Wheel hub blank detection method and system

Abstract

The application provides a wheel hub blank detection method and system, and relates to the technical field of industrial detection. The method is provided with multiple specific identification stations in a wheel hub blank detection device and is automatically controlled by using a roller bed. The wheel type identification, wheel width identification and air valve hole angle extraction of the wheel hub blank can be completed in the first identification station equipped with a visual camera. The appearance reconstruction and size detection of the wheel hub blank can be completed in the second identification station equipped with a laser profiler. The clamping point acquisition and laser marking process can be completed in the third identification station equipped with a laser marker. Therefore, the automatic detection of the wheel hub blank is efficiently realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial detection, in particular to a wheel hub blank detection method and system. BACKGROUND

[0002] In the production and processing process of the wheel hub blank, the appearance of the blank after pressure casting and heat treatment needs to be accurately and effectively detected and monitored, and the detection result is fed back to the machining sequence in time when the detection result is abnormal, so as to adjust the subsequent wheel hub blank production and processing process, thereby reducing the production loss as much as possible. It can be seen that how to quickly and effectively detect the wheel hub blank is crucial in the wheel hub production process.

[0003] Due to the large number of wheel hub blank models and complex structure, the existing scheme is mostly local detection, lacks efficient and unified control logic, and is difficult to adapt to the integrated detection needs of multiple types of wheel hubs. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a wheel hub blank detection method and system, which sets multiple specific identification stations in the wheel hub blank detection equipment and uses a roller bed for automatic control, can complete wheel type identification, wheel width identification and valve hole angle extraction of the wheel hub blank in the identification station equipped with a vision camera, can complete appearance reconstruction and size detection of the wheel hub blank in the identification station equipped with a laser profiler, and can complete the acquisition of the clamping point and the laser marking process in the identification station equipped with a laser marker, thereby efficiently realizing the automatic detection of the wheel hub blank.

[0005] In a first aspect, the present application provides a wheel hub blank detection method, which is applied to a wheel hub blank detection equipment; wherein the wheel hub blank detection equipment transmits the wheel hub blank to be detected to multiple identification stations in sequence through a roller bed for detection;

[0006] The method comprises:

[0007] The first identification station identification step: after transmitting the wheel hub blank to the first identification station by the roller bed, the vision camera in the wheel hub blank detection equipment is controlled to collect the digital image corresponding to the wheel hub blank, and the first detection result of the wheel hub blank is determined by using the digital image;

[0008] The second identification station identification step: after transmitting the wheel hub blank to the second identification station by the roller bed, the laser profiler in the wheel hub blank detection equipment is controlled to generate the reconstructed point cloud corresponding to the wheel hub blank, and the second detection result corresponding to the wheel hub blank is determined by using the reconstructed point cloud;

[0009] The third identification station identification step: after the wheel hub blank is conveyed to the third identification station by the roller way, the clamping point corresponding to the wheel hub blank is determined by using the reconstructed point cloud, and the third detection result corresponding to the wheel hub blank is determined according to the identification result of the clamping point by the laser marking machine in the wheel hub blank detection device.

[0010] The detection result acquisition step: the detection result corresponding to the wheel hub blank is determined based on the first detection result, the second detection result and the third detection result.

[0011] Optionally, the first identification station identification step comprises:

[0012] When it is detected that the wheel hub blank in the roller way is transmitted to the first identification station, the wheel hub blank is clamped by controlling the preset clamping jaw in the wheel hub blank detection device, and then the digital image corresponding to the wheel hub blank is collected by using the vision camera;

[0013] The wheel hub feature area corresponding to the wheel hub blank in the digital image is identified and acquired, and the wheel type identification result corresponding to the wheel hub blank is determined by using the wheel hub feature area;

[0014] The size parameter corresponding to the wheel hub blank is determined by using the wheel type identification result, and the wheel width identification result corresponding to the wheel hub blank is determined according to the size;

[0015] The valve hole area in the wheel hub blank in the digital image is identified and acquired, and the valve hole identification result corresponding to the wheel hub blank is determined by using the valve hole area;

[0016] The first detection result corresponding to the wheel hub blank is determined according to the wheel type identification result, the wheel width identification result and the valve hole identification result.

[0017] Optionally, the step of identifying and acquiring the wheel hub feature area corresponding to the wheel hub blank in the digital image and determining the wheel type identification result corresponding to the wheel hub blank by using the wheel hub feature area comprises:

[0018] The outer contour area, the spoke area and the center hole area corresponding to the wheel hub blank in the digital image are identified, and the wheel hub feature area is acquired based on the outer contour area, the spoke area and the center hole area;

[0019] The outer contour curve data corresponding to the outer contour area in the wheel hub feature area, the spoke shape data corresponding to the spoke area and the center hole result data corresponding to the center hole area are acquired, and the height data of the wheel hub blank is collected and acquired by controlling the preset height sensor in the wheel hub blank detection device;

[0020] The wheel type data corresponding to the wheel hub blank is determined by using the outer contour curve data, the spoke shape data, the center hole result data and the height data, and the wheel type identification result corresponding to the wheel hub blank is determined according to the matching result of the wheel type data and the preset wheel type template library.

[0021] Optionally, the valve hole region in the hub blank in the digital image is identified and acquired, and the step of determining the valve hole recognition result corresponding to the hub blank by using the valve hole region comprises:

[0022] A valve hole template image corresponding to the valve hole of the hub blank is determined, and after template matching recognition of the digital image by using the valve hole template image, the valve hole region corresponding to the valve hole in the digital image is determined according to the matching recognition result;

[0023] The valve hole digital image corresponding to the valve hole region in the digital image and the outer contour digital image corresponding to the outer contour region are acquired;

[0024] The valve hole angle corresponding to the valve hole is determined according to the position data of the valve hole digital image in the outer contour digital image, and the valve hole recognition result corresponding to the hub blank is determined according to the valve hole angle.

[0025] Optionally, the second recognition station recognition step comprises:

[0026] When it is detected that the hub blank in the roller bed is transmitted to the second recognition station, the preset multiple laser profilers in the hub blank detection equipment are controlled to collect the initial point cloud corresponding to the hub blank;

[0027] The shaft calibration block and the module calibration block contained in the hub blank are acquired, the initial point cloud corresponding to the reconstructed point cloud is obtained by reconstructing and scanning the hub blank by using the shaft calibration block and the module calibration block, and the topography reconstruction result corresponding to the hub blank is determined according to the reconstructed point cloud;

[0028] One or more size parameters of the hub blank are obtained by the reconstructed point cloud, the size parameters include one or more of the wheel width, the inner side wheel outer diameter, the outer side wheel outer diameter, the end face flatness, the rim runout, the end face to bolt hole distance, the rim depth, the rim wall thickness, the inner side fender height, the outer side fender height, the flange and end face parallelism, the inner fender and end face parallelism, the inner side wheel inner diameter, the rim roundness, and the end face runout, and the size detection result corresponding to the hub blank is determined according to the size parameters;

[0029] The second detection result corresponding to the hub blank is determined according to the topography reconstruction result and the size detection result.

[0030] Optionally, the initial point cloud corresponding to the reconstructed point cloud is obtained by reconstructing and scanning the hub blank by using the shaft calibration block and the module calibration block, and the step comprises:

[0031] A shaft coordinate system corresponding to the shaft calibration block is determined, and a first pose matrix corresponding to the laser profiler and the shaft coordinate system is determined according to the rotation parameters corresponding to the shaft calibration block;

[0032] Determine the translation coordinate system corresponding to the module calibration block, and determine the second pose matrix of the laser profilometer corresponding to the translation coordinate system based on the translation parameters corresponding to the module calibration block;

[0033] The wheel hub blank is reconstructed using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud.

[0034] Optionally, the step of reconstructing the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud includes:

[0035] Determine the rotation matrix corresponding to the laser profilometer based on the rotation angle corresponding to the rotating shaft calibration block;

[0036] The pose matrix of the laser profilometer in the world coordinate system is determined by the dot product of the second pose matrix, the rotation matrix, and the first pose matrix.

[0037] Obtain the corresponding point cloud data from the initial point cloud, and use the dot product of the pose matrix and the point cloud data to reconstruct the wheel hub blank, thus obtaining the reconstructed point cloud corresponding to the initial point cloud.

[0038] Optionally, the third identification station identification step includes:

[0039] When the wheel hub blank in the roller conveyor is detected to be transferred to the third recognition station, the first point cloud data corresponding to the end face area of ​​the wheel hub blank and the second point cloud data corresponding to the outer wheel area of ​​the wheel hub blank are obtained in the reconstructed point cloud.

[0040] The inner rim area corresponding to the wheel hub blank is determined using the first point cloud data. The main axis direction of the wheel hub blank is determined based on the axis of symmetry of the inner rim area, and the axial clamping point of the wheel hub blank is determined based on the main axis direction.

[0041] The fitting circle region corresponding to the wheel hub blank is determined using the second point cloud data, and the radial clamping point of the wheel hub blank is determined based on the center and radius of the fitting circle region.

[0042] The laser marking machine in the wheel hub blank inspection equipment is controlled to perform laser marking on the axial clamping points and radial clamping points, and the third inspection result is determined based on the laser marking results corresponding to the clamping points in the wheel hub blank.

[0043] Optionally, the steps of determining the inner rim region corresponding to the wheel hub blank using the first point cloud data, determining the main axis direction of the wheel hub blank based on the axis of symmetry of the inner rim region, and determining the axial clamping point of the wheel hub blank based on the main axis direction include:

[0044] The blank end face region corresponding to the wheel hub blank is obtained by using the first point cloud data, and the blank end face region is subjected to spatial plane fitting processing by the least squares method to obtain the end face fitting plane corresponding to the blank end face region.

[0045] The rim region corresponding to the wheel hub blank is obtained using the first point cloud data, and the axis of the rim region is fitted based on the rotational symmetry axis of the wheel hub blank to obtain the main shaft corresponding to the wheel hub blank.

[0046] The center reference point is determined based on the intersection of the end face fitting plane and the spindle, and a clamping reference line is constructed based on the spindle direction corresponding to the spindle, passing through the center reference point and perpendicular to the end face fitting plane.

[0047] Multiple candidate positions are obtained from the wheel hub blanks at multiple angles evenly distributed on the clamping reference line, and the axial clamping point corresponding to the wheel hub blank is determined based on the candidate positions.

[0048] Secondly, the present invention provides a wheel hub blank inspection system, which is applied to a wheel hub blank inspection equipment; wherein, the wheel hub blank inspection equipment sequentially transmits the wheel hub blank to be inspected to multiple identification stations for inspection via roller conveyors;

[0049] The system includes:

[0050] The first identification station identification module is used to control the vision camera in the wheel hub blank detection equipment to collect the digital image corresponding to the wheel hub blank after the wheel hub blank is transported to the first identification station by the roller conveyor, and to use the digital image to determine the first detection result of the wheel hub blank.

[0051] The second identification station identification module is used to control the laser profilometer in the wheel hub blank detection equipment to generate the reconstructed point cloud corresponding to the wheel hub blank after the wheel hub blank is transported to the second identification station by the roller conveyor, and to use the reconstructed point cloud to determine the second detection result corresponding to the wheel hub blank.

[0052] The third identification station identification module is used to determine the clamping point corresponding to the wheel hub blank after the wheel hub blank is conveyed to the third identification station by the roller conveyor, and to determine the third detection result corresponding to the wheel hub blank based on the marking result of the clamping point by the laser marking machine in the wheel hub blank detection equipment.

[0053] Detection result acquisition module: used to determine the detection result corresponding to the wheel hub blank based on the first detection result, the second detection result, and the third detection result.

[0054] Thirdly, the present invention also provides a wheel hub blank inspection device, which uses roller conveyors to sequentially transport the wheel hub blank to be inspected to multiple identification stations for inspection.

[0055] The wheel hub blank inspection equipment includes at least: a vision camera, a laser profilometer, a laser marking machine, and a controller; the controller is connected to the vision camera, the laser profilometer, and the laser marking machine respectively;

[0056] The controller includes a processor and a memory. The memory stores computer-executable instructions that can be executed by the processor. The processor executes the computer-executable instructions to implement the steps of the wheel hub blank detection method provided in the first aspect.

[0057] Fourthly, embodiments of the present invention also provide a storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the steps of the wheel hub blank detection method provided in the first aspect.

[0058] This invention provides a method and system for detecting wheel hub blanks, applied to a wheel hub blank detection equipment. The equipment uses roller conveyors to sequentially transport the wheel hub blanks to multiple identification stations for detection. During the detection process, the method first transports the wheel hub blank to a first identification station via roller conveyors, then controls a vision camera in the equipment to acquire a digital image of the blank, and uses this image to determine a first detection result. Next, the equipment transports the blank to a second identification station via roller conveyors, then controls a laser profilometer in the equipment to generate a reconstructed point cloud of the blank, and uses this point cloud to determine a second detection result. Subsequently, the equipment transports the blank to a third identification station via roller conveyors, uses the reconstructed point cloud to determine the clamping point of the blank, and uses the laser marking machine in the equipment to mark the clamping point to determine a third detection result. Finally, the final detection result is determined based on the first, second, and third detection results. This method sets up multiple specific recognition stations in the wheel hub blank inspection equipment and uses roller conveyors for automated control. It can complete wheel shape recognition, wheel width recognition and valve hole angle extraction of the wheel hub blank in the first recognition station equipped with a vision camera, complete the shape reconstruction and size detection of the wheel hub blank in the second recognition station equipped with a laser profilometer, and complete the acquisition of clamping points and laser marking process in the third recognition station equipped with a laser marking machine, thereby efficiently realizing the automated inspection of wheel hub blanks.

[0059] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0060] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0061] 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.

[0062] Figure 1 A flowchart of a wheel hub blank inspection method provided in an embodiment of the present invention;

[0063] Figure 2 This is a flowchart of the first identification station identification step S101 of a wheel hub blank detection method provided in an embodiment of the present invention;

[0064] Figure 3 A flowchart of step S202 of a wheel hub blank inspection method provided in an embodiment of the present invention;

[0065] Figure 4 A flowchart of step S204 of a wheel hub blank inspection method provided in an embodiment of the present invention;

[0066] Figure 5 This is a flowchart of the second identification station identification step S102 of a wheel hub blank detection method provided in an embodiment of the present invention;

[0067] Figure 6 In step S502 of the wheel hub blank detection method provided in this embodiment of the invention, a flowchart is shown in which the reconstructed point cloud corresponding to the wheel hub blank is obtained by reconstructing the initial point cloud using the shaft calibration block and the module calibration block.

[0068] Figure 7 A flowchart of step S603 of a wheel hub blank inspection method provided in an embodiment of the present invention;

[0069] Figure 8 A flowchart of the third identification station identification step S103 of a wheel hub blank detection method provided in an embodiment of the present invention;

[0070] Figure 9 A flowchart of step S802 of a wheel hub blank inspection method provided in an embodiment of the present invention;

[0071] Figure 10 This is an illustration of the wheel shape recognition effect in a wheel hub blank detection method provided by an embodiment of the present invention;

[0072] Figure 11 This is a schematic diagram of valve hole stretching in a wheel hub blank inspection method provided by an embodiment of the present invention;

[0073] Figure 12 This is a schematic diagram of the shaft calibration block in a wheel hub blank inspection method provided in an embodiment of the present invention;

[0074] Figure 13 This is a schematic diagram of a module calibration block in a wheel hub blank inspection method provided in an embodiment of the present invention;

[0075] Figure 14 This is a schematic diagram of the reconstructed point cloud in a wheel hub blank detection method provided in an embodiment of the present invention;

[0076] Figure 15 This is a schematic diagram of a wheel hub blank inspection system provided in an embodiment of the present invention;

[0077] Figure 16 This is a schematic diagram of the structure of a wheel hub blank inspection device provided in an embodiment of the present invention;

[0078] Figure 17 This is a schematic diagram of the structure of a controller provided in an embodiment of the present invention.

[0079] icon:

[0080] 1510 - First identification station identification module; 1520 - Second identification station identification module; 1530 - Third identification station identification module; 1540 - Detection result acquisition module;

[0081] 101 - Processor; 102 - Memory; 103 - Bus; 104 - Communication interface. Detailed Implementation

[0082] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0083] To facilitate understanding of this embodiment, a method for detecting wheel hub blanks disclosed in this invention will first be described in detail. This method is applied to a wheel hub blank detection equipment; wherein, the wheel hub blank detection equipment sequentially transports the wheel hub blanks to be detected to multiple identification stations via roller conveyors for detection.

[0084] like Figure 1 As shown, the method includes:

[0085] First identification station identification step S101: After the wheel hub blank is transported to the first identification station by the roller conveyor, the vision camera in the wheel hub blank detection equipment is controlled to collect the digital image corresponding to the wheel hub blank, and the first detection result of the wheel hub blank is determined by the digital image.

[0086] When the wheel hub blank is conveyed to the first identification station via roller conveyor, the inspection process begins. At this time, the wheel hub blank inspection equipment controls a vision camera to acquire multi-angle, high-definition images of the wheel hub blank, generating corresponding digital images. These images will be used to analyze the appearance features of the wheel hub blank to perform wheel shape recognition, size recognition, and valve hole recognition. For example, it will check for obvious defects such as cracks, scratches, and dents on the surface, as well as the basic geometric parameters of the wheel hub (such as diameter and whether the overall outline meets the preliminary specifications). Finally, after processing by the image recognition algorithm, the first inspection result is obtained to determine whether the appearance and basic shape of the wheel hub are qualified.

[0087] Second identification station identification step S102: After the wheel hub blank is transported to the second identification station using a roller conveyor, the laser profilometer in the wheel hub blank detection equipment is controlled to generate the reconstructed point cloud corresponding to the wheel hub blank, and the reconstructed point cloud is used to determine the second detection result corresponding to the wheel hub blank.

[0088] The wheel hub blanks, after being inspected at the first station, continue to be conveyed by roller conveyor to the second identification station. Here, the laser profilometer begins operation, scanning the entire surface of the wheel hub blank with a laser beam to collect a large amount of three-dimensional coordinate data, thereby reconstructing a complete point cloud model of the wheel hub blank. This point cloud model can accurately reflect the three-dimensional morphology of the wheel hub blank, including fine structural information such as the thickness, curvature, and relative position of each part. Based on this, the second identification station can detect potential problems in the internal structure and dimensional accuracy of the wheel hub blank (such as uneven local thickness, dimensional deviations at key positions, etc.), thus generating a second inspection result reflecting whether the three-dimensional structure and dimensional accuracy of the wheel hub meet the standards.

[0089] Third identification station identification step S103: After the wheel hub blank is conveyed to the third identification station using the roller conveyor, the clamping point corresponding to the wheel hub blank is determined by reconstructing the point cloud, and the third detection result corresponding to the wheel hub blank is determined according to the marking result of the clamping point by the laser marking machine in the wheel hub blank detection equipment.

[0090] After completing the second-station inspection, the wheel hub blank is conveyed by roller conveyor to the third identification station. The wheel hub blank inspection equipment uses the reconstructed point cloud model generated at the second station to accurately calculate the clamping points (i.e., the key positions for fixing the wheel hub) required for subsequent processing or handling. Then, a laser marking machine marks the calculated clamping points. The third inspection result is primarily based on the accuracy of the marking, checking whether the laser marking accurately hits the clamping points and whether the clarity and completeness of the marking meet the requirements of subsequent processes, thus ensuring that the clamping point markings provide reliable guidance for subsequent clamping operations.

[0091] Step S104 for obtaining test results: Determine the test results corresponding to the wheel hub blank based on the first test result, the second test result and the third test result.

[0092] The first test result is mainly used to determine whether the appearance is qualified. The second test result is used to guide the finishing of the blank (grinding and cutting). The third test result is used to guide the clamping of the blank wheel hub to ensure clamping stability. The above three test results can be used to comprehensively determine whether the wheel hub blank is qualified.

[0093] Optionally, the first identification station identification step S101, such as... Figure 2 As shown, it includes:

[0094] Step S201: After the wheel hub blank in the roller conveyor is detected to be transferred to the first identification station, the preset grippers in the wheel hub blank detection equipment are controlled to clamp the wheel hub blank, and the digital image corresponding to the wheel hub blank is acquired by the vision camera.

[0095] Once the relevant sensors confirm that the wheel hub blank has been accurately conveyed to the first identification station by the roller conveyor, the wheel hub blank inspection equipment will first activate the preset gripper mechanism. The gripper adaptively clamps the wheel hub blank according to its approximate size, ensuring its stability during the inspection process and avoiding the impact of roller conveyor vibration or slight positional shifts on image acquisition accuracy. After clamping and fixing, the vision camera (usually an industrial-grade high-definition camera, possibly equipped with a ring light source to ensure uniform illumination) takes multi-angle pictures of the wheel hub blank, generating clear digital images and providing high-quality image data for subsequent feature analysis.

[0096] Step S202: Identify and obtain the wheel hub feature region corresponding to the wheel hub blank in the digital image, and use the wheel hub feature region to determine the wheel type recognition result corresponding to the wheel hub blank.

[0097] This step preprocesses the acquired digital images (e.g., denoising and contrast enhancement), then automatically locates and extracts the core feature regions of the wheel hub blank using image recognition algorithms. These include distinctive structural features such as the rim profile, spoke layout, and center hole shape. Based on these feature regions, a comparison is made with a pre-set standard wheel type database to determine the wheel type category of the wheel hub blank (e.g., low-pressure cast wheel, forged wheel, etc.) and whether its overall profile matches the standard template of the corresponding wheel type. Finally, a wheel type recognition result is generated (e.g., wheel type matching is qualified or there is an anomaly).

[0098] Step S203: Use the wheel shape recognition result to determine the corresponding size parameters of the wheel hub blank, and determine the wheel width recognition result corresponding to the wheel hub blank based on the size.

[0099] After determining the wheel type, the corresponding standard size parameter model is called. Combining the pixel ratio of the wheel hub feature area in the digital image with the conversion relationship between the actual physical size, the key dimensions of the wheel hub blank (such as the rim diameter, center hole diameter, etc.) are accurately calculated. Among them, the wheel width is an important parameter. The distance between the two edges of the rim is measured and compared with the standard wheel width range to determine whether it is within the allowable tolerance range, thereby generating a wheel width recognition result (such as wheel width is acceptable, too narrow, or too wide).

[0100] Step S204: Identify and acquire the valve hole region in the wheel hub blank in the digital image, and use the valve hole region to determine the valve hole identification result corresponding to the wheel hub blank.

[0101] The purpose of this step is to identify and extract the valve bore area from the digital image. The valve bore is a key through hole on the wheel hub used to install the valve stem. Through image analysis, it is possible to detect whether the valve bore exists (to avoid missed machining), whether its position deviates from the standard coordinates, whether the bore diameter meets the specifications, and whether there are defects such as burrs or damage on the edge of the bore, thereby generating valve bore identification results (such as whether the valve bore is qualified or whether there is an abnormal position controller / controller shape).

[0102] Step S205: Determine the first detection result corresponding to the wheel hub blank based on the wheel shape recognition result, wheel width recognition result, and valve hole recognition result.

[0103] Based on the combined results of wheel type identification (determining whether the wheel type matches), wheel width identification (determining whether the width dimension meets the standard), and valve hole identification (determining whether the valve hole meets the requirements), and according to the preset qualification judgment rules (if all three items are qualified, the whole is qualified; if one item is seriously abnormal, the whole is unqualified), the final comprehensive inspection result of the wheel hub blank at the first identification station is determined, which is the first inspection result.

[0104] Optionally, step S202 involves identifying and acquiring the wheel hub feature region corresponding to the wheel hub blank in the digital image, and using the wheel hub feature region to determine the wheel type recognition result corresponding to the wheel hub blank, as follows: Figure 3 As shown, it includes:

[0105] Step S301: Identify the outer contour region, spoke region, and center hole region corresponding to the wheel hub blank in the digital image, and obtain the wheel hub feature region based on the outer contour region, spoke region, and center hole region.

[0106] First, the digital image undergoes meticulous processing. Using algorithms such as edge detection and region segmentation, three key structural regions of the wheel hub blank are accurately identified and separated: the outer contour region (the outermost edge contour of the wheel hub, reflecting the overall shape boundary), the spoke region (the supporting structure connecting the wheel hub center to the outer contour, including features such as the number and distribution of spokes), and the center hole region (the circular through-hole in the center of the wheel hub, a crucial part for installation and fixation). These three regions together constitute the core wheel hub feature areas for wheel shape recognition, laying the foundation for subsequent precise analysis.

[0107] Step S302: Obtain the outer contour curve data corresponding to the outer contour area, the spoke shape data corresponding to the spoke area, and the center hole result data corresponding to the center hole area in the wheel hub feature area, and control the preset height sensor in the wheel hub blank detection equipment to collect and obtain the height data of the wheel hub blank.

[0108] After extracting the characteristic regions of the wheel hub, quantitative analysis is performed on each region: for the outer contour region, continuous outer contour curve data (including parameters such as curvature changes and diameter changes) is generated through curve fitting technology; for the spoke region, spoke shape data such as the number of spokes, cross-sectional shape, and arrangement angle are extracted; for the center hole region, center hole result data such as hole diameter and hole edge flatness are obtained. Simultaneously, the equipment controls a pre-set height sensor (such as a laser displacement sensor) to scan the wheel hub blank vertically, collecting height data at different positions (reflecting the thickness changes and overall three-dimensional shape of the wheel hub), supplementing the three-dimensional dimensional information that cannot be provided by two-dimensional images.

[0109] Step S303: Determine the wheel type data corresponding to the wheel blank using the outer contour curve data, spoke shape data, center hole result data, and height data, and determine the wheel type recognition result corresponding to the wheel blank based on the matching result of the wheel type data and the preset wheel type template library.

[0110] The aforementioned outer contour curve data, spoke shape data, center hole result data, and height data are fused to construct a unique wheel profile data model for the wheel blank. Subsequently, this model is compared in multiple dimensions with the equipment's preset wheel profile template library (containing the characteristic parameter threshold ranges of various standard wheel profiles). By calculating the feature matching degree (such as contour fit, spoke parameter consistency, height deviation range, etc.), it is determined whether the wheel profile of the wheel blank matches a certain standard wheel profile in the template library, and the accuracy of the matching. Finally, a wheel profile recognition result is generated (e.g., matching a certain standard wheel profile, wheel profile abnormality, or no matching of the corresponding model).

[0111] Optionally, step S204 involves identifying and acquiring the valve hole region in the wheel hub blank from the digital image, and using the valve hole region to determine the valve hole identification result corresponding to the wheel hub blank. Figure 4 As shown, it includes:

[0112] Step S401: Determine the valve hole template image corresponding to the valve hole of the wheel hub blank. After performing template matching and recognition on the digital image using the valve hole template image, determine the valve hole region corresponding to the valve hole in the digital image based on the matching and recognition results.

[0113] First, a pre-defined valve bore template image is invoked. This template is a baseline image pre-constructed based on the shape, size, and edge features of a standard valve bore. Then, using a template matching algorithm (such as normalized cross-correlation matching), the image is searched and compared within the digital image of the wheel hub blank acquired at the first recognition station. By calculating the similarity between local areas of the image and the template, the region in the digital image that matches the valve bore features is accurately located, thus determining the specific location of the valve bore in the overall image—the valve bore region. This step effectively eliminates interference from other similar holes or stains on the wheel hub surface, ensuring the accuracy of valve bore region recognition.

[0114] Step S402: Obtain the valve hole digital image corresponding to the valve hole region and the outer contour digital image corresponding to the outer contour region in the digital image.

[0115] After identifying the valve bore region, this region was cropped and extracted from the original digital image to obtain a digital image of the valve bore containing only its details (clearly showing local features such as the valve bore's edge, diameter, and shape). Simultaneously, the digital image of the outer contour corresponding to the previously identified outer contour region was extracted (reflecting the complete shape of the outermost edge of the wheel hub blank). The acquisition of these two sub-images provides an independent and clear image data foundation for subsequent analysis of the relative position of the valve bores.

[0116] Step S403: Determine the valve hole angle corresponding to the valve hole based on the position data of the valve hole digital image in the outer contour digital image, and determine the valve hole recognition result corresponding to the wheel hub blank based on the valve hole angle.

[0117] By calculating image coordinates, the relative position data of the valve bore digital image within the outer contour digital image is obtained (e.g., the angle between the line connecting the valve bore center and a reference point on the outer contour edge, with the hub center as the origin). This allows the determination of the valve bore's angular parameters relative to the overall hub contour (i.e., the valve bore angle). Subsequently, this angular parameter is compared with a preset standard valve bore angle range to determine if it is within acceptable tolerances (e.g., whether the angle deviation exceeds the design requirement of ±1°). Finally, based on the comparison results, a valve bore recognition result is generated, including information such as the presence and accuracy of the valve bore (whether the angle is compliant).

[0118] Optionally, the second identification station identification step S102, such as... Figure 5 As shown, it includes:

[0119] Step S501: After the wheel hub blank in the roller conveyor is detected to be transferred to the second identification station, the multiple laser profilometers preset in the wheel hub blank detection equipment are controlled to collect the initial point cloud corresponding to the wheel hub blank.

[0120] Once the sensor detects that the wheel hub blank has been precisely conveyed to the second identification station via the roller conveyor, the wheel hub blank inspection equipment activates multiple preset laser profilometers (typically distributed at multiple angles along the circumference of the wheel hub to ensure coverage of the entire surface). These laser profilometers simultaneously emit laser beams to perform high-speed scanning of the wheel hub blank, acquiring a large amount of discrete three-dimensional coordinate data by capturing laser reflection signals, forming an initial point cloud that can preliminarily reflect the surface morphology of the wheel hub. Collaborative acquisition by multiple devices avoids blind spots from single-angle scanning, ensuring the integrity of the point cloud data.

[0121] Step S502: Obtain the shaft calibration block and module calibration block contained in the wheel hub blank. After reconstructing and scanning the wheel hub blank using the shaft calibration block and module calibration block, obtain the reconstructed point cloud corresponding to the initial point cloud. Determine the shape reconstruction result corresponding to the wheel hub blank based on the reconstructed point cloud.

[0122] First, the pre-set rotation axis calibration block (used to calibrate the rotation axis reference) and module calibration block (used to unify the coordinate system of the multi-laser profilometer) are located on the wheel hub blank. Using the known parameters of these two calibration blocks, the initial point cloud is subjected to coordinate calibration, noise filtering, and data fusion processing. Specifically, this may include eliminating coordinate system deviations in data collected from different devices, filling in missing points in the scanning gaps, and optimizing the point cloud density distribution. Finally, a complete reconstructed point cloud that accurately reflects the three-dimensional shape of the wheel hub is reconstructed. Based on this reconstructed point cloud, a three-dimensional topographic model of the wheel hub blank is generated, i.e., the topographic reconstruction result, which can intuitively display the overall structure, surface changes, and subtle morphological features of the wheel hub.

[0123] Step S503: Obtain one or more dimensional parameters corresponding to the wheel blank, such as wheel width, inner wheel outer diameter, outer wheel outer diameter, end face flatness, rim runout, distance from end face to bolt hole, rim depth, rim wall thickness, inner rim height, outer rim height, flange and end face parallelism, inner rim and end face parallelism, inner wheel inner diameter, rim roundness, and end face runout, by reconstructing the point cloud. Determine the dimensional inspection results corresponding to the wheel blank based on the dimensional parameters.

[0124] This step involves in-depth 3D dimensional analysis of the reconstructed point cloud, extracting several key dimensional parameters. These parameters include: wheel width (distance between the two end faces of the wheel hub), inner / outer wheel diameter (diameter of the inner and outer edges of the rim), end face flatness (flatness of the end face), rim runout (radial deviation during rim rotation), distance from the end face to the bolt hole (distance from the end face reference to the center of the bolt hole), rim depth and wall thickness, inner / outer rim height, parallelism between the flange and the end face, parallelism between the inner rim and the end face, inner wheel diameter, rim roundness (circularity of the rim cross-section), and end face runout (axial deviation during end face rotation). These parameters are then compared one by one with the standard dimensional range of the corresponding wheel type to determine if each parameter is within the allowable tolerance, thus generating comprehensive dimensional inspection results.

[0125] Step S504: Determine the second inspection result corresponding to the wheel hub blank based on the shape reconstruction result and the size inspection result.

[0126] This step integrates the morphology reconstruction results and dimensional inspection results, and analyzes them according to a preset judgment logic. For example, if the morphology reconstruction shows no obvious deformation and all dimensional parameters are within tolerance range, it is judged as qualified; if a key dimension is out of tolerance or there is a significant morphological abnormality, it is judged as unqualified. Finally, the integrated results of the wheel hub blank at the second identification station are obtained, which is the second inspection result.

[0127] Optionally, after reconstructing and scanning the wheel hub blank using the axle calibration block and module calibration block, the reconstructed point cloud corresponding to the initial point cloud can be obtained, such as... Figure 6 As shown, it includes:

[0128] Step S601: Determine the rotation axis coordinate system corresponding to the rotation axis calibration block, and determine the first pose matrix of the laser profilometer corresponding to the rotation axis coordinate system based on the rotation parameters corresponding to the rotation axis calibration block.

[0129] First, the rotation axis calibration block on the wheel hub blank is identified (its position and geometric parameters are known benchmarks). A rotation axis coordinate system is then established using this calibration block as a reference. Typically, the rotation axis of the wheel hub is taken as the origin, and coordinate axes are set along the axial and radial directions to form a reference coordinate system for uniform rotation angle measurement. Subsequently, based on rotation parameters such as angle changes and axis offset collected by the rotation axis calibration block during rotation, a coordinate transformation algorithm is used to calculate the spatial transformation relationship between the measurement coordinate system of each laser profilometer and the rotation axis coordinate system, i.e., the first pose matrix. This matrix contains rotation and translation parameters, which can transform the point cloud data collected by the laser profilometer from its own coordinate system to the rotation axis coordinate system, ensuring that data scanned at different angles maintain consistency in the rotational dimension.

[0130] Step S602: Determine the translation coordinate system corresponding to the module calibration block, and determine the second pose matrix corresponding to the laser profilometer and the translation coordinate system based on the translation parameters corresponding to the module calibration block.

[0131] This step first acquires the module calibration block on the wheel hub blank (used to calibrate the relative positions between multiple laser profilometers), and establishes a moving coordinate system based on this calibration block. This coordinate system typically uses a fixed reference point of the equipment as a reference to unify the spatial positions of different laser profilometers. Based on the displacement, installation deviation, and other translation parameters recorded by the module calibration block during translation, the spatial transformation relationship between the measurement coordinate system and the moving coordinate system of each laser profilometer is calculated, i.e., the second pose matrix. This matrix eliminates measurement deviations caused by differences in the installation positions of different laser profilometers, unifying the point cloud data collected by each device under the same moving coordinate system, ensuring data consistency in the translation dimension.

[0132] Step S603: Reconstruct the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud.

[0133] This step fuses the first pose matrix (responsible for rotational dimension calibration) and the second pose matrix (responsible for translational dimension calibration) to form a complete coordinate transformation model. This model is then used to perform comprehensive coordinate transformation and stitching processing on the initial point cloud data, including aligning point cloud fragments acquired from different angles and positions, eliminating conflicts in overlapping data areas, and filling in missing points in scanning blind spots. The resulting reconstructed point cloud is a complete 3D data model after multi-dimensional calibration and fusion, accurately and consistently reflecting the overall shape of the wheel hub blank, providing a reliable 3D data foundation for subsequent morphology analysis and dimensional inspection.

[0134] Optionally, step S603 involves reconstructing the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud, as shown in step S603. Figure 7 As shown, it includes:

[0135] Step S701: Determine the rotation matrix corresponding to the laser profilometer based on the rotation angle corresponding to the rotating axis calibration block.

[0136] This step first extracts the precise rotation angle parameters (including rotation components around the X, Y, and Z axes) recorded by the axis calibration block during rotation. These angle parameters are then converted into a rotation matrix corresponding to the laser profilometer using methods such as Euler angle transformation or rotation vector calculation. This rotation matrix accurately describes the rotational attitude of the laser profilometer in three-dimensional space, providing a mathematical basis for the subsequent orientation calibration of point cloud data and ensuring that point clouds acquired from different angles maintain consistency in spatial orientation.

[0137] Step S702: Determine the pose matrix of the laser profilometer in the world coordinate system based on the dot product of the second pose matrix, the rotation matrix and the first pose matrix.

[0138] This step uses the world coordinate system (the device's preset global reference coordinate system, used to unify the spatial reference of all detection data) as a benchmark, and performs matrix operations on the acquired second pose matrix (describing the transformation relationship between the laser profilometer and the translation coordinate system), rotation matrix (describing the rotational attitude of the laser profilometer), and first pose matrix (describing the transformation relationship between the laser profilometer and the rotation axis coordinate system). Specifically, through dot product operations of the three (matrix multiplication is performed in a preset order), the local coordinate system of the laser profilometer is gradually transformed and unified to the world coordinate system, finally obtaining the complete pose matrix of the laser profilometer in the world coordinate system. This matrix contains comprehensive parameters of translation and rotation, fully describing the position and attitude of the laser profilometer in global space.

[0139] Step S703: Obtain the corresponding point cloud data in the initial point cloud, and reconstruct the wheel hub blank using the dot product result of the pose matrix and the point cloud data to obtain the reconstructed point cloud corresponding to the initial point cloud.

[0140] This step first extracts the 3D coordinate data of all discrete points in the initial point cloud (based on the local coordinate system of each laser profilometer). Then, it performs a dot product operation between these point cloud data and the pose matrix obtained in step S702, that is, using the coordinate transformation formula, it converts the local coordinates of each point into global coordinates in the world coordinate system. Specifically, the above process is as follows:

[0141]

[0142] in: Pose relationship from the line laser profilometer to world coordinates. The first data collected by the line laser profilometer Frame point cloud data; The pose relationship between the line laser profilometer and the rotating axis; : Rotation of the shaft The rotation matrix corresponding to the angle; This indicates that the module starts from its initial position, and each frame follows the direction vector. With fixed step size Perform a linear translation.

[0143] After this conversion, point cloud data from different laser profilometers and different angles are unified into the same spatial reference system. The converted point cloud is then deduplicated, stitched together, and optimized (such as filling data gaps and smoothing noise points) to finally form a reconstructed point cloud that can completely and accurately reflect the three-dimensional shape of the wheel hub blank, providing a unified and accurate three-dimensional data foundation for subsequent shape analysis and size detection.

[0144] Optionally, the third identification station identification step S103, such as... Figure 8 As shown, it includes:

[0145] Step S801: After the wheel hub blank in the roller conveyor is detected to be transferred to the third recognition station, the first point cloud data corresponding to the end face area of ​​the wheel hub blank and the second point cloud data corresponding to the outer wheel area of ​​the wheel hub blank are obtained in the reconstructed point cloud.

[0146] Once the sensor detects that the wheel hub blank has been conveyed to the third identification station via the roller conveyor, point cloud data for two key regions are precisely extracted from the reconstructed point cloud generated at the second identification station: First, the first point cloud data corresponding to the end face region of the wheel hub blank (reflecting the three-dimensional morphology of the axial end face of the wheel hub, including information such as end face flatness and edge contour); second, the second point cloud data corresponding to the outer rim region of the wheel hub (reflecting the three-dimensional structure of the radial outer edge of the wheel hub, including features such as rim curvature and outer diameter contour). These two types of point cloud data provide a precise three-dimensional coordinate basis for subsequent clamping point calculations.

[0147] Step S802: Use the first point cloud data to determine the inner rim area corresponding to the wheel hub blank, determine the main axis direction of the wheel hub blank according to the axis of symmetry of the inner rim area, and determine the axial clamping point of the wheel hub blank based on the main axis direction.

[0148] This step analyzes the first point cloud data, using a region segmentation algorithm to locate the inner rim region of the wheel hub blank (the inner structure of the rim near the center hole, which is the key stress area for axial clamping). Then, a feature extraction algorithm is used to identify the axis of symmetry of this inner rim region (usually a baseline coinciding with the wheel hub's central axis), thereby determining the wheel hub's main axis direction (i.e., the axial reference). Based on the main axis direction and the structural strength distribution of the inner rim region, axial clamping points that ensure clamping stability are calculated (mostly several points symmetrically distributed along the main axis, used to fix the wheel hub axially).

[0149] Step S803: Use the second point cloud data to determine the fitting circle region corresponding to the wheel hub blank, and determine the radial clamping point of the wheel hub blank based on the center and radius of the fitting circle region.

[0150] For the second point cloud data, a circle fitting algorithm is used to process the point cloud of the outer rim of the wheel hub to generate a fitted circle region that accurately reflects the rim profile (eliminating the influence of local minor deformations on the reference). Based on the center coordinates (corresponding to the center of the wheel hub) and radius parameters of this fitted circle region, combined with the structural strength characteristics of the rim, radial clamping points are calculated (mostly points evenly distributed along the circumference of the fitted circle, used to fix the wheel hub radially and prevent rotational offset).

[0151] Step S804: Control the laser marking machine in the wheel hub blank inspection equipment to perform laser marking on the axial clamping point and the radial clamping point, and determine the third inspection result based on the laser marking result corresponding to the clamping point in the wheel hub blank.

[0152] This step first involves controlling a laser marking machine to perform high-precision laser markings (such as tiny dots or crosshairs) on the surface of the wheel hub blank according to the calculated axial and radial clamping point coordinates. After marking is completed, the accuracy of the laser markings is checked through image recognition: including whether the markings accurately cover the clamping point positions, whether the marking clarity meets the recognition requirements of subsequent processes, and whether the number of markings is consistent with the calculation results. Based on these inspection results, a third inspection result is generated to determine whether the clamping point markings are qualified, ensuring the reliability of clamping operations in subsequent processing.

[0153] Optionally, step S802 involves using the first point cloud data to determine the inner rim region corresponding to the wheel hub blank, determining the main axis direction of the wheel hub blank based on the axis of symmetry of the inner rim region, and determining the axial clamping point of the wheel hub blank based on the main axis direction. Figure 9 As shown, it includes:

[0154] Step S901: Use the first point cloud data to obtain the blank end face region corresponding to the wheel hub blank, and use the least squares method to perform spatial plane fitting processing on the blank end face region to obtain the end face fitting plane corresponding to the blank end face region.

[0155] Step S902: Use the first point cloud data to obtain the rim area corresponding to the wheel hub blank, and after fitting the axis of the rim area based on the rotational symmetry axis of the wheel hub blank, obtain the main shaft corresponding to the wheel hub blank.

[0156] Step S903: Determine the center reference point based on the intersection of the end face fitting plane and the spindle, and construct a clamping reference line that passes through the center reference point and is perpendicular to the end face fitting plane based on the spindle direction corresponding to the spindle.

[0157] Step S904: Obtain multiple candidate positions from the wheel hub blanks evenly distributed along the clamping reference line at multiple angles, and determine the axial clamping point corresponding to the wheel hub blank based on the candidate positions.

[0158] The clamping point selection process first establishes a unified reference coordinate system based on the best-fit registration results between the reconstructed wheel blank point cloud and the blank digital model. Then, using the geometric datum in the blank model as the positioning basis, high-precision clamping point selection is achieved. In the axial clamping point calculation process, point cloud data of the wheel blank end face region is first extracted, and spatial plane fitting is performed using the least squares method to obtain the end face fitting plane. Simultaneously, the inner rim region is segmented from the point cloud, and axis fitting is performed based on its rotationally symmetric structure to obtain the main axis direction of the wheel blank. Subsequently, the algorithm calculates the intersection point of the fitted axis and the fitted end face plane, defining it as the ideal center reference point, and constructs a theoretical clamping reference line passing through this point and perpendicular to the end face. Using three evenly distributed angular directions (120° apart) on this intersection line as candidate positions, the algorithm further searches the end face point cloud for the actual point cloud points closest to these candidate directions, using the minimum perpendicular distance to the fitted intersection line as the criterion, and finally selects the three end face points with the best clamping stability as axial clamping points.

[0159] In the selection of radial clamping points, point cloud data of the outer wheel region of the wheel blank is first extracted. This region typically has a well-defined circular symmetry structure. Using the least squares method or other robust fitting algorithms, spatial circle fitting is performed on the point cloud of this region to obtain the center position and radius of the fitted circle. This center is defined as the central reference for radial clamping. Subsequently, using the fitted circle center as a reference, three angle points (120° apart) are selected at equal intervals along the circumference as candidate radial clamping directions. Rays radiating from the center to these three directions are constructed in space, and point cloud points closest to these directions are searched on the fitted circle. For each direction, the point whose Euclidean distance from its point cloud to the center is closest to the fitted radius is selected as the final radial clamping point, ensuring that the clamping position is located on a high-precision circular profile of the actual wheel-out shape. Furthermore, to improve the overall stability of the clamping, the three axial clamping points on the end face and the three radial clamping points on the wheel-out circumference are staggered by 60°.

[0160] Specifically, the above embodiments mainly involve the identification process at three workstations. Specifically, at the first workstation, a vision camera captures a frontal image of the wheel while it is held by the grippers, obtaining key information such as its diameter and shape features. Simultaneously, combined with the overall wheel height data obtained from the height sensor in front of the feed roller conveyor, template matching is performed with a preset wheel type template library using multi-dimensional features such as the outer contour curve, spoke shape, and center hole structure to complete automatic wheel type identification. The wheel hub blank template, the rotated wheel hub blank, and the matching results are shown below. Figure 10 As shown.

[0161] After wheel shape recognition, wheel width recognition is performed. By retrieving the standard size parameter library corresponding to the wheel shape and comparing it with the standard parameters, the preset loading of wheel size parameters, including diameter and height information, is achieved. Subsequently, mechanical movement and parameter configuration need to be performed based on the recognition results. The module is dynamically scheduled to move the camera according to the recognition results to accommodate wheel hub blanks of different sizes.

[0162] Geometric features of the valve hole region are extracted using a frontal image processing algorithm. Combined with wheel shape recognition results, corresponding template information is retrieved. A template matching algorithm is then used to automatically identify and label the valve hole position, thus completing valve hole recognition. The valve hole position will be used as the initial point in subsequent clamping point calculation and marking positioning stages. The valve hole angle is used to correct the wheel hub pose to the origin. Valve hole angle = (identified valve hole pixel position / total number of pixels in the stretched image) * 360. The stretched image of the valve hole is shown below. Figure 11 As shown.

[0163] The topography reconstruction and dimensional inspection processes are performed at the second station. A schematic diagram of the shaft calibration block is shown below. Figure 12 As shown, the shaft calibration process first involves scanning the dedicated calibration block from multiple angles using a line laser profilometer to obtain its brightness and height maps at different turntable angles. Multiple known marker points on the calibration block's digital model are then extracted from the scanning results. Corresponding coordinates of the marker points under the line laser profilometer And simultaneously record the corresponding turntable rotation angle. Based on a set Data, complete the pose matrix from the line laser profilometer to the rotation axis coordinate system. Initial estimation. Subsequently, a complete 3D point cloud of the calibration block at various angles is constructed and registered with its digital model. The distance from the midpoint of the reconstruction result to the surface of the digital model is calculated as the optimization objective function. A nonlinear least squares method (LM algorithm) is used to optimize the pose matrix. Perform global optimization.

[0164] A schematic diagram of the module calibration block is shown below. Figure 13 As shown, nonlinear least squares optimization is first performed based on the intersection points obtained from the 4 intersection points and 5 planes each time to obtain the relative pose relationship between the intersection points and the coordinate system of the line laser profiler, and finally the motion vector of the line laser profiler is obtained. For each move The direction vector.

[0165] The shaft calibration and module calibration are combined with the point cloud collected by the line laser profilometer for reconstruction. A schematic diagram of the reconstructed point cloud from the line laser profilometer and the registered wheel hub digital model is shown below. Figure 14 As shown.

[0166] Among them, the The global 3D points of the frame are:

[0167] (5)

[0168] in: Pose relationship from the line laser profilometer to world coordinates. The first data collected by the line laser profilometer Frame point cloud data; The pose relationship between the line laser profilometer and the rotating axis; : Rotation of the shaft The rotation matrix corresponding to the angle; This indicates that the module starts from its initial position, and each frame follows the direction vector. With fixed step size Perform a linear translation.

[0169] Dimensional inspection is performed using the reconstructed point cloud. The dimensional inspection mainly includes the following: wheel width, inner wheel outer diameter, outer wheel outer diameter, end face flatness, rim runout, distance from end face to bolt hole, rim depth, rim wall thickness, inner rim height, outer rim height, flange and end face parallelism, inner rim and end face parallelism, inner wheel inner diameter, rim roundness, and end face runout.

[0170] The third station outputs and marks the clamping points. To improve clamping accuracy and machining consistency in subsequent processes, after selecting the clamping points, a laser marking machine marks the selected clamping point positions, forming identifiable clamping reference marks on the surface of the wheel hub blank. This provides a basis for the precise alignment of the subsequent automatic fixture, improving the positioning rigidity and stability of the wheel hub blank during clamping. This stability is particularly crucial in the rough machining stage of the wheel hub blank, effectively reducing deformation and cutting errors caused by clamping deviations or uneven stress.

[0171] As can be seen from the wheel hub blank inspection method in the above embodiments, the method sets up multiple specific recognition stations in the wheel hub blank inspection equipment and uses roller conveyors for automated control. It can complete the wheel shape recognition, wheel width recognition and valve hole angle extraction of the wheel hub blank in the first recognition station equipped with a vision camera, and complete the shape reconstruction and size detection of the wheel hub blank in the second recognition station equipped with a laser profilometer. It can also complete the acquisition of clamping points and laser marking process in the third recognition station equipped with a laser marking machine, thereby efficiently realizing the automated inspection of wheel hub blanks.

[0172] Corresponding to the above embodiments of the wheel hub blank inspection method, this embodiment of the invention also provides a wheel hub blank inspection system, which is applied to a wheel hub blank inspection equipment; wherein, the wheel hub blank inspection equipment sequentially transmits the wheel hub blank to be inspected to multiple identification stations for inspection via roller conveyors;

[0173] like Figure 15 As shown, the system includes:

[0174] First identification station identification module 1510: After the wheel hub blank is transported to the first identification station by the roller conveyor, it controls the vision camera in the wheel hub blank detection equipment to collect the digital image corresponding to the wheel hub blank, and uses the digital image to determine the first detection result of the wheel hub blank.

[0175] The second identification station identification module 1520 is used to control the laser profilometer in the wheel hub blank detection equipment to generate the reconstructed point cloud corresponding to the wheel hub blank after the wheel hub blank is transported to the second identification station by the roller conveyor, and to use the reconstructed point cloud to determine the second detection result corresponding to the wheel hub blank.

[0176] The third identification station identification module 1530 is used to determine the clamping point corresponding to the wheel hub blank after the wheel hub blank is conveyed to the third identification station by the roller conveyor, and to determine the third detection result corresponding to the wheel hub blank based on the marking result of the clamping point by the laser marking machine in the wheel hub blank detection equipment.

[0177] Detection result acquisition module 1540: used to determine the detection result corresponding to the wheel hub blank based on the first detection result, the second detection result, and the third detection result.

[0178] As can be seen from the above-mentioned wheel hub blank inspection system, the system is equipped with multiple specific recognition stations and uses roller conveyors for automated control. It can complete wheel shape recognition, wheel width recognition and valve hole angle extraction of the wheel hub blank in the first recognition station equipped with a vision camera, and complete the shape reconstruction and size detection of the wheel hub blank in the second recognition station equipped with a laser profilometer. It can also complete the acquisition of clamping points and laser marking process in the third recognition station equipped with a laser marking machine, thereby efficiently realizing the automated inspection of wheel hub blanks.

[0179] The wheel hub blank inspection system provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned wheel hub blank inspection method embodiment. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned wheel hub blank inspection method embodiment.

[0180] like Figure 16As shown, this embodiment also provides a wheel hub blank inspection device. The wheel hub blank inspection device sequentially transmits the wheel hub blank to be inspected to multiple identification stations for inspection via roller conveyors. The wheel hub blank inspection device includes at least: a vision camera, a laser profilometer, a laser marking machine, and a controller. The controller is connected to the vision camera, the laser profilometer, and the laser marking machine, respectively.

[0181] The wheel hub blank inspection equipment contains three stations. Specifically, station 1 is based on visual recognition to complete wheel shape recognition, wheel width recognition and valve hole angle extraction. It also dynamically schedules the module to move the camera according to the recognition results to take into account wheel hub blanks of different sizes. The valve hole angle recognition is used to correct the wheel hub pose to the origin.

[0182] In station 2, a joint calibration algorithm for the module's linear motion direction vector and the rotation axis's spatial attitude parameters is introduced. Through nonlinear least-squares optimization (such as the LM algorithm), the spatial pose of the line laser profilometer is accurately calculated at each frame acquisition time, thereby improving the stitching accuracy of multi-angle scanning. The resulting high-quality point cloud can be precisely registered with the blank digital model, thus achieving more accurate contour analysis and dimensional evaluation.

[0183] In station 3, a spatial fitting algorithm combining point cloud and digital model is introduced. By combining end face plane fitting and inner rim axis fitting, three axial clamping points are extracted. By combining spatial circle fitting of the outer wheel circular region, three radial clamping points are extracted. The clamping points are spatially divided and staggered (60° offset) to form a highly stable six-point clamping layout. At the same time, the clamping points are physically marked by an automatic marking system, realizing the transformation from virtual calculation reference to physical clamping reference.

[0184] The structural diagram of the controller is shown below. Figure 17 As shown, the device includes a processor 101 and a memory 102; wherein, the memory 102 is used to store one or more computer instructions, which are executed by the processor to implement the steps of the above-described wheel hub blank detection method.

[0185] Figure 17 The controller shown also includes a bus 103 and a communication interface 104. The processor 101, the communication interface 104, and the memory 102 are connected via the bus 103.

[0186] The memory 102 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. The bus 103 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 17The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0187] The communication interface 104 is used to connect to at least one user terminal and other network units through a network interface, and to send encapsulated IPv4 packets or IPv4 packets to the user terminal through the network interface.

[0188] Processor 101 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 101 or by instructions in software form. The processor 101 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 102. The processor 101 reads the information in memory 102 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0189] It should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered 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 claims.

Claims

1. A method for inspecting wheel hub blanks, characterized in that, The method is applied to a wheel hub blank inspection equipment; wherein, the wheel hub blank inspection equipment sequentially transmits the wheel hub blank to be inspected to multiple identification stations for inspection via roller conveyors; The method includes: The first identification station identification step: After the wheel hub blank is transported to the first identification station by the roller conveyor, the vision camera in the wheel hub blank detection equipment is controlled to collect the digital image corresponding to the wheel hub blank, and the first detection result of the wheel hub blank is determined by the digital image. The second identification station identification step: After the wheel hub blank is transported to the second identification station using the roller conveyor, the laser profilometer in the wheel hub blank detection equipment is controlled to generate the reconstructed point cloud corresponding to the wheel hub blank, and the second detection result corresponding to the wheel hub blank is determined using the reconstructed point cloud. The third identification station identification step: After the wheel hub blank is conveyed to the third identification station using the roller conveyor, the clamping point corresponding to the wheel hub blank is determined using the reconstructed point cloud, and the third detection result corresponding to the wheel hub blank is determined according to the marking result of the clamping point by the laser marking machine in the wheel hub blank detection equipment; wherein, the clamping point corresponds to the inner rim area and the fitted circle area of ​​the wheel hub blank; Steps for obtaining test results: Determine the test result corresponding to the wheel hub blank based on the first test result, the second test result, and the third test result; The second identification station identification step includes: When the wheel hub blank in the roller conveyor is detected to be transferred to the second identification station, the multiple laser profilometers preset in the wheel hub blank detection equipment are controlled to collect the initial point cloud corresponding to the wheel hub blank. The axle calibration block and module calibration block contained in the wheel hub blank are obtained. The wheel hub blank is reconstructed and scanned using the axle calibration block and the module calibration block to obtain the reconstructed point cloud corresponding to the initial point cloud. The shape reconstruction result corresponding to the wheel hub blank is determined based on the reconstructed point cloud. The reconstructed point cloud is used to obtain one or more dimensional parameters of the wheel blank, including wheel width, inner wheel outer diameter, outer wheel outer diameter, end face flatness, rim circular runout, distance from end face to bolt hole, rim depth, rim wall thickness, inner rim height, outer rim height, flange and end face parallelism, inner rim and end face parallelism, inner wheel inner diameter, rim roundness, and end face runout. The dimensional detection results of the wheel blank are then determined based on the dimensional parameters. The second detection result corresponding to the wheel hub blank is determined based on the morphology reconstruction result and the size detection result; After reconstructing and scanning the wheel hub blank using the axle calibration block and the module calibration block, the reconstructed point cloud corresponding to the initial point cloud is obtained, including: Determine the rotation axis coordinate system corresponding to the rotation axis calibration block, and determine the first pose matrix of the laser profilometer corresponding to the rotation axis coordinate system based on the rotation parameters corresponding to the rotation axis calibration block; Determine the translation coordinate system corresponding to the module calibration block, and determine the second pose matrix of the laser profilometer and the translation coordinate system based on the translation parameters corresponding to the module calibration block; The wheel blank is reconstructed using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud; The step of reconstructing the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud includes: The rotation matrix corresponding to the laser profilometer is determined based on the rotation angle corresponding to the rotating axis calibration block. The pose matrix of the laser profilometer in the world coordinate system is determined based on the dot product of the second pose matrix, the rotation matrix, and the first pose matrix. Obtain the corresponding point cloud data in the initial point cloud, and reconstruct the wheel hub blank using the dot product result of the pose matrix and the point cloud data to obtain the reconstructed point cloud corresponding to the initial point cloud.

2. The wheel hub blank inspection method according to claim 1, characterized in that, The first identification station identification step includes: When the wheel hub blank in the roller conveyor is detected to be transferred to the first identification station, the preset grippers in the wheel hub blank detection equipment are controlled to clamp the wheel hub blank, and the vision camera is used to collect the digital image corresponding to the wheel hub blank. Identify and obtain the wheel hub feature region corresponding to the wheel hub blank in the digital image, and use the wheel hub feature region to determine the wheel type recognition result corresponding to the wheel hub blank; The wheel shape recognition result is used to determine the size parameters corresponding to the wheel hub blank, and the wheel width recognition result corresponding to the wheel hub blank is determined based on the size. Identify and acquire the valve hole region in the wheel hub blank in the digital image, and use the valve hole region to determine the valve hole identification result corresponding to the wheel hub blank; The first detection result corresponding to the wheel hub blank is determined based on the wheel shape recognition result, the wheel width recognition result, and the valve hole recognition result.

3. The wheel hub blank inspection method according to claim 2, characterized in that, The steps of identifying and acquiring the wheel hub feature region corresponding to the wheel hub blank in the digital image, and using the wheel hub feature region to determine the wheel type recognition result corresponding to the wheel hub blank, include: Identify the outer contour region, spoke region, and center hole region corresponding to the wheel hub blank in the digital image, and obtain the wheel hub feature region based on the outer contour region, the spoke region, and the center hole region; The system acquires the outer contour curve data corresponding to the outer contour region, the spoke shape data corresponding to the spoke region, and the center hole result data corresponding to the center hole region in the wheel hub feature region, and controls the preset height sensor in the wheel hub blank detection device to collect and acquire the height data of the wheel hub blank. The wheel type data corresponding to the wheel hub blank is determined by using the outer contour curve data, the spoke shape data, the center hole result data, and the height data, and the wheel type recognition result corresponding to the wheel hub blank is determined by matching the wheel type data with the preset wheel type template library.

4. The wheel hub blank inspection method according to claim 3, characterized in that, The steps of identifying and acquiring the valve hole region in the wheel hub blank in the digital image, and determining the valve hole identification result corresponding to the wheel hub blank using the valve hole region, include: Determine the valve hole template image corresponding to the valve hole of the wheel hub blank, and use the valve hole template image to perform template matching and recognition on the digital image. Then, determine the valve hole region corresponding to the valve hole in the digital image based on the matching and recognition results. Acquire the valve hole digital image corresponding to the valve hole region and the outer contour digital image corresponding to the outer contour region in the digital image; The valve hole angle corresponding to the valve hole is determined based on the position data of the valve hole digital image in the outer contour digital image, and the valve hole recognition result corresponding to the wheel hub blank is determined based on the valve hole angle.

5. The wheel hub blank inspection method according to claim 1, characterized in that, The third identification station identification step includes: When the wheel hub blank in the roller conveyor is detected to be transferred to the third identification station, the first point cloud data corresponding to the end face area of ​​the wheel hub blank and the second point cloud data corresponding to the outer wheel area of ​​the wheel hub blank are obtained in the reconstructed point cloud. The inner rim area corresponding to the wheel hub blank is determined using the first point cloud data, the main axis direction of the wheel hub blank is determined according to the axis of symmetry of the inner rim area, and the axial clamping point of the wheel hub blank is determined based on the main axis direction. The fitting circle region corresponding to the wheel hub blank is determined using the second point cloud data, and the radial clamping point of the wheel hub blank is determined based on the center and radius of the fitting circle region. The laser marking machine in the wheel hub blank inspection equipment is controlled to perform laser marking on the axial clamping point and the radial clamping point, and the third inspection result is determined based on the laser marking result corresponding to the clamping point in the wheel hub blank.

6. The wheel hub blank inspection method according to claim 5, characterized in that, The steps of determining the inner rim region corresponding to the wheel hub blank using the first point cloud data, determining the main axis direction of the wheel hub blank based on the axis of symmetry of the inner rim region, and determining the axial clamping point of the wheel hub blank based on the main axis direction include: The blank end face region corresponding to the wheel hub blank is obtained by using the first point cloud data, and the blank end face region is subjected to spatial plane fitting processing by the least squares method to obtain the end face fitting plane corresponding to the blank end face region. The rim area corresponding to the wheel hub blank is obtained by using the first point cloud data, and the rim area is fitted with the axis based on the rotational symmetry axis of the wheel hub blank to obtain the main shaft corresponding to the wheel hub blank. The center reference point is determined based on the intersection of the end face fitting plane and the spindle, and a clamping reference line is constructed based on the spindle direction corresponding to the spindle, passing through the center reference point and perpendicular to the end face fitting plane. Multiple candidate positions are obtained from the wheel hub blank by means of multiple angle directions evenly distributed on the clamping reference line, and the axial clamping point corresponding to the wheel hub blank is determined based on the candidate positions.

7. A wheel hub blank inspection system, characterized in that, The system is applied to a wheel hub blank inspection equipment; wherein, the wheel hub blank inspection equipment uses roller conveyors to sequentially transport the wheel hub blank to be inspected to multiple identification stations for inspection; The system includes: First identification station identification module: After the wheel hub blank is transported to the first identification station by the roller conveyor, the vision camera in the wheel hub blank detection equipment is controlled to collect the digital image corresponding to the wheel hub blank, and the first detection result of the wheel hub blank is determined by the digital image. The second identification station identification module is used to control the laser profilometer in the wheel hub blank detection equipment to generate a reconstructed point cloud corresponding to the wheel hub blank after the wheel hub blank is transported to the second identification station using the roller conveyor, and to use the reconstructed point cloud to determine the second detection result corresponding to the wheel hub blank. The third identification station identification module is used to determine the clamping point corresponding to the wheel hub blank after the wheel hub blank is conveyed to the third identification station using the roller conveyor, and to determine the third detection result corresponding to the wheel hub blank based on the marking result of the clamping point by the laser marking machine in the wheel hub blank detection equipment; wherein, the clamping point corresponds to the inner rim area and the fitted circle area of ​​the wheel hub blank; Detection result acquisition module: used to determine the detection result corresponding to the wheel hub blank based on the first detection result, the second detection result and the third detection result; The second identification station identification module is further configured to, upon detecting that the wheel hub blank in the roller conveyor has been transferred to the second identification station, control multiple laser profilometers preset in the wheel hub blank detection equipment to acquire the initial point cloud corresponding to the wheel hub blank; acquire the shaft calibration block and module calibration block contained in the wheel hub blank; use the shaft calibration block and module calibration block to reconstruct and scan the wheel hub blank to obtain the reconstructed point cloud corresponding to the initial point cloud; and determine the shape reconstruction result corresponding to the wheel hub blank based on the reconstructed point cloud; through the... The method involves reconstructing the point cloud to obtain one or more dimensional parameters corresponding to the wheel blank, including wheel width, inner wheel outer diameter, outer wheel outer diameter, end face flatness, rim circular runout, distance from end face to bolt hole, rim depth, rim wall thickness, inner rim height, outer rim height, flange and end face parallelism, inner rim and end face parallelism, inner wheel inner diameter, rim roundness, and end face runout. Based on these dimensional parameters, the method determines the dimensional detection result corresponding to the wheel blank. Finally, based on the morphology reconstruction result and the dimensional detection result, the method determines the second detection result corresponding to the wheel blank. The second identification station identification module, in the process of obtaining the reconstructed point cloud corresponding to the initial point cloud after reconstructing the wheel hub blank using the rotating shaft calibration block and the module calibration block, is further configured to: determine the rotating shaft coordinate system corresponding to the rotating shaft calibration block; determine the first pose matrix corresponding to the rotating shaft coordinate system based on the rotation parameters corresponding to the rotating shaft calibration block; determine the translation coordinate system corresponding to the module calibration block; determine the second pose matrix corresponding to the translation coordinate system based on the translation parameters corresponding to the module calibration block; and reconstruct the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud. The second identification station identification module, in the process of reconstructing the wheel hub blank using the first pose matrix and the second pose matrix to obtain the reconstructed point cloud corresponding to the initial point cloud, is further configured to: determine the rotation matrix corresponding to the laser profilometer based on the rotation angle corresponding to the rotating shaft calibration block; determine the pose matrix of the laser profilometer in the world coordinate system based on the dot product result of the second pose matrix, the rotation matrix, and the first pose matrix; acquire the point cloud data corresponding to the initial point cloud; and reconstruct the wheel hub blank using the dot product result of the pose matrix and the point cloud data to obtain the reconstructed point cloud corresponding to the initial point cloud.

Images