeds the preset radial runout judgment threshold, and determine the edge point sequence within the corresponding phase as an abnormal point sequence. Perform logical splitting of abnormal phases and normal phases to obtain the runout interference phase identifier set.

[0089] Based on the generated radial displacement sequence, abnormal phases are screened and marked. A radial runout threshold is preset, set according to the tolerance range of the wheel hub machining accuracy, for example, 0.5 mm. Each displacement value in the radial displacement sequence is traversed, and its absolute value is compared with the threshold. If the absolute value of the displacement of a phase pair exceeds 0.5 mm, it is determined that abnormal and severe runout has occurred at that position, possibly caused by loose clamps, wheel hub burrs, or foreign object attachment. This phase index is marked as abnormal, and the corresponding entire column of edge points within that phase is also classified as an abnormal point column. Subsequently, a logical splitting operation is performed to separate all marked abnormal phase indices and store them in a runout interference phase identifier set, while unmarked phases are retained as normal phases.

[0090] Table 3 shows the displacement detection results and judgment status of some phase pairs.

[0091] Table 3. Radial runout detection results

[0092]

[0093] Table 3 illustrates the anomaly screening process. Based on the data in the table, the radial displacement value of phase pair 102 to 103 is read as 0.65 mm, and the judgment threshold is 0.50 mm. A comparison operation is performed, determining that 0.65 is greater than 0.50, thus the condition is met. A marking action is then triggered, recording index 103 into the interference set. Simultaneously, the percentage of abnormal phases is calculated. If the total number of detected phases is 3600 and the number of abnormal phases is 36, the anomaly rate is 1%. The advantage of this operational logic is that it achieves automated initial quality screening, enabling rapid location of process deviations on the production line.

[0094] S303: Call the jitter interference phase identifier set, perform a removal operation on the edge point sequence corresponding to the abnormal phase in the initial sequence set of image edge points, and re-aggregate the edge point sequences in the remaining phase according to the phase order to form a data structure consisting only of continuous phase edge points, and filter and establish a continuous image edge point sequence set.

[0095] The system invokes the aberration interference phase identifier set and returns to the initial sequence set of image edge points to perform a cleaning operation. For each aberration phase index recorded in the identifier set, the corresponding edge point data frame is directly removed, set to empty, or marked as invalid. After the removal operation, discontinuous gaps will appear in the data sequence. To restore the topological continuity of the data, the edge point sequences within the remaining normal phases are re-aggregated. Following the original phase order, the scattered data frames are tightly arranged, eliminating index jumps and forming a new data structure consisting only of continuous and smooth phase edge points. Finally, a continuous image edge point sequence set is established. This set represents the clean contour data after removing high-frequency noise and outliers, realistically restoring the main geometric features of the contour.

[0096] In the data re-aggregation example, the original sequence is assumed to contain phases 1, 2, 3, 4, and 5, with phase 3 marked as an anomaly. After removal, the sequence becomes 1, 2, 4, and 5. The data frame of phase 4 is read, and its logical index is remapped to 3. The data frame of phase 5 is read and remapped to 4. Simultaneously, the spatial distance between the end of phase 2 and the beginning of the original phase 4 is calculated. Assuming the coordinates of the end of phase 2 are (100, 50.0) and the coordinates of the beginning of the original phase 4 are (102, 50.1), the Euclidean distance is calculated using the Pythagorean theorem: the square of (102 - 100) plus the square of (50.1 - 50.0). The square root of the result is approximately 2.002. If this distance is less than a preset continuity threshold of 5.0, the boundary smoothness after re-aggregation is confirmed. The advantage of this operational logic is that it retains effective geometric information to the greatest extent while removing noise, ensuring the realism of the model reconstruction.

[0097] Please see Figure 5 The specific steps of S4 are as follows:

[0098] S401: Call the continuous image edge point sequence set, compare the coordinate offsets of adjacent edge point columns in the axial and radial directions according to the angle phase number, and adjust the position parameters of the edge point columns in the axial coordinate system according to the comparison results, so that all edge point sequences are aligned with the reference datum in the axial direction, and establish an aligned edge point sequence set.

[0099] A continuous set of image edge point sequences is invoked to begin fine-grained geometric correction. Due to potential slight axial movement of the rotating fixture, the contours at different angles may not be perfectly aligned axially. The coordinate offsets of adjacent edge point sequences in the axial and radial directions are compared one by one according to the angle phase number. A reference phase, usually the starting phase, is selected as the reference coordinate system. For each subsequent phase, its axial distance from the previous phase at a feature point (such as the inflection point of a wheel hub flange) is calculated. Based on the comparison results, an axial correction vector is calculated and applied to all coordinate points of that edge point sequence, adjusting its position parameters in the axial coordinate system. This process ensures that the entire edge point sequence remains aligned with the reference reference in the axial direction, eliminating the "snake-like" error caused by mechanical movement and establishing an aligned edge point sequence set.

[0100] In the alignment calculation, phases K and K+1 are selected. The axial coordinate value of the rim inflection point in phase K is identified as 200.5 mm, and the corresponding axial coordinate value in phase K+1 is 200.8 mm. A subtraction operation is performed, calculating 200.5 minus 200.8 to obtain an offset of -0.3 mm. It is determined that a positive translation operation needs to be performed on all data points in phase K+1, with a translation amount of 0.3 mm. The original axial coordinate of a point in phase K+1 is read as 300.0 mm, and the translation amount of 0.3 mm is added to obtain the corrected coordinate of 300.3 mm. The advantage of this calculation logic is that it compensates for the insufficient mechanical precision of the hardware fixture through software algorithms, significantly reducing the cost of relying on precision machinery.

[0101] S402: Based on the set of aligned edge point sequences, set equally spaced reference positions in the axial direction in each edge point column, extract the coordinates of radial boundary points on both sides of the reference positions, and perform a linear connection operation on the coordinates of adjacent boundary points to construct boundary line segments at the corresponding angles and obtain a set of multi-phase boundary line segments.

[0102] Align the set of edge point sequences and set equally spaced reference positions along the axial direction in each edge point sequence. For example, define a sampling slice every 0.1 mm along the axial direction. At each reference position, extract the coordinates of the nearest radial boundary points on both sides of that position. Since the actual acquired point cloud may be discrete, use linear interpolation to determine the precise radial values ​​at the reference positions. Subsequently, perform a linear connection operation on the coordinates of adjacent boundary points, i.e., connect adjacent sampling points with straight line segments to construct continuous boundary line segments at the corresponding angles. Repeat this step for all angular phases to obtain a set of multi-phase boundary line segments. This set transforms the discrete point cloud into a series of continuous vector contour lines, facilitating subsequent geometric measurements.

[0103] In the example of constructing boundary line segments, the axial reference position X is set to 150.0 mm. The nearest point A (149.8, 60.2) with X less than 150.0 and the nearest point B (150.2, 60.4) with X greater than 150.0 are found in the dataset. The axial distance between A and B is calculated to be 0.4 mm, and the radial distance is 0.2 mm. The axial offset of the reference position 150.0 relative to point A is calculated to be 0.2 mm. Using the proportional relationship, the radial increment is calculated as (0.2 ÷ 0.4) × 0.2 = 0.1 mm. Adding the increment 0.1 to the radial coordinate of point A (60.2) yields the interpolated radial coordinate of 60.3 mm at the reference position. The advantage of this operational logic is that it achieves sub-pixel-level contour reconstruction, making the measurement results unrestricted by sampling resolution.

[0104] S403: Call the multi-phase boundary line segment set, perform structural combination of all boundary line segment data frames under all angle phases in phase number order, and perform coordinate transformation on the combined boundary line segments to construct a set of contour section wireframe diagrams;

[0105] The multi-phase boundary segment set is invoked, and the boundary segment data frames under all angular phases are structurally combined according to their phase numbers. These segments form a cylindrical mesh structure in space. To unify the measurement standards, the combined boundary segments undergo a coordinate transformation, converting the relative sensor coordinates to the wheel hub's own cylindrical coordinate system. Using a calibration matrix, the (x, z) sensor coordinates of each point are converted to (r, h, theta) cylindrical coordinates, where r represents the radius, h represents the height, and theta represents the angle. Through this transformation, a set of profile cross-section wireframe diagrams is constructed. This set completely describes the geometric skeleton of the wheel hub in three-dimensional space, providing a fully digital twin model for the final dimensional tolerance evaluation.

[0106] In the specific calculations of coordinate transformation, the offset of the sensor coordinate system origin relative to the hub rotation center is set to (dx = 500 mm, dz = 100 mm). For the sensor coordinates of a certain point (x = 20 mm, z = 50 mm), its absolute radius r is calculated. An addition operation is performed: r equals dx plus x, i.e., 500 plus 20 equals 520 mm. The absolute height h is calculated: h equals dz plus z, i.e., 100 plus 50 equals 150 mm. If the angle theta corresponding to the current phase is 90 degrees, then the coordinates of this point in the cylindrical coordinate system are (520, 150, 90). The advantage of this calculation logic is that it unifies local measurement data to a global reference, allowing for direct comparison and evaluation of geometric features at different locations.

[0107] Please see Figure 6 The specific steps of S5 are as follows:

[0108] S501: Call the contour section wireframe diagram set, extract the distance values ​​of the coordinate points between the boundary line pairs in the radial direction in each section diagram, calculate the coordinate difference values ​​of the upper and lower boundaries at each axial position, and organize all distance data according to the axial sequence to establish a section radial distance sequence set;

[0109] The set of profile section wireframe diagrams is invoked to enter the dimensional analysis stage. In each section diagram, the radial distance values ​​of the coordinate points between boundary line pairs are extracted sequentially. Specifically, for features such as the rim thickness or spoke width of the wheel hub, its upper and lower boundaries are identified. The coordinate difference between the upper and lower boundaries at each axial position is calculated. For example, at axial position h, if the upper boundary radius is r_outer and the lower boundary radius is r_inner, then the radial distance is r_outer minus r_inner. All distance data are organized according to the axial sequence to establish a set of radial distance sequences for the sections. This set of sequences quantifies the solid thickness distribution of various parts of the wheel hub.

[0110] In the distance calculation example, a section at an axial position h = 50 mm is selected. The radial coordinates of the upper boundary point are measured to be 225.50 mm, and the radial coordinates of the lower boundary point are 215.50 mm. A subtraction operation is performed, subtracting 215.50 from 225.50, yielding a radial distance of 10.00 mm. To verify compliance, a standard value of 10.00 mm and a tolerance range of ±0.10 mm are introduced. The absolute value of the difference between the measured value and the standard value is calculated to be 0.00 mm, indicating that the dimension at that location is acceptable. The advantage of this calculation logic is that it enables full-section wall thickness detection, allowing for the discovery of shrinkage cavities or uneven wall thickness defects during the casting process.

[0111] S502: Based on the cross-sectional radial distance sequence set, in the boundary wireframe structure corresponding to the angle phase, interpolation reference points are set at uniform intervals in the axial direction, a diameter line is constructed according to the corresponding radial distance value, and the outer boundary line shape graphic is superimposed and drawn in each cross-sectional view to obtain the angle phase contour graphic sequence;

[0112] Based on the radial distance sequence set of the cross-section, a visualized contour graphic is further constructed. In the boundary wireframe structure corresponding to the angular phase, interpolation reference points are set at uniform intervals along the axial direction, for example, one point per millimeter. According to the corresponding radial distance value, a diameter line is constructed on the screen; that is, a circle or arc is drawn with the rotation center as the center and half the radial distance as the radius. The outer boundary line shape graphic is overlaid on each cross-sectional view, connecting the data points into a smooth curve. In this way, an angular phase contour graphic sequence is obtained. These graphics not only contain numerical information but also intuitively display the contour features of the wheel hub with geometric shapes, facilitating visual inspection by operators.

[0113] In the logic of graphic construction, the radial distance value of a certain point is processed as 450.0 mm. A division operation is performed, dividing 450.0 by 2, to obtain the radius value of 225.0 mm. Using the drawing engine, a circle with a radius of 225.0 pixels is drawn at the center coordinates (1000, 1000) of the canvas. At the same time, the radius of the adjacent point, 225.1 mm, is read, and a line segment is drawn between the two points. The slope of the line segment is calculated as (225.1 - 225.0) ÷ 1.0 = 0.1. The advantage of this operation logic is that it transforms abstract data into an intuitive engineering drawing view, helping engineers quickly understand deformation trends.

[0114] S503: Call the sequence of angle phase contour graphics, aggregate all graphic structures according to the phase number order, and perform coordinate transformation and structural reorganization on all angle bitmaps to construct a set of automotive wheel hub measurement contour graphics;

[0115] The process involves calling the sequence of angle phase contour graphics and performing the final 3D compositing. Based on the phase numbering order, all graphic structures are aggregated. Two-dimensional cross-sectional graphics are rotated and arranged around the central axis, filling the gaps to form closed curved surfaces. All angle bitmaps undergo coordinate transformation and structural reorganization, and using OpenGL or DirectX rendering technology, a set of automotive wheel hub measurement contour graphics is constructed. This set is a complete high-fidelity 3D model, which users can rotate, scale, and section on the interface to comprehensively examine the wheel hub's machining quality.

[0116] In the final example of data aggregation, the number of vertices in the entire model was counted. Assuming 3600 phases per cycle, 1000 points per phase, and a total of 3.6 million points, the memory usage of each point was calculated. Assuming each point is represented by 3 floating-point numbers (12 bytes), the total data volume is approximately 43.2 megabytes. This 43.2 megabytes of data was loaded into video memory, and lighting was applied to render the model. The model's surface area was calculated by summing the areas of all the tiny triangular facets. Assuming the calculated total surface area is 1.5 square meters, it was compared with the theoretical surface area of ​​1.48 square meters from the CAD design model. The deviation rate was calculated as (1.5 - 1.48) divided by 1.48, approximately 1.35%. The advantage of this calculation logic is that it provides a macro-level quality evaluation indicator, making the consistency assessment of the product more objective and scientific.

[0117] Please see Figure 7 An automated measurement system for automotive wheel hub dimensions includes:

[0118] The trigger timing generation module is used to implement S1: control the industrial computer to set the light source state and laser exposure cycle, read the reference trigger signal and the equal angle trigger signal, record the response time of the angle phase, arrange the response times in angle order, and construct an optical trigger timing set;

[0119] The edge point extraction module is used to implement S2: call the optical trigger timing set to control image acquisition, extract edge points in the axial and radial directions of the image frame, classify edge points according to angle phase number, reorganize edge points with the same angle phase in axial order, and generate an initial sequence set of image edge points;

[0120] The interference removal module is used to implement S3: call the initial sequence set of image edge points, compare the distribution of edge points in the radial direction of adjacent angular phases, identify and remove jumping interference point sequences, and filter the continuous image edge point sequence set;

[0121] The contour construction module is used to implement S4: call the continuous image edge point sequence set, align the edge point column based on the angle phase number, set the axial reference position, extract boundary points and draw boundary lines, and combine them to generate a set of contour section wireframe diagrams;

[0122] The measurement diagram generation module is used to implement S5: call the contour section wireframe diagram set, extract the radial distance between boundary lines, draw the diameter line and outer boundary line graphics in the order of angle and phase, and combine them to generate a set of automobile wheel hub measurement contour graphics.

[0123] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the technical solution.

Claims

1. An automated measurement method for the dimensions of automobile wheel hubs, characterized in that, Includes the following steps: S1: Control the industrial computer to set the light source status and laser exposure cycle, read the reference trigger signal and the equal angle trigger signal, record the response time of the angle phase, arrange the response times in angle order, and construct an optical trigger timing set; S2: Call the optical trigger timing set to control image acquisition, extract edge points in the axial and radial directions of the image frame, classify the edge points according to the angle phase number, reorganize the edge points of the same angle phase in axial order, and generate an initial sequence set of image edge points. S3: Call the initial sequence set of image edge points, compare the distribution of edge points in the radial direction of adjacent angular phases, identify and remove jittery interference point sequences, and filter the continuous image edge point sequence set; S4: Call the continuous image edge point sequence set, align the edge point column based on the angle phase number, set the axial reference position, extract the boundary points and draw the boundary lines, and combine them to generate a set of contour section wireframe diagrams; S5: Call the set of contour section wireframe diagrams, extract the radial distance between boundary lines, draw the diameter line and outer boundary line graphics in the order of angle and phase, and combine them to generate a set of automotive wheel hub measurement contour graphics.

2. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The optical trigger timing set includes the reference trigger signal time, the equal angle trigger signal time, the angle phase number, and the trigger time arrangement sequence. The image edge point initial sequence set includes the axial edge point set, the radial edge point set, the phase number correspondence, and the edge point classification result. The continuous image edge point sequence set includes radial distribution change points, jump interference marker points, and continuous filtering edge point sequences. The contour cross-section wireframe diagram set includes equally spaced reference points, boundary line node structures, and angle phase corresponding wireframe diagrams. The automobile wheel hub measurement contour graphic set includes diameter distribution line diagrams, contour outer boundary line diagrams, and angle sequence combination graphic structures.

3. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The process of identifying and removing interfering points involves comparing the distribution of edge points in the radial direction of adjacent angular phases, identifying changing point sequences, and removing them from the edge point sequence.

4. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The classification of edge points by angle and phase number refers to classifying and organizing the extracted edge points according to their corresponding angle and phase numbers.

5. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Based on the industrial control computer control signal, the exposure period of the light source control signal and the laser profile acquisition is configured with parameters. The exposure duration and emission period parameters in the laser profiler and the light source controller are called to establish the time synchronization relationship between the light source emission and the laser acquisition, and an exposure period synchronization configuration group is generated. S102: Based on the exposure cycle synchronization configuration group, collect the reference trigger signal and the constant angle trigger signal output by the rotating fixture, extract the time index representing the wheel hub angle, and normalize and rearrange all constant angle trigger times to generate the wheel hub constant angle response time sequence. S103: Call the wheel hub angle response time series, sort them sequentially according to the angle position within the rotation cycle based on the time index, perform linear fitting processing, obtain the time mapping point corresponding to the angle, and establish an optical trigger timing set.

6. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Call the optical trigger timing set, control the line laser profilometer to acquire contour image frames according to the trigger time, synchronously start image acquisition for each trigger time point, and number and store the acquired image frames according to the trigger sequence number to establish a contour image frame sequence set; S202: Based on the contour image frame sequence set, detect the gray-level gradient changes in the axial and radial regions of each frame image, extract image coordinate points whose edge intensity values ​​are greater than the preset edge extraction threshold, perform structural splitting according to the direction of the coordinates, and obtain a set of direction-separated edge points. S203: Call the direction-separated edge point set, classify the edge point data frames according to the angle phase number corresponding to the image, and reorganize the edge point sequence according to the axial position index within each angle phase to establish an initial sequence set of image edge points.

7. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Call the initial sequence set of image edge points, perform continuity detection on the image edge point sequences corresponding to all angle phases, construct adjacent phase pairs according to the phase number order, and calculate the position difference of the edge point coordinates of adjacent phases in the radial direction to generate a phase radial displacement sequence. S302: Based on the phase radial displacement sequence, compare the radial displacement of the phase pair with the preset radial runout judgment threshold, mark the phase index that exceeds the preset radial runout judgment threshold, and determine the edge point sequence within the corresponding phase as an abnormal point sequence. Perform logical splitting of abnormal phases and normal phases to obtain a runout interference phase identifier set. S303: Call the jitter interference phase identifier set, perform a removal operation on the edge point sequence corresponding to the abnormal phase in the initial sequence set of image edge points, and re-aggregate the edge point sequences in the remaining phase according to the phase order to form a data structure consisting only of continuous phase edge points, and filter and establish a continuous image edge point sequence set.

8. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Call the continuous image edge point sequence set, compare the coordinate offsets of adjacent edge point columns in the axial and radial directions according to the angle phase number, and adjust the position parameters of the edge point columns in the axial coordinate system according to the comparison results, so that all edge point sequences are aligned with the reference reference in the axial direction, and establish an aligned edge point sequence set. S402: Based on the set of aligned edge point sequences, set equally spaced reference positions in the axial direction in each edge point column, extract the coordinates of radial boundary points on both sides of the reference positions, and perform a linear connection operation on the coordinates of adjacent boundary points to construct boundary line segments at the corresponding angles and obtain a set of multi-phase boundary line segments. S403: Call the multi-phase boundary line segment set, perform structural combination of boundary line segment data frames under all angle phases in phase number order, and perform coordinate transformation on the combined boundary line segments to construct a contour section wireframe set.

9. The automated measurement method for the dimensions of automobile wheel hubs according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Call the set of outline cross-section wireframe diagrams, extract the distance values ​​of coordinate points between boundary line pairs in the radial direction in each cross-section diagram, calculate the coordinate difference values ​​of the upper and lower boundaries at each axial position, and organize all distance data according to the axial sequence to establish a cross-section radial distance sequence set; S502: Based on the radial distance sequence set of the cross section, in the boundary wireframe structure corresponding to the angle phase, interpolation reference points are set at uniform intervals in the axial direction, a diameter line is constructed according to the corresponding radial distance value, and the outer boundary line shape graphic is superimposed and drawn in each cross section to obtain the angle phase contour graphic sequence. S503: Call the angle phase contour graphic sequence, aggregate all graphic structures according to the phase number order, and perform coordinate transformation and structural reorganization on all angle bitmaps to construct a set of automotive wheel hub measurement contour graphics.

10. An automated measurement system for the dimensions of automobile wheel hubs, characterized in that, The system is used to implement the automated measurement method for the dimensions of automobile wheel hubs according to any one of claims 1-9, and the system includes: The trigger timing generation module is used to implement S1: control the industrial computer to set the light source state and laser exposure cycle, read the reference trigger signal and the equal angle trigger signal, record the response time of the angle phase, arrange the response times in angle order, and construct an optical trigger timing set; The edge point extraction module is used to implement S2: calling the optical trigger timing set to control image acquisition, extracting edge points in the axial and radial directions of the image frame, classifying edge points by angle phase number, reorganizing edge points of the same angle phase in axial order, and generating an initial sequence set of image edge points; The interference removal module is used to implement S3: call the initial sequence set of image edge points, compare the distribution of edge points in the radial direction of adjacent angular phases, identify and remove jumping interference point sequences, and filter the continuous image edge point sequence set; The contour line construction module is used to implement S4: calling the continuous image edge point sequence set, aligning the edge point column based on the angle phase number, setting the axial reference position, extracting boundary points and drawing boundary lines, and combining them to generate a contour section wireframe set; The measurement diagram generation module is used to implement S5: call the set of contour section wireframe diagrams, extract the radial distance between boundary lines, draw the diameter line and outer boundary line graphics in the order of angle and phase, and combine them to generate a set of automobile wheel hub measurement contour graphics.

Images