Method for detecting precision of 3D (three-dimensional) vision wheel position indicator equipment

Abstract

The invention discloses a method for detecting the precision of 3D visual wheel position indicator equipment, which is used for detecting the precision of the 3D visual wheel position indicator equipment on site. The method comprises the following steps: fixing left and right target plates at two ends of a calibration rod, acquiring images of the target plates before and after the calibration rod rotates around a fixed shaft by using cameras on the left and right sides of a positioning instrument, calculating space parameters of a rotating shaft in a left and right target plate coordinate system, and constructing a virtual rotating shaft; the method comprises the following steps: moving a calibration rod in a measurement space to obtain poses of a plurality of positions, unifying virtual rotating shaft parameters of all the positions to the same measurement coordinate system, calculating a consistency error of the virtual rotating shaft parameters, and judging the precision of the position indicator according to a preset precision threshold value. According to the method, the virtual rotating shaft is constructed as the precision detection reference, so that field precision evaluation can be carried out on the 3D vision wheel position indicator equipment without a complex precision device. The optimal scheme supports single-rotation rapid detection or multiple-rotation fitting optimization, and errors can be output as geometric parameters or visual vehicle positioning parameters. The method has the advantages of simple steps and high reliability, and can timely and conveniently evaluate the equipment precision condition on the user site.

Description

Technical Field

[0001] This invention relates to the field of optical measurement instrument accuracy testing technology, and in particular to a method for on-site accuracy testing of wheel alignment equipment based on 3D vision. Background Technology

[0002] 3D vision wheel alignment machines are widely used in the automotive repair field. Their measurement accuracy is highly dependent on the accurate calibration of the positional relationship of multiple cameras. Factors such as vibration and temperature changes during use may cause the calibration parameters to drift, affecting the measurement accuracy. Therefore, a convenient and effective on-site accuracy testing method is needed.

[0003] Currently, the accuracy detection of 3D vision wheel alignment instruments faces fundamental technical challenges: existing detection methods are all based on the direct comparison mode of "physical reference - measurement result": the national standard GB / T 33570-2017 relies on complex mechanical devices as physical references, which are bulky and costly; on-site vehicle detection methods are unreliable due to the uncertainty of vehicle conditions; although patented solutions CN217542418U and CN222895723U have made some improvements, they still require high-precision mechanical structures as detection references.

[0004] This "physical benchmark" model has inherent drawbacks: first, the benchmark itself requires regular calibration, increasing maintenance costs; second, the machining precision of the mechanical structure limits the upper limit of detection accuracy; and third, it is not convenient for rapid on-site deployment. Especially for 3D vision technology, its characteristic of calculating spatial coordinates through image processing should be more suitable for algorithm-based detection methods, but existing technologies have failed to fully utilize this advantage.

[0005] Therefore, there is an urgent need to break through the traditional physical benchmark model and develop a detection method based on virtual benchmarks to fundamentally solve the problems of convenience and reliability in on-site detection. Summary of the Invention

[0006] This invention provides a method for detecting the accuracy of a 3D vision wheel alignment device. The method uses a constructed virtual rotation axis to detect the accuracy of the 3D vision wheel alignment device. While reducing the size of the detection device, simplifying its structure, and lowering the difficulty of machining and assembly, the method ensures that the detection steps are simple and feasible. This allows wheel alignment device manufacturers, suppliers, users, and relevant regulatory authorities to better assess the accuracy of the equipment in the field and promptly determine whether recalibration is needed.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] Firstly, this application provides a method for detecting the accuracy of a 3D vision wheel alignment device. The core innovation of this application lies in: constructing a stable virtual rotation axis as a detection benchmark through simple physical rotational motion, combined with image processing and spatial geometric calculations. Then, the virtual rotation axis is measured at multiple different positions and angles in the measurement space. The measurement accuracy of the 3D vision wheel alignment device is judged by comparing the consistency of the measurement results. This virtual rotation axis does not rely on high-precision mechanical structures, thereby greatly reducing the size of the detection device, simplifying its structure, and reducing the difficulty of machining and assembly. This achieves a fundamental shift in detection technology from relying on "mechanical accuracy" to relying on "algorithm accuracy." The method includes:

[0009] Step 1: Fix the left and right target plates to both ends of the calibration rod so that the left and right target plates are respectively within the field of view of the left and right cameras of the 3D vision wheel positioning instrument;

[0010] Step 2: Rotate the calibration rod around the fixed axis. The left and right cameras of the 3D vision wheel positioning instrument will respectively capture images of the left and right target plates before and after rotation, and calculate the pose of each target plate before and after rotation.

[0011] Step 3: Based on the geometric characteristics of rigid body rotational motion, even if the calibration rod has manufacturing errors, a virtual rotation axis with an ideal straight line can be constructed using spatial geometric algorithms by collecting the pose data when rotating around a fixed axis. Specifically, based on the pose of each target plate before and after rotation, the spatial parameters of the virtual rotation axis in the coordinate systems of the left and right target plates are calculated through spatial geometric transformation.

[0012] Step 4: Move the calibration rod to multiple different positions, ensuring that the left and right target plates are within the field of view of the left and right cameras respectively at each position;

[0013] Step 5: At each position, the left and right cameras respectively acquire the poses of the left and right target plates, and unify the virtual rotation axis spatial parameters of the left and right target plates in their respective coordinate systems into the global measurement coordinate system;

[0014] Step 6: Since the virtual rotation axis is a geometrically ideal straight line with a fixed length, theoretically the measurement results at different positions in the measurement space should be completely consistent. Any measurement deviation reflects the systematic error of the 3D vision positioning instrument. Therefore, based on the unified data, the consistency error of the virtual rotation axis parameters is calculated. The measurement accuracy of the 3D vision wheel positioning instrument can be judged according to the preset accuracy threshold.

[0015] Preferably, steps 2 and 3 involve calculating the spatial parameters of the virtual rotation axis by rotating the calibration rod once around the fixed axis.

[0016] Preferably, in steps 2 and 3, the virtual rotation axis spatial parameters are optimized by rotating the calibration rod around the same fixed axis at least twice and fitting the pose data from the multiple rotations.

[0017] Preferably, the spatial parameters of the rotation axis in step 3 in the respective coordinate systems of the left and right target plates include: the direction vector of the rotation axis and the reference point of the rotation axis at the end of the target plate.

[0018] Preferably, the reference point of the rotation axis at the end of the target plate is the coordinate of the intersection point of the rotation axis and the plane of the target plate.

[0019] Preferably, the reference point of the rotation axis at the end of the target plate is the coordinate of the intersection point of the rotation axis and a fixed plane (such as the clamp plane) in the coordinate system of the target plate.

[0020] Preferably, in step 6, the consistency error is converted into a custom error type and output, wherein the error type includes at least one of geometric parameter error and vehicle positioning parameter error.

[0021] Preferably, the geometric parameter error includes at least one of the direction vector angle error of the rotation axis and the length error.

[0022] Preferably, the vehicle positioning parameter error includes at least one of toe angle error, camber angle error, and track width error.

[0023] The beneficial effects of the present invention are: (1) The technical path of “building a virtual reference by rotational motion” replaces the traditional “mechanical structure as physical reference” mode, which fundamentally simplifies the detection process: The present invention can complete the accuracy detection of the 3D vision wheel positioning device by simply rotating and moving the calibration rod. The steps are simple and feasible, and there is no need to carry out cumbersome mechanical assembly and calibration steps on site, which effectively reduces the intensity and complexity of manual operation.

[0024] (2) Using the original target plate of the positioning instrument, only a simple calibration rod is needed to complete the detection, and the device structure is extremely simplified: The main structure of the device of this invention is a rigid, movable, and rotatable calibration rod. The two ends of the rod have connectors, which can rigidly fix the left and right target plates of the 3D vision wheel positioning instrument (no need to prepare additional target plates) to the calibration rod respectively. The length of the calibration rod should be sufficient to ensure that the left and right target plates can appear in the field of view of the cameras on both sides of the 3D vision wheel positioning instrument after being rigidly connected. The calibration rod should be able to remain stationary when the camera acquires the target plate image. Directly using the target plate of the 3D vision wheel positioning instrument for detection can also detect the inaccuracy of the equipment caused by the deformation of the target plate, etc.

[0025] (3) Low requirements for machining and assembly of the device: The present invention only requires that the calibration rod does not deform during the detection process, that the relative positional relationship between the two target plates does not change, and that they are accurately identified within the field of view of the left and right cameras of the 3D vision wheel positioning instrument; it is not necessary for the rotation axis and the axis of the calibration rod to be coaxial, there are no strict requirements for the straightness of the calibration rod, the calibration rod can also be assembled in sections, and there are no precise requirements for the installation angle and position of the target plate relative to the calibration rod, so the requirements for machining and assembly of the device are not high. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 A schematic flowchart illustrating a method for detecting the accuracy of a 3D wheel alignment device provided in an embodiment of the present invention;

[0028] Figure 2 This is a three-dimensional schematic diagram of the target plate rotating around a fixed axis in a 3D wheel alignment device accuracy detection method provided in an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram illustrating the virtual rotation axis construction principle in a 3D wheel alignment device accuracy detection method provided by an embodiment of the present invention;

[0030] Figure 4 This is a schematic diagram illustrating that the virtual rotating axis does not rotate along the central axis of the calibration rod in a 3D wheel alignment device accuracy detection method provided in an embodiment of the present invention.

[0031] Figure 5 This is a schematic diagram illustrating the arbitrary horizontal and cross-sectional movement of the calibration rod to several positions in a 3D wheel alignment device accuracy detection method provided in an embodiment of the present invention.

[0032] Figure 6 This is a schematic diagram of a 3D wheel alignment device accuracy detection apparatus provided in an embodiment of the present invention;

[0033] Figure 7 This is a schematic diagram of the calibration rod after segmented assembly in a 3D wheel alignment device accuracy testing apparatus provided in an embodiment of the present invention;

[0034] Figure 8 This is a schematic diagram of a calibration rod fixing a target plate via an adapter plate in a 3D vision wheel positioning device accuracy detection device provided in an embodiment of the present invention. Detailed Implementation

[0035] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail and completely below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] like Figure 1 The diagram shown is a schematic flowchart of a 3D vision wheel positioning device accuracy detection method provided by an embodiment of the present invention.

[0037] First, in step S01, the left and right target plates are fixedly connected to both ends of the calibration rod, so that the left and right target plates are respectively located within the field of view of the left and right cameras of the 3D vision wheel positioning instrument.

[0038] Then, in step S02, the calibration rod is rotated around the fixed axis, and the left and right cameras of the 3D vision wheel positioning instrument respectively acquire images of the left and right target plates before and after rotation, and calculate the pose of each target plate before and after rotation.

[0039] Furthermore, in step S03, based on the geometric characteristics of rigid body rotational motion, even if the calibration rod has manufacturing errors, a virtual rotation axis of an ideal straight line can be constructed using spatial geometric algorithms by using the pose data collected when rotating around a fixed axis. Specifically, based on the poses of each target plate before and after rotation, the spatial parameters of the virtual rotation axis in the coordinate systems of the left and right target plates are calculated through spatial geometric transformation.

[0040] Furthermore, in step S04, the calibration rod is moved to multiple different positions, each position ensuring that the left and right target plates are respectively within the field of view of the left and right cameras;

[0041] Further, in step S05, at each position, the left and right cameras respectively acquire the poses of the left and right target plates, and unify the virtual rotation axis spatial parameters of the left and right target plates in their respective coordinate systems into the global measurement coordinate system;

[0042] Finally, in step S06, since the virtual rotation axis is a geometrically ideal straight line with a fixed length, theoretically the measurement results at different positions in the global measurement space should be completely consistent. Any measurement deviation reflects the systematic error of the 3D vision positioning instrument. Based on the unified data, the consistency error of the virtual rotation axis parameters is calculated, and the measurement accuracy of the 3D vision wheel positioning instrument is judged according to the preset accuracy threshold.

[0043] like Figure 2As shown, this is a three-dimensional schematic diagram of the target plate rotating around a fixed axis in a 3D vision wheel positioning device accuracy detection method provided by an embodiment of the present invention.

[0044] The target board has a known pattern or markings. The initial position of the target board is 110. After rotating around a fixed axis 40 by a certain angle, the position after rotation is 111. The rotation axis 40 intersects the plane of the target board at point 41.

[0045] like Figure 3 As shown, this is a schematic diagram illustrating the virtual rotation axis construction principle in a 3D vision wheel positioning device accuracy detection method provided by an embodiment of the present invention.

[0046] The calibration rod 20 is a rigid rod; the left and right target plates 11' and 11" are fixed to both ends of the calibration rod 20, and there are known patterns or markings on the plates; the calibration rod 20 rotates around the fixed axis 40; the left and right cameras 12' and 12" of the 3D vision wheel positioning instrument observe the left and right target plates 11' and 11" respectively; C0 is the global measurement coordinate system, C1' is the coordinate system of the left target plate, C1" is the coordinate system of the right target plate, C2' is the coordinate system of the left camera, and C2" is the coordinate system of the right camera;

[0047] In step 2, with the calibration rod 20 in its initial position, the left and right cameras 12' and 12" respectively acquire images of the initial positions 11'0 and 11"0 of the left and right target plates 11' and 11" and calculate the initial pose of the left target plate 11' in the left camera coordinate system C2'. Calculate the initial pose of the right target plate 11” in the right camera coordinate system C2”. ;

[0048] The calibration rod 20 is rotated around the fixed axis 40 by a certain angle. The left and right cameras respectively acquire images of the rotated positions 11'1 and 11"1 of the two target plates 11' and 11" on the left and right sides. The rotated pose of the left target plate 11' in the left camera coordinate system C2' is calculated. Calculate the rotated pose of the target plate 11” on the right side in the right camera coordinate system C2”. ;

[0049] Furthermore, in step 3, the virtual rotation axis 40', 40" of the rotation axis 40 in the respective coordinate systems C1', C1" of the left and right target plates 11', 11" and the reference point 41', 41" on the rotation axis 40 can be calculated. The reference point 41', 41" on the rotation axis 40 can be the point where the rotation axis 40 intersects with the plane of the left and right target plates 11', 11" or the reference point 41', 41" can be determined at other positions near the target plates 11', 11".

[0050] Taking the left camera 12' and the left target plate 11' as an example, the following is a process for calculating the virtual rotation axis 40' and the reference point 41' on the rotation axis in the coordinate system C1' of the left target plate 11':

[0051] In the left camera coordinate system C2', the initial pose of the left target plate 11' is... Post-rotation pose ,

[0052]

[0053]

[0054] in, , It is a 3×3 rotation matrix. , It is a 3×1 translation vector;

[0055] In the left camera coordinate system C2', the rotational transformation of the left target plate 11 from its initial position 11'0 to its rotated position 11'1. for:

[0056]

[0057]

[0058] in, It is a 3×3 rotation matrix. It is a 3×1 translation vector; this transformation The rotational motion about the fixed axis 40, expressed in the left camera coordinate system C2', can be further obtained as follows:

[0059]

[0060]

[0061] The vector representation of it is its rotation axis vector. ;

[0062] Because, when rotating around a fixed axis, the point on the axis of rotation The coordinates remain unchanged, so the following condition is met:

[0063]

[0064] Then transform it from the left camera coordinate system C2' to the left target board 11' coordinate system C1';

[0065] The transformation from the target board coordinate system C1' to the camera coordinate system C2' is already in place:

[0066]

[0067] Rotation axis vector Since it is a free vector, it is only affected by rotation, so the rotation axis vector in the target plate coordinate system C1' is:

[0068]

[0069] Points on the axis of rotation It is a position vector, affected by rotation and translation, so the point on the rotation axis in the target plate coordinate system C1' is:

[0070]

[0071] The parametric equations of the rotation axis in the target plate coordinate system C1' are:

[0072]

[0073] Then, the reference point 41' is calculated in the target plate coordinate system C1'. Here, the calculation takes the intersection of the rotation axis 40' and the plane of the target plate 11' as the reference point 41'. z=0, that is, the component in z is 0, so:

[0074]

[0075] in, express The component in z, solve for s:

[0076]

[0077] in, That is, the axis of rotation cannot be parallel to the plane of the target plate;

[0078] Substituting s back into the equation of rotation axis parameters, we obtain the coordinates of reference point 41' in the target plate coordinate system C1':

[0079]

[0080] Of course, reference point 41' can also be selected from other points on the axis of rotation;

[0081] This allows us to calculate the virtual rotation axis 40' and the reference point 41' on the rotation axis in the coordinate system C1' of the left target plate 11', respectively, using the rotation vector. and points Similarly, the virtual rotation axis 40” and the reference point 41” on the rotation axis in the coordinate system C1” of the right target plate 11” can be calculated respectively using the rotation vector. and points express;

[0082] Once the virtual rotation axes 40', 40” are constructed, their positional relationship relative to the target plate 11', 11” remains fixed throughout the entire measurement process. Therefore, in step 5 described later, at each placement position, the pose of the virtual rotation axes 40 in the global measurement coordinate system C0 can be determined based on the pose of the target plate 11', 11”.

[0083] In the global measurement coordinate system C0, the pose of the left target plate 11' is as follows:

[0084]

[0085] Position of target plate 11 on the right:

[0086]

[0087] Rotation axis vector It can then be directly converted into the rotation axis vector in the global measurement coordinate system C0:

[0088]

[0089] Reference point It can then be directly converted into a reference point in the global measurement coordinate system C0:

[0090]

[0091] Rotation axis vector This can be directly converted into the rotation axis vector in the global measurement coordinate system C0 as follows:

[0092]

[0093] Reference point It can then be directly converted into a reference point in the global measurement coordinate system C0:

[0094]

[0095] Furthermore, based on the reference points at both ends of the rotation axis 40 , A rotation axis vector can also be calculated:

[0096]

[0097] And the length of the axis of rotation:

[0098]

[0099] Using the methods described above, we can use , , and Four different attributes are used to describe the same virtual axis of rotation. Ideally, the calibration rod is calculated at one position. , , They should be collinear, the length of the axis of rotation The measurements are the same regardless of the location.

[0100] In some embodiments, in order to ensure that the established virtual rotation axis is more stable, steps 2 and 3 involve rotating the calibration rod around the same fixed axis at least twice, and fitting and optimizing the spatial parameters of the virtual rotation axis based on the pose data of multiple rotations. At the same time, it can be determined whether the virtual rotation axis is shaking, which can further reduce the error of the established virtual rotation axis spatial parameters.

[0101] like Figure 4 The diagram shown illustrates how the virtual rotation axis does not rotate along the central axis of the calibration rod in a 3D vision wheel positioning device accuracy detection method provided by an embodiment of the present invention.

[0102] Because the rotating axes 40', 40” and the reference points 41', 41” on the rotating axes are only virtual axes 40 and points on them that were rotated out during this test, and do not necessarily have to be the axis center of the calibration rod 20 and the points on the axis center, the accuracy is not significantly related to the machining and assembly accuracy of the calibration rod 20.

[0103] If the machining and assembly accuracy of the calibration rod 20 is not high, and its straightness and coaxiality are not good, the virtual rotation axis 40 around which the calibration rod 20 is rotated may not be the central axis of the calibration rod 20. The left and right cameras 12 of the 3D vision wheel positioning instrument 10 also acquire images of the left and right target plates 11', 11" before and after rotation, and calculate the poses of the target plates 11', 11" before and after rotation. Similarly, the virtual rotation axis 40', 40" under the coordinate system C1', C1" of the left and right target plates 11', 11" and a reference point 41', 41" intersecting the plane PLN', PLN" which has a fixed positional relationship with the left and right target plates 11', 11" on the rotation axis can be calculated using the same method. The plane PLN', PLN" is the plane with a fixed positional relationship with the left and right target plates 11', 11". It can be the plane of the left and right target plates 11', 11" or other planes, such as the clamping plane formed by the clamping jaws.

[0104] like Figure 5 The diagram shown illustrates the arbitrary horizontal and cross-directional movement of the calibration rod to several positions in a 3D vision wheel positioning device accuracy detection method provided in this embodiment of the invention.

[0105] In step 4, the calibration rod 20 is moved horizontally and crosswise to n different positions P1, P2…P nEach position ensures that the two target plates 11' and 11" are within the field of view of the left and right cameras 12 of the 3D vision wheel positioning instrument 10 and can be accurately identified.

[0106] In step 5, at each position, the 3D vision wheel positioning instrument 10 acquires the poses of the two target plates 11', 11" on the left and right sides of the calibration rod 20, respectively. The rotation axes 40', 40" and reference points 41', 41" calculated in step 3 within their respective coordinate systems C1', C1" of the left and right target plates 11', 11" are unified to the same global measurement coordinate system C0, resulting in n sets of rotation axis vectors and rotation axis lengths.

[0107] , , , ,

[0108] Where i represents the i-th placement position of calibration rod 20. , , , These represent the rotation axis vector calculated for the left target plate, the rotation axis vector calculated for the right target plate, the rotation axis vector calculated by the reference points at both ends of rotation axis 40, and the rotation axis length calculated by the reference points at both ends of rotation axis 40, respectively, under the global measurement coordinate system C0.

[0109] Because the measurement principle of the 3D vision wheel alignment instrument does not rely independently on various tilt sensors and other electronic devices, its accuracy is mainly determined by a series of calibration parameters of multiple cameras 12. Through these calibration parameters, the coordinate systems of each camera can be unified under a global measurement coordinate system C0. Therefore, by verifying the error of the angle and length of the virtual axis 40, it is possible to determine whether the calibration parameters of multiple cameras 12 are accurate, without having to verify each angle that the wheel alignment instrument needs to measure one by one. By judging whether the virtual rotation axis 40 is globally consistent, it is possible to judge whether the angle measurement is accurate. The globally consistent length of the virtual rotation axis 40 placed front and back and intersecting indicates that if a standard rectangle exists, the measurement result of the alignment instrument 10 is also a standard rectangle. Thus, it is not necessary to mechanically construct a precise rectangular zero-position frame or test platform.

[0110] In step 6, the consistency of the angles and distances between the virtual rotation axes 40',40" and the reference points 41',41" on the left and right target plates 11',11" under the same global measurement coordinate system C0 is calculated. Under the same global measurement coordinate system C0, the angle between the rotation axes 40', 40" and the vector 40',40" at each position should theoretically be 0° or 180°. The distance ||41'41"|| between 41',41" at each position should be the same. If the angle between the rotation axes 40', 40" and the vector 40',40" at each position is not 0°, it indicates that there is an error in the angle measurement. If the distance ||41'41"|| between 41',41" at each position is not the same, it indicates that there is an error in the dimension measurement.

[0111] Specifically, for the i-th placement position, calculate respectively... , The included angle between:

[0112]

[0113] Calculate separately , The included angle between:

[0114]

[0115] Calculate separately , The included angle between:

[0116]

[0117] Set the threshold for radian representation ,when ,and ,and If the angle accuracy at that position is within the acceptable range, it is considered acceptable; otherwise, it is considered unacceptable.

[0118] If the angular accuracy at all positions meets the requirements, then the global angular accuracy is considered to meet the requirements; otherwise, it is considered unqualified.

[0119] For the length of the virtual axis at all positions, simply calculate the difference between the minimum and maximum values.

[0120]

[0121] Set threshold ,when The length accuracy is considered to be within the acceptable range; otherwise, it is considered unacceptable.

[0122] In some embodiments, in step 6, in addition to simply using the absolute error value as the basis for error judgment, mean square error, root mean square error, etc. can also be used as the basis for error judgment; the angle error is calculated in radians, and in order to conform to the habits of positioning device users, the consistency error can be converted into fractions, degrees, or a custom error type for output.

[0123] In some embodiments, when the rotating shaft 40 approaches the axle direction of the vehicle positioning measurement, the calibration rod 20 is moved to a different position on the vehicle positioning measurement plane, and the measurements P1, P2…P at each position are compared. n When the two target plates 11', 11" are consistent in terms of the angle and distance between the virtual rotation axes 40', 40" and the reference points 41', 41" on the rotation axes under the unified global measurement coordinate system C0, the corresponding angles can be converted and displayed as errors in the toe-in and camber directions, and the distance error can be converted into the measurement error of the wheel track at the front and rear positions. This makes it easier for users to intuitively judge whether the accuracy of the 3D vision wheel alignment instrument 10 is within the allowable error range.

[0124] like Figure 6 As shown in the figure, a schematic diagram of a 3D vision wheel positioning device accuracy detection device provided in an embodiment of the present invention is provided. The device includes: a 3D vision wheel positioning device 10, a calibration rod 20, and a data processing module 30.

[0125] In the schematic diagram of this example device, the 3D vision wheel positioning device 10 uses a camera 12 to acquire images of a target plate 11 fixed to the wheel rim by a clamp 15 for wheel positioning measurement. The device includes multiple cameras 12, multiple target plates 11, a processing unit 13 and a corresponding connecting and fixing bracket 14. It has been calibrated and has initial calibration parameters, and can obtain the pose of the target plate 11.

[0126] In the schematic diagram of this example device, the calibration rod 20 is a rigid, movable rod that can rotate around a fixed axis. The two ends of the rod have connectors that can rigidly fix the left and right target plates 11', 11" of the 3D vision wheel positioning instrument to the calibration rod 20 respectively. The length of the calibration rod 20 must be sufficient to ensure that the left and right target plates 11', 11" can appear in the field of view of the cameras 12 on both sides of the 3D vision wheel positioning instrument after being rigidly connected. The calibration rod 20 must remain stationary when the cameras 12 acquire images of the target plates 11', 11";

[0127] In the schematic diagram of this example device, the data processing module 30 is configured to perform the calculation and judgment steps as described in claim 1.

[0128] like Figure 7The diagram shown is a schematic of the calibration rod after segmented assembly in a 3D vision wheel positioning device accuracy detection device provided in an embodiment of the present invention.

[0129] The calibration rod 20 is rigid to ensure that it will not deform during the detection process. It is not necessary for the rotation axis 40 and the axis of the calibration rod 20 to be coaxial. There is no strict accuracy requirement for the straightness of the calibration rod 20. The calibration rod 20 can be integral or segmented and rigidly assembled by connecting parts. The calibration rod 20 needs to be able to rotate around a certain fixed axis 40. When the camera 12 acquires the image of the target plate 11, the calibration rod should be able to remain stationary. In order to facilitate rotation and fixation, brackets 21 can be provided at both ends of the calibration rod 20 to facilitate rotation and fixation.

[0130] like Figure 8 The diagram shown illustrates how the calibration rod fixes the target plate via an adapter plate in a 3D vision wheel positioning device accuracy testing apparatus provided in this embodiment of the invention.

[0131] The connection between the calibration rod 20 and the target plate 11 is required to be a rigid fixed connection. After the connection, the two target plates 11 can be photographed by the left and right cameras 12 of the positioning instrument 10 at the same time. There are no restrictions on the structure and angle of the connecting parts. The target plate 11 can be directly fixed to the calibration frame 20, or the positioning instrument target plate 11 together with the clamp 15 can be fixed to the calibration rod through an adapter plate 22.

[0132] In some embodiments, the accuracy detection method of the 3D vision wheel alignment device is characterized in that the method can use one calibration rod 20 to replace a certain row of wheel axles of the vehicle, or use multiple calibration rods 20 to replace multiple rows of wheel axles of the vehicle, to simulate the vehicle positioning measurement process and measure whether its toe-in, camber, body size, etc. are accurate. The virtual rotation axis 40 obtained by rotating the calibration rod 20 corresponds to the left and right wheel axles, and its total toe-in and camber should be 0.

[0133] Those skilled in the art will also understand that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements to the technical solutions of the present invention made by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for detecting the accuracy of a 3D vision wheel aligner device, characterized in that, The method comprises: Step 1: fixing and connecting the left and right target plates to the two ends of the calibration rod, so that the left and right target plates are respectively located in the visual range of the left and right side cameras of the 3D visual wheel alignment instrument; Step 2: rotating the calibration rod around the fixed shaft, and the left and right side cameras of the 3D visual wheel alignment instrument respectively acquire images of the left and right target plates before and after rotation, and calculate the poses of the target plates before and after rotation; Step 3: by rotating around the fixed shaft, even if the straightness of the calibration rod is not good, a virtual rotation axis of an absolute standard straight line can be constructed, and the spatial parameters of the virtual rotation axis in the coordinate system of each target plate are calculated according to the poses of the target plates before and after rotation; Step 4: moving the calibration rod to multiple different positions, and ensuring that the left and right target plates are respectively located in the visual range of the left and right side cameras at each position; Step 5: at each position, the left and right side cameras respectively acquire the poses of the left and right target plates, and the spatial parameters of the virtual rotation axis in the coordinate system of each target plate are unified to the global measurement coordinate system; Step 6: based on the unified data, calculating the consistency error of the virtual rotation axis parameters, and judging the measurement accuracy of the 3D visual wheel alignment instrument according to a preset accuracy threshold.

2. The method of claim 1, wherein, The steps 2 and 3 calculate the spatial parameters of the virtual rotation axis by rotating the calibration rod around the fixed shaft once.

3. The method of claim 1, wherein the 3D vision wheel aligner device precision detection method is characterized by, The steps 2 and 3 calculate the spatial parameters of the virtual rotation axis by rotating the calibration rod around the same fixed shaft at least twice, and fitting and optimizing the spatial parameters of the virtual rotation axis according to the pose data of multiple rotations.

4. The method of claim 1, wherein, The spatial parameters of the rotation axis in the coordinate system of each target plate in the step 3 include a direction vector of the rotation axis and a reference point of the rotation axis at the end of the target plate.

5. The method of claim 4, wherein, The reference point of the rotation axis at the end of the target plate is the intersection coordinate of the rotation axis and the target plate plane.

6. The method of claim 4, wherein, The reference point of the rotation axis at the end of the target plate is the intersection coordinate of the rotation axis and a certain fixed plane (such as a clamp plane) in the coordinate system of the target plate.

7. The method of claim 1, wherein, In the step 6, the consistency error is converted into a self-defined error type for output, and the error type includes at least one of a geometric parameter error and a vehicle positioning parameter error.

8. The method of claim 7, wherein, The geometric parameter error includes at least one of a direction vector angle error and a length error of the rotation axis.

9. The method of claim 7, wherein, The vehicle positioning parameter error includes at least one of a toe angle error, a camber angle error, and a wheelbase error.

Images