Wheel positioning parameter detection method and system based on face light field wheel rim feature extraction

Abstract

The application discloses a kind of based on face light field wheel rim feature extraction wheel positioning parameter detection method and system, to solve the problem of automobile wheel positioning parameter detection based on face light field wheel rim feature extraction.The automobile wheel positioning parameter detection method based on face light field wheel rim feature extraction mainly includes five steps: image acquisition of camera (2) calibration, the coordinates of laser plane under the camera (2) coordinate system are solved, the three-dimensional feature points of laser feature points under the camera (2) coordinate system are solved, automobile wheel rim space circle fitting and solving circle center and normal vector and solving automobile wheel positioning parameter composition.A kind of based on face light field wheel rim feature extraction wheel positioning parameter detection method and system can be used for non-contact, stable performance.

Description

Technical Field

[0001] This invention relates to a measurement method and measuring equipment in the field of automobile inspection, and more specifically, it is a method and system for detecting wheel alignment parameters based on surface light field wheel rim feature extraction. Background Technology

[0002] Automotive inspection technology plays a crucial role in ensuring vehicle safety, improving vehicle reliability and durability, and enhancing driving comfort and experience. The significance of visual inspection of vehicle wheel alignment parameters lies in ensuring vehicle safety and stability. Wheel alignment parameters significantly impact a vehicle's handling, driving, and braking performance. Deviations in these parameters can lead to problems such as vehicle skidding, accelerated tire wear, increased energy consumption, and a deterioration in driving experience, potentially even jeopardizing driving safety. Visual inspection can promptly detect these deviations and make timely adjustments and corrections, thereby ensuring vehicle safety, stability, and comfort. Currently, the widely used method for detecting vehicle wheel alignment parameters is based on visual 3D four-wheel alignment inspection, which mostly employs two-dimensional target markers. However, using two-dimensional target markers for vehicle wheel alignment parameter inspection suffers from drawbacks such as contact measurement, cumbersome operation, susceptibility to external influences, time-consuming process, and dependence on vehicle type. Therefore, a wheel positioning parameter detection method and system based on surface light field rim feature extraction is proposed. It adopts an independent camera and surface laser, eliminating the need for a two-dimensional target marking plate, and realizes accurate detection of wheel geometric elements. This forms an active visual detection method for automotive wheel positioning parameters that can directly and non-contactly detect wheel rims, is highly reliable, can be calibrated on-site, and has traceable detection values. Summary of the Invention

[0003] This invention addresses the problems of contact measurement, cumbersome operation, susceptibility to external influences, and high time consumption in the detection of automotive wheel alignment parameters using two-dimensional target markers. It proposes a lightweight, flexible, and simple system that achieves high precision and eliminates the need for two-dimensional target markers by extracting surface light field rim features. The system consists of a camera and a surface laser. By detecting the intersection of the actively projected laser plane family with the wheel rim, it optimizes the calculation of wheel rim geometric elements such as the rim ellipse, the spatial coordinates of the rim center, and the normal vector, achieving high-precision, direct, non-contact visual measurement of automotive wheel alignment parameters.

[0004] Referring to the accompanying drawings, the present invention is implemented using the following technical solution:

[0005] The specific steps of the wheel positioning parameter detection method based on surface light field rim feature extraction are as follows:

[0006] Step 1: Place the camera bracket on the ground and fix the camera on the camera bracket. According to the needs of the detection range of the car wheel alignment parameters, fix the position of the camera. The camera captures target images of the planar target at different positions. The camera parameters are calibrated by the DLT calibration method to obtain the intrinsic parameters K of the camera.

[0007] Step 2: Place the surface laser directly in front of the wheel rim, ensuring the laser plane can hit the rim. Turn on the surface laser and move the planar target into the camera's field of view. Use the camera to capture i sets of target images, one with laser lines and one without. Use the difference image method to determine the position of the laser feature points in the target images. In the camera coordinate system, the laser plane determined by the laser emitted by the k-th (k = 1, 2, 3, 4, 5) surface laser at the i-th position is π. i，k ;

[0008] The image coordinates x of the center feature point of the laser stripe formed by the intersection of the planar target and the planar laser plane are extracted using the Hessian matrix method. i，k Let the coordinates of the center feature point of the laser stripe in the camera 2 coordinate system be... The coordinates in the target coordinate system are: The transformation relationship between the target coordinate system and the image coordinate system satisfies

[0009]

[0010] Among them, P i =K[R i t i K is the camera's intrinsic parameters obtained from the first step of calibration, and R is the camera's extrinsic parameters obtained from QR decomposition. i t i ;

[0011] Based on the transformation relationship between the target coordinate system and the camera coordinate system, the center point of the laser stripe in the target coordinate system is determined. Transform to the camera coordinate system, and find the coordinates of the laser bar's center point in the target coordinate system. Coordinates of the center point of the laser strip in the camera coordinate system satisfy

[0012]

[0013] Due to the coordinates of the laser center point in the camera coordinate system On the plane of light, therefore satisfying

[0014]

[0015] Based on the calculated coordinates of the laser center point in the camera coordinate system And using the above equation, the coordinates π of the laser plane in the camera coordinate system are calculated using the SVD decomposition method. i，k ;

[0016] Step 3: When the laser is activated and the wheel rim is within the camera's field of view, based on the image captured by the camera, the laser feature points on the edge of the wheel rim are l. k，m m (m = 1, 2, ... 10) represents the m-th feature point on the k-th laser plane on the edge of the wheel rim. Let l be the three-dimensional coordinates of the laser feature points in the camera coordinate system. Based on the transformation relationship between the image coordinate system and the camera coordinate system, the image coordinates of the laser strip feature points are l. k，m The coordinates of the laser stripe feature points in the camera coordinate system satisfy

[0017]

[0018] Where K is the intrinsic parameter of the camera obtained from the first step of calibration, and the coordinates of the laser stripe feature points in the camera coordinate system. The coordinates of the laser stripe feature points in the image coordinate system. k，m It can be represented as and l k，m =(u i v i ,1) T ;

[0019] Based on laser feature points in the camera coordinate system On the laser plane, therefore satisfying

[0020]

[0021] Among them, the coordinates π of the laser plane in the camera coordinate system obtained in the second step are... i，k = (π1, π2, π3, π4) T Combining the above formulas, we have

[0022]

[0023] The three-dimensional coordinates of the laser strip feature points in the camera coordinate system are obtained using the above formula.

[0024] Step 4: Based on the coordinates of the laser bar feature points in the camera coordinate system Located on the same wheel rim plane, the plane equation is expressed as:

[0025] ax + by + cz - 1 = 0

[0026] Written in matrix form as

[0027]

[0028] Based on the laser bar feature points in the aforementioned camera coordinate system The normal vector N = (a, b, c) of the wheel rim plane obtained by fitting the spatial circle of the wheel rim. T The center O and radius r of the wheel rim circle are obtained by using the least squares method.

[0029] Step 5: Based on the center O of the wheel rim space circle obtained from the above formula, the rotation centers of the four wheels of the car in the global coordinate system can be obtained as O1, O2, O3, and O4. These can be used to determine the reference plane of the car body, where the normal vector OZ axis of the horizontal plane of the car body is...

[0030]

[0031] The centerline of a car body is the line connecting the centers of the front and rear axles, i.e., the reference OY axis of the car body.

[0032]

[0033] The OX axis is the outer product of the normal vector of the horizontal plane of the car body, the OZ axis, and the reference OY axis of the car body.

[0034] n OX =n OY ×n OZ

[0035] During the wheel alignment parameter detection process, based on the wheel rim plane normal vector N obtained in step four, the alignment parameters of the four wheels are calculated by projecting the direction vector N onto the vehicle body reference plane obtained above.

[0036]

[0037] Where θ is when q = 1, 2, 3, 4 q These are the kingpin inclination angle, kingpin caster angle, wheel toe angle, and wheel camber angle, respectively. n1 is the normal vector of the vehicle wheel positioning reference plane. When q = 1 or 2, N represents the kingpin axis of the wheel, and n2 is the normal vector of the horizontal plane of the vehicle body. When q = 3 or 4, N represents the steering knuckle axis of the wheel, and n2 is the normal vector of the longitudinal plane of the vehicle body.

[0038] The wheel positioning parameter detection system based on surface light field rim feature extraction includes a camera bracket, a camera, a laser board box, a surface laser and a planar target;

[0039] The camera bracket and laser board box are placed on the ground. The camera is fixedly connected to the top of the camera bracket by the threaded hole at the bottom and the bolt thread at the top. A set of surface lasers are inserted into a set of through holes in the laser board box and are in tight contact with the surface of the laser board box.

[0040] The camera bracket described in the technical solution is a height-adjustable tripod.

[0041] The camera described in the technical solution is a wide-angle industrial camera.

[0042] The laser plate box described in the technical solution is a part made of steel plate, and a set of through holes are machined on the side of the steel plate.

[0043] The surface laser described in the technical solution is a cylindrical part that can emit laser light from a flat surface.

[0044] The planar target described in the technical solution is a flat plate made of acrylic material, with a checkerboard target paper attached to its outer surface.

[0045] The beneficial effects of this invention are:

[0046] (1) The method of the present invention addresses the problem of detecting automotive wheel positioning parameters. It uses a camera and a surface laser to optimize and calculate the wheel geometric elements such as the wheel rim ellipse, the spatial coordinates of the wheel rim center, and the normal vector, thereby achieving high-precision direct non-contact visual measurement of automotive wheel positioning parameters.

[0047] (2) The method of this invention introduces a surface laser as a bridge between the camera and the wheel. Compared with a clamping device that relies on a rigid connection, the surface light field enhances the flexibility of the vision inspection system with higher flexibility, providing convenience for adjusting the distance and angle of the surface light field during the experiment. This forms an active vision inspection method for automobile wheel positioning parameters that can directly and non-contactly inspect the wheel rim, is highly reliable, can be calibrated on-site, and has traceable detection values.

[0048] (3) The system of the present invention has a wide measurement range, reliable performance, simple structure, easy operation and low cost. Attached Figure Description

[0049] Figure 1 This is a flowchart of a wheel positioning parameter detection method based on surface light field rim feature extraction;

[0050] Figure 2 This is an axonometric view of a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0051] Figure 3 It is a wheel positioning parameter detection system based on surface light field rim feature extraction and an axonometric map of planar target 5;

[0052] Figure 4This is an isometric view of camera bracket 1 in a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0053] Figure 5 This is an isometric view of camera 2 in a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0054] Figure 6 This is an axonometric view of laser plate box 3 in a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0055] Figure 7 This is an axonometric view of the surface laser 4 in a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0056] Figure 8 This is an axonometric view of planar target 5 in a wheel positioning parameter detection system based on surface light field rim feature extraction;

[0057] In the diagram: 1. Camera bracket, 2. Camera, 3. Laser board box, 4. Surface laser, 5. Planar target. Detailed Implementation

[0058] The present invention will now be described in further detail with reference to the accompanying drawings:

[0059] See Figure 1 The wheel positioning parameter detection method based on surface light field rim feature extraction can be divided into the following five steps:

[0060] Step 1: Place camera bracket 1 on the ground and fix camera 2 on camera bracket 1. According to the needs of the detection range of car wheel positioning parameters, fix the position of camera 2. Camera 2 captures target images of planar target 5 at different positions. The parameters of camera 2 are calibrated by DLT calibration method to obtain the intrinsic parameters K of camera 2.

[0061] Step 2: Place the surface laser 4 directly in front of the wheel rim, ensuring the laser plane can hit the wheel rim. Turn on the surface laser 4 and move the planar target 5 into the field of view of camera 2. Use camera 2 to capture i sets of target images with and without laser lines. Use the difference image method to determine the position of the laser feature points in the target images. In the coordinate system of camera 2, the laser plane determined by the laser emitted by the k-th (k = 1, 2, 3, 4, 5) surface laser 4 at the i-th position is π. i，k ;

[0062] The image coordinates x of the center feature point of the laser stripe formed by the intersection of the planar target 5 and the planar laser plane are extracted using the Hessian matrix method. i，k Let the coordinates of the center feature point of the laser stripe in the camera 2 coordinate system be... The coordinates in the target coordinate system are: The transformation relationship between the target coordinate system and the image coordinate system satisfies

[0063]

[0064] Among them, P i =K[R i t i K is the intrinsic parameter of camera 2 obtained from the first step of calibration, and R is the extrinsic parameter of camera 2 obtained from QR decomposition. i t i ;

[0065] Based on the transformation relationship between the target coordinate system and the camera 2 coordinate system, the center point of the laser stripe in the target coordinate system is determined. Transform to the camera 2 coordinate system, and find the coordinates of the laser bar's center point in the target coordinate system. Coordinates of the center point of the laser stripe in the camera 2 coordinate system satisfy

[0066]

[0067] Due to the coordinates of the laser center point in the camera 2 coordinate system On the plane of light, therefore satisfying

[0068]

[0069] Based on the calculated coordinates of the laser center point in the camera 2 coordinate system And using the above equation, the coordinates π of the laser plane in the camera 2 coordinate system are calculated by the SVD decomposition method. i，k ;

[0070] Step 3: When the surface laser 4 is turned on and the wheel rim is within the field of view of camera 2, according to the image captured by camera 2, the laser feature points on the edge of the wheel rim are l. k，m m (m = 1, 2, ... 10) represents the m-th feature point on the k-th laser plane on the edge of the wheel rim. Let l be the three-dimensional coordinates of the laser feature points in the camera 2 coordinate system. Based on the transformation relationship between the image coordinate system and the camera 2 coordinate system, the image coordinates of the laser strip feature points are l. k，m Coordinates of laser stripe feature points in camera 2 coordinate system satisfy

[0071]

[0072] Where K is the intrinsic parameter of camera 2 obtained through the first step of calibration, and the coordinates of the laser stripe feature points in the camera 2 coordinate system. The coordinates of the laser stripe feature points in the image coordinate system. k，m It can be represented as and l k，m =(u i v i ,1) T ;

[0073] Based on the laser feature points in the camera 2 coordinate system On the laser plane, therefore satisfying

[0074]

[0075] Among them, the coordinates π of the laser plane in the camera 2 coordinate system obtained in the second step are... i，k = (π1, π2, π3, π4) T Combining the above formulas, we have

[0076]

[0077] The three-dimensional coordinates of the laser stripe feature points in the camera 2 coordinate system are obtained using the above formula.

[0078] Step 4: Based on the coordinates of the laser bar feature points in the camera 2 coordinate system Located on the same wheel rim plane, the plane equation is expressed as:

[0079] ax + by + cz - 1 = 0

[0080] Written in matrix form as

[0081]

[0082] Based on the laser bar feature points in the coordinate system of camera 2 mentioned above The normal vector N = (a, b, c) of the wheel rim plane obtained by fitting the spatial circle of the wheel rim. T The center O and radius r of the wheel rim circle are obtained by using the least squares method.

[0083] Step 5: Based on the center O of the wheel rim space circle obtained from the above formula, the rotation centers of the four wheels of the car in the global coordinate system can be obtained as O1, O2, O3, and O4. These can be used to determine the reference plane of the car body, where the normal vector OZ axis of the horizontal plane of the car body is...

[0084]

[0085] The centerline of a car body is the line connecting the centers of the front and rear axles, i.e., the reference OY axis of the car body.

[0086]

[0087] The OX axis is the outer product of the normal vector of the horizontal plane of the car body, the OZ axis, and the reference OY axis of the car body.

[0088] n OX =n OY ×n OZ

[0089] During the wheel alignment parameter detection process, based on the wheel rim plane normal vector N obtained in step four, the alignment parameters of the four wheels are calculated by projecting the direction vector N onto the vehicle body reference plane obtained above.

[0090]

[0091] Where θ is when q = 1, 2, 3, 4 q These are the kingpin inclination angle, kingpin caster angle, wheel toe angle, and wheel camber angle, respectively. n1 is the normal vector of the vehicle wheel positioning reference plane. When q = 1 or 2, N represents the kingpin axis of the wheel, and n2 is the normal vector of the horizontal plane of the vehicle body. When q = 3 or 4, N represents the steering knuckle axis of the wheel, and n2 is the normal vector of the longitudinal plane of the vehicle body.

[0092] See Figures 2 to 8 The wheel positioning parameter detection system based on surface light field wheel rim feature extraction includes a camera bracket 1, a camera 2, a laser board box 3, a surface laser 4, and a planar target 5.

[0093] Camera bracket 1 is a height-adjustable tripod bracket. Camera bracket 1 and laser plate box 3 are placed on the ground. Camera 2 is a wide-angle industrial camera. Camera 2 is fixedly connected to the top of camera bracket 1 by a threaded hole at the bottom and a bolt thread. Laser plate box 1 is a part made of steel plate. A set of through holes is machined on the side of the steel plate. A set of surface lasers 4 are inserted into a set of through holes in laser plate box 3 and are in close contact with the surface of laser plate box 3. Surface lasers 4 are cylindrical parts that can emit laser planes. Planar target 5 is a flat plate made of acrylic material with checkerboard target paper attached to the outer surface.

Claims

1. A method for detecting wheel positioning parameters based on surface light field rim feature extraction, characterized in that, The specific steps of the wheel positioning parameter detection method based on surface light field rim feature extraction are as follows: Step 1: Place the camera bracket (1) on the ground and fix the camera (2) on the camera bracket (1). According to the needs of the detection range of the car wheel positioning parameters, fix the position of the camera (2). The camera (2) captures the target image of the plane target (5) at different positions. The parameters of the camera (2) are calibrated by the DLT calibration method to obtain the intrinsic parameters K of the camera (2). Step 2: Place the surface laser (4) directly in front of the wheel rim, ensuring that the laser plane can hit the wheel rim. Turn on the surface laser (4) and move the planar target (5) into the field of view of the camera (2). Use the camera (2) to take i sets of target images with and without laser lines. Use the difference image method to determine the position of the laser feature point in the target image. In the coordinate system of the camera (2), the laser plane determined by the laser emitted by the kth (k=1,2,3,4,5) surface laser (4) at the i-th position is π. i，k ; The image coordinates x of the center feature point of the laser stripe formed by the intersection of the planar target (5) and the planar laser plane are extracted using the Hessian matrix method. i，k Let the coordinates of the center feature point of the laser stripe in the coordinate system of camera (2) be... The coordinates in the target coordinate system are: The transformation relationship between the target coordinate system and the image coordinate system satisfies Among them, P i =K[R i t i K is the intrinsic parameter of camera (2) obtained through the first step of calibration, and the extrinsic parameter R of camera (2) is obtained through QR decomposition. i t i ; Based on the transformation relationship between the target coordinate system and the camera (2) coordinate system, the center point of the laser stripe in the target coordinate system is... Transform to the camera (2) coordinate system, and the coordinates of the center point of the laser stripe in the target coordinate system. Coordinates of the center point of the laser strip in the camera (2) coordinate system satisfy Due to the coordinates of the laser center point in the camera (2) coordinate system On the plane of light, therefore satisfying Based on the calculated coordinates of the laser center point in the camera (2) coordinate system And using the above formula, the coordinates π of the laser plane in the camera (2) coordinate system are calculated by SVD decomposition. i，k ; Step 3: When the surface laser (4) is turned on and the wheel rim is within the field of view of the camera (2), according to the image acquired by the camera (2), the laser feature points of the laser plane on the edge of the wheel rim are l. k，m m (m = 1, 2, ... 10) represents the m-th feature point on the k-th laser plane on the edge of the wheel rim. Given the three-dimensional coordinates of the laser feature points in the camera (2) coordinate system, and based on the transformation relationship between the image coordinate system and the camera (2) coordinate system, the image coordinates of the laser strip feature points are l. k，m The coordinates of the laser stripe feature points in the camera (2) coordinate system satisfy Where K is the intrinsic parameter of the camera (2) obtained through the first step of calibration, and the coordinates of the laser stripe feature points in the coordinate system of the camera (2). The coordinates of the laser stripe feature points in the image coordinate system. k，m It can be represented as and l k，m =(u i v i ,1) T ; According to the laser feature points in the coordinate system of camera (2) On the laser plane, therefore satisfying Among them, the coordinates π of the laser plane in the camera (2) coordinate system obtained in the second step are... i，k = (π1, π2, π3, π4) T Combining the above formulas, we have The three-dimensional coordinates of the laser stripe feature points in the camera (2) coordinate system are obtained according to the above formula. Step 4: Based on the coordinates of the laser bar feature points in the camera (2) coordinate system Located on the same wheel rim plane, the plane equation is expressed as ax + by + cz - 1 = 0 Written in matrix form as Based on the laser bar feature points in the coordinate system of the aforementioned camera (2) The normal vector N = (a, b, c) of the wheel rim plane obtained by fitting the spatial circle of the wheel rim. T The center O and radius r of the wheel rim circle are obtained by using the least squares method. Step 5: Based on the center O of the wheel rim spatial circle obtained from the above formula, the rotation centers of the four wheels of the car in the global coordinate system are obtained as O1, O2, O3, and O4. These are used to determine the reference plane of the car body, where the normal vector OZ axis of the horizontal plane of the car body is... The centerline of a car body is the line connecting the centers of the front and rear axles, i.e., the reference OY axis of the car body. The OX axis is the outer product of the normal vector of the horizontal plane of the car body, the OZ axis, and the reference OY axis of the car body. n OX =n OY ×n OZ During the wheel alignment parameter detection process, based on the wheel rim plane normal vector N obtained in step four, the alignment parameters of the four wheels are calculated by projecting the direction vector N onto the vehicle body reference plane obtained above. Where θ is when q = 1, 2, 3, 4 q These are the kingpin inclination angle, kingpin caster angle, wheel toe angle, and wheel camber angle, respectively. n1 is the normal vector of the vehicle wheel positioning reference plane. When q = 1 or 2, N represents the kingpin axis of the wheel, and n2 is the normal vector of the horizontal plane of the vehicle body. When q = 3 or 4, N represents the steering knuckle axis of the wheel, and n2 is the normal vector of the longitudinal plane of the vehicle body.

2. The detection system of the wheel positioning parameter detection method based on surface light field rim feature extraction according to claim 1 includes a camera bracket (1), a camera (2), a laser board box (3), a surface laser (4), and a planar target (5); The camera bracket (1) and laser board box (3) are placed on the ground, and the camera (2) is fixedly connected to the top of the camera bracket (1) by a threaded hole at the bottom. The camera is characterized by... A set of surface lasers (4) are inserted into a set of through holes in the laser plate box (3) and are in close contact with the surface of the laser plate box (3).

Images