Motor vehicle aligner with improved accuracy and no-stop positioning using driving direction calculation

Abstract

Vehicle alignment systems and methods are disclosed that operate based on calculations of a "drive direction," or the direction in which the vehicle is moving. Since the vehicle can be assumed to be a rigid body, each wheel has the same drive direction. Thus, alignment parameters for one wheel can be compared to the same parameters for another wheel by equating their drive directions, thereby eliminating the requirement for the aligner to "see" both sides of the vehicle simultaneously. An embodiment includes a system having one or more cameras located on a fixture that carries calibration elements for an ADAS system, and one or more targets placed on the vehicle to measure the drive direction of the vehicle. The drive direction is assumed to be parallel to the vehicle thrust line and can be used as a line to orient the fixture to the vehicle.

Description

[0001] Related Applications

[0002] This application claims priority to U.S. Provisional Application No. 62 / 730,727, filed on September 13, 2018, entitled “Wheel Aligner with Improved Accuracy and Stop-Free Positioning Using Drive Direction Calculation,” the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present subject matter relates to machine vision type vehicle aligners. The present disclosure is particularly applicable to wheel aligners for cars and trucks and aligners for calibrating various vehicle sensors. Background Art

[0004] Machine vision vehicle alignment systems (also referred to as "image aligners") using a movable camera and a target attached to the wheel are well known. The target can be seen by the camera so that image data acquired for a given alignment process can be used to calculate vehicle alignment angles for display through a user interface, typically a computer display. Early system implementations included a rigid beam connecting the cameras so that the position and orientation of the cameras relative to each other could be determined and relied upon as not changing. Later system implementations were introduced including the use of cameras that were not rigidly connected to each other, but rather using separate camera / target systems to continuously calibrate the position of one target viewing camera to another target viewing camera. This type of system is described in U.S. Patents 5,535,522; 6,931,340; 6,959,253 and 6,968,282, the entire contents of which are incorporated herein by reference in their entirety. An example of a vehicle wheel aligner using such image processing is the Visualiner 3D or "V3D", which is commercially available from John Bean, Inc. of Conway, Arkansas (a subsidiary of Snap-on Incorporated).

[0005] There is a need for an aligner that includes a camera that can measure quickly and continuously enough to measure, detect, and correct or warn of problems with the alignment measurements. In addition, there is a need for an aligner that measures, detects, or corrects all possible problems that may make the alignment reading incorrect. In addition, current camera-based aligners require a positioning or runout process to measure the wheel axle current (also called the "wheel spindle" or "axis of rotation") and a steering swing process to measure the caster and steering axis inclination (SAI). This process is time consuming and often requires the technician to hold the vehicle steady at certain points of the positioning and caster swing. There is a need for a faster positioning process and a faster caster swing process, where the technician can perform the process without stopping or delaying.

[0006] A camera-based system that addresses these needs is disclosed in U.S. Pat. No. 10,072,926, which is attached hereto and incorporated herein by reference in its entirety. In such a system, at col. 9, lines 38 et seq. of the '926 patent, image data of a target is captured as the wheel and target are continuously rotated through a plurality of degrees of rotation without pausing. The image data is used to calculate at least one pose of the target for every five degrees of rotation as the wheel and target are continuously rotated. At least one of the cameras includes a data processor for performing the steps of preprocessing the image data and calculating alignment parameters of the vehicle based on the preprocessed image data. Thus, alignment-related measurements can be made in rapid succession.

[0007] These conventional camera-based aligners require that the left and right cameras be rigidly mounted to a beam and the relative positions of these cameras need to be known. Other types of conventional camera-based aligners use additional cameras to view a target between the measurement cameras (i.e., across the vehicle) to correlate the relative positions of the measurement cameras. All of these aligners require expensive components, or have structures that are obstructive to use, or both. Therefore, there is a need for an aligner that does not require additional hardware to align the calibration cameras to the relative positions of the cameras, and / or an aligner that does not require a rigid beam between the two cameras to free up space for shops and technicians.

[0008] The system of the '926 patent provides a significant improvement in speed, convenience, accuracy, and functionality over conventional machine vision aligners. However, these systems may not be suitable for aligning wheels on certain vehicles, such as large trucks and other heavy equipment, due to the size of the vehicle and the resulting geometry between the system's wheel mounted targets, the system's cameras, and the amount of wheel rotation required. An aligner is needed to quickly and easily measure large vehicles, ideally as the aligner is driven through tight spaces.

[0009] Advanced driver assistance systems (ADAS) in vehicles assist the driver and provide an additional level of vehicle safety. This includes systems such as adaptive cruise control, lane departure warning, collision avoidance, and rear collision warning. These systems use a combination of sensors (such as cameras, sonar, and radar) that may require calibration to function correctly. Calibration is performed by placing elements such as reflectors, mirrors, targets, etc. in a known relationship to the vehicle. There are conventional systems that use simple manual measurements by drawing a grid on the floor or using a laser line to line up the elements to perform vehicle / carriage orientation. There are also systems that use computerized wheel alignment systems, where the elements are placed on the carriage to orient and position the calibration elements relative to the vehicle. There are many advantages to using a wheel alignment system. However, it is desirable to create a low-cost, simpler system that does not require a full wheel alignment system while still providing a user with guided operations to determine the position and orientation of the carriage and measure and record the position of the calibration elements. Summary of the invention

[0010] The present disclosure provides a vehicle alignment system, the vehicle alignment system comprising a first image sensor for viewing a first target disposed on a first wheel of the vehicle on a first side of the vehicle and for capturing image data of the first target while the vehicle is in motion and the first wheel and the target rotate; a second image sensor for viewing a second target disposed on a second wheel of the vehicle on a second side of the vehicle opposite to the first side of the vehicle and for capturing image data of the second target while the vehicle is in motion and the second wheel and the target rotate. A first gravity sensor is attached to the first image sensor for measuring a sensed orientation relative to gravity on the first side of the vehicle; and a second gravity sensor is attached to the second image sensor for measuring a sensed orientation relative to gravity on the second side of the vehicle. A data processor is provided for performing the following steps: using image data to calculate multiple postures of each of a first target and a second target when the first wheel and target and the second wheel and target rotate; using the calculated posture of the first target to calculate the driving direction of a first side of the vehicle, and using the calculated posture of the second target to calculate the driving direction of a second side of the vehicle; using a measured orientation of a first gravity sensor relative to gravity to calculate the gravity direction of the first side of the vehicle, and using a measured orientation of a second gravity sensor relative to gravity to calculate the gravity direction of the second side of the vehicle; calculating the vehicle driving direction by transforming the driving direction and gravity direction of the first side of the vehicle into a common coordinate system common to the driving direction and gravity direction of the second side of the vehicle; and calculating wheel alignment parameters using the vehicle driving direction.

[0011] According to one aspect of the present disclosure, a third target and a fourth target are attached to a third wheel and a fourth wheel of the vehicle, and an image sensor captures image data of all four targets. A data processor is used to calculate a driving direction of a first side of the vehicle using the calculated poses of the first target and the third target and to calculate a driving direction of a second side of the vehicle using the calculated poses of the second target and the fourth target.

[0012] According to another aspect of the present disclosure, the first image sensor and the second image sensor are mounted to a fixture for carrying an ADAS calibration element, the orientation between the image sensor and the ADAS calibration element is known, and the fixture is located in front of the vehicle. The data processor is used to guide a user to orient the ADAS calibration element to the vehicle driving direction using a user interface.

[0013] Objects and advantages of embodiments of the disclosed subject matter will become apparent from the following description when considered in conjunction with the accompanying drawings.In addition, the different configurations discussed in the following sections can be performed in a different order or simultaneously with each other. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The embodiments will be described in detail below with reference to the accompanying drawings, wherein the same reference numerals represent the same elements. The accompanying drawings are not necessarily drawn to scale. Where applicable, certain functions may not be shown to help describe the basic features.

[0015] Figure 1A is a diagram illustrating a driving direction operation principle according to the present disclosure.

[0016] Figure 1B is a schematic plan view of a vehicle wheel aligner according to an embodiment of the present disclosure.

[0017] 2a-c illustrate vectors associated with the drive direction operating principle according to the present disclosure.

[0018] Figure 3 A method of aligning a lifting device relative to gravity orientation is shown.

[0019] Figure 4A is a schematic plan view of a 3D vision vehicle wheel aligner.

[0020] Figure 4B is a schematic plan view of a 3D vision vehicle wheel aligner according to an embodiment of the present disclosure.

[0021] Figure 5A and 5B is a schematic perspective view of a 3D vision vehicle wheel aligner according to an embodiment of the present disclosure.

[0022] Figure 6 is a schematic plan view showing various wheel alignment parameters of a vehicle.

[0023] Fig. 7A is a schematic top view of an alignment system according to various embodiments.

[0024] Figure 7B yes Fig. 7A Schematic perspective view of the alignment system.

[0025] Figure 7C is a perspective view illustrating an exemplary camera beam in accordance with various embodiments.

[0026] Fig.7D A perspective view of a plurality of different targets according to various embodiments is included

[0027] Fig. 8A is a front perspective view of an ADAS calibration element fixture according to an embodiment of the present disclosure.

[0028] Figure 8B is in use Fig. 8A A floor plan of the fixture to calibrate the vehicle's ADAS sensors.

[0029] Figure 8C Relevant alignment parameters for positioning ADAS calibration elements are shown.

[0030] Fig. 9 is a schematic side view of a driving assistance system according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] It should be understood that the principles described herein are not limited in application to the construction details or component arrangements set forth in the following description or shown in the accompanying drawings. The principles may be embodied in other embodiments and may be practiced or implemented in various ways. In addition, it should be understood that the words and terms used herein are for descriptive purposes and should not be considered restrictive.

[0032] Drive Direction Operating Principle

[0033] The disclosed alignment systems and methods operate based on the calculation of a parameter called "drive direction," which is the direction in which the vehicle is moving. Since the vehicle can be assumed to be a rigid body, each wheel (and each axle) has the same drive direction. Thus, the alignment parameters of one wheel or axle can be compared to the same parameters of another wheel or axle by equating the drive directions of the wheels and axles. For example, by equating the drive directions of each axle, the toe angle of each axle can be compared to the toe angle of the other axle. Thus, the relative toe angle of two axles (i.e., axle offset) can be measured without all the cameras of a conventional vision aligner seeing both axes at the same time or without wheel position or positioning information from one side of the vehicle to the other.

[0034] The basic concept of drive direction alignment is to measure geometrical properties of interest for wheel alignment without directly measuring lateral (i.e., "left to right") position or orientation information about system components. Instead, the disclosed aligner measures information indirectly that couples measurements from the left and right sides, allowing measurements from one side of the vehicle to be transformed into a common coordinate system common to measurements from the other side of the vehicle. This can be accomplished by measuring two or more directions jointly from both sides of the vehicle.

[0035] Will refer to Figure 1A -B explains the basic principle. In the illustrated embodiment, two common directions measured on both the left and right sides of the vehicle 30 are the driving direction and the gravity direction. The driving direction DDL, DDR is measured from a calibrated camera 10L, 10R on each side of the vehicle 30, and the gravity direction GDL, GDR is measured from a calibrated inclinometer 102L, 102R rigidly coupled to each camera 10L, 10R. Alternatively, the inclinometer 102L, 102R can be coupled to any fixed target. The transformation of each inclinometer measurement to its coupled camera is known from a previous calibration, which is described below.

[0036] Gravity direction measurement

[0037] exist Figure 1A In the embodiment shown in FIG. 1-B, gravity is measured by inclinometers 102L, 102R located in each of the left and right sides of the vehicle 30. Depending on the type of inclinometer used, those skilled in the art will appreciate that the measurement output may be in a 3D vector format representing the direction of gravity in the inclinometer coordinate system, or may be represented as a set of (θX, θY) rotation angles describing the inclination about the (X, Y) axis of the inclinometer. If the output is a 3D vector describing the gravity vector in the inclinometer coordinate system, it may be used directly in the processing chain.

[0038] If the output format is a set of (θX, θY) tilt angles, these angles must be transformed into a 3D gravity vector for use in the above processing chain. This can be achieved in a number of ways. In one embodiment, an initial vector representing the gravity orientation in the inclinometer coordinate system is encoded as a 3D vector X=0, Y=0, Z=1. This 3D vector is then rotated around the inclinometer X axis by a rotation angle θX. The rotated 3D vector is then rotated around the inclinometer Y axis by a rotation angle θY. Now, assuming the inclinometer is in a (θX, θY) tilt angle, this rotated 3D vector describes the gravity orientation in the inclinometer coordinate system and can be used in the processing chain described.

[0039] The above discussion assumes that a three-dimensional wheel alignment process is performed. However, the novelty of the present disclosure is not limited to pure 3D alignment. It may be desirable to perform a 2D alignment measurement. In this case, gravity is not measured as a 3D vector or a set of 2D angles, but as a pitch angle from a single axis sensor. Under this construction, it is assumed that all tilts between the cameras are in the direction of the vehicle camber. The measured tilt angles on both sides of the vehicle are then used to adjust the relative left-right tilt angles of the cameras on both sides of the vehicle. This relative tilt angle between the two sides of the vehicle is then used as an offset to measure the camber angle on both sides of the vehicle relative to a common reference frame. Deviations in the driving direction measurements of both cameras in the camber direction will be ignored.

[0040] Transformation of gravity direction from inclinometer to camera coordinate system

[0041] On both sides of the vehicle 30, the gravity direction and the driving direction must be expressed in a common coordinate system. This means that geometric quantities measured in one coordinate system must be transformed into the same coordinate basis so that they can be used in downstream calculations. Figure 1A In the system shown in Figure 1-B, this is achieved by transforming the measured gravity directions GDL, GDR from each inclinometer coordinate system to its rigidly coupled camera coordinate system. This well-known transformation requires a calibration that quantifies how the measurements from the inclinometer coordinate system are transformed into the camera coordinate system. The calibration describes how to rotate from the inclinometer coordinate system to the camera coordinate system. At run time, the measured 3D gravity vector in each inclinometer coordinate system is multiplied by the inclinometer to camera coordinate system rotation matrix. The end result is that the gravity measured in the inclinometer coordinate system can now be expressed in the camera coordinate system on each side of the vehicle.

[0042] Vehicle driving direction measurement

[0043] exist Figure 1A In the embodiment shown in -B, the driving direction DDL, DDR is measured on each side of the vehicle using a respective camera 10L, 10R. The camera can be used to measure the driving direction in a variety of ways, but in Figure 1A -B shows a system in which a set of targets are attached to the wheels of a vehicle. Figure 1B , targets 80a-d are shown attached to vehicle wheels 22L, 22R, 24L, 24R. Conventional targets consist of a set of reference points of known geometry. As the vehicle moves through a minimum distance, the reference points are identified and located in a series of camera images. At each image where the target is visible, the 3D position of the target is calculated as a monocular pose estimate in each camera coordinate system by a process known to those skilled in the art.

[0044] After the running motion is completed, the measured position of the target at all positions is used to calculate the optimal driving direction. In order to calculate the driving direction, the target position must be measured in at least two different vehicle running positions. Depending on the placement of the target on the running vehicle, some orthogonalization of the measured target coordinates may be required. If the target is imaged while attached to the frame or body of the vehicle or the center of the wheel, the target should move in a straight line. However, if the target is positioned on the vehicle wheel, for example, the target will typically draw a cycloidal trajectory. For this case, the direction of the best fit line through the center of the target will depend on the phase angle of the target on the wheel at each data acquisition position. In other words, the target will oscillate with some translational component in a direction orthogonal to the true vehicle driving direction.

[0045] Deviations from the true vehicle drive direction can be subtracted from the measurement target position by reference to an external measurement that is generally orthogonal to the vehicle direction. For example, by using the gravity plane or the plane along which the vehicle operates, the normal to the gravity plane or the plane of operation can be used as the direction to remove the orthogonal component of the target vibration. This reduces uncontrolled variability in the vehicle drive direction measurement, resulting in a more accurate and repeatable drive direction measurement.

[0046] Once the target positions have been orthogonalized (if necessary) as described above, the array of 3D center positions is used as input to a well-known least squares calculation algorithm. The optimal driving direction is calculated using the least squares method to determine the target motion principal directions on each side of the vehicle. The final result of this calculation, performed independently for the left and right sides, is the vehicle driving direction DDL, DDR measured in each of the left camera 10L and right camera 10R coordinate systems.

[0047] It must also be noted that for vehicles with front wheel steering (either due to the front wheels turning, or because the individual toe angles are severely out of specification), the target imaged while attached to the front wheels will experience a slightly different trajectory. This problem is exacerbated when the run distance is greater and the vehicle is turned through a larger semi-circle. However, for shorter run distances, the effect of the steering angle should be quite limited.

[0048] In cases where vehicle steering is not negligible, steering effects can be detected and compensated in various ways. One approach is to calculate the axis of rotation between successive positions of a wheel-mounted target in running motion and use the deviation of the wheel axle from the wheel running angle to determine the steering axis and steering angle. Using the steering axis and angle, the nonlinear target trajectory can be corrected independently on each side of the vehicle to obtain steering-adjusted driving direction.

[0049] Calculate left-to-right rotation from common direction measurements

[0050] The problem of determining the best rotation between the left and right camera coordinate systems is an instance of what is known to those skilled in the art as the Wahba problem. The basic question for this approach is: given that two or more directions are measured in an initial coordinate system and those same directions are measured in a second coordinate system, what is the best rotation between the two coordinate systems? This problem can be solved in a number of ways. If the number of common directions measured in the two coordinate systems is exactly two, the so-called Triad method can be used to solve for the best rotation between the two coordinate systems. For two or more common measurements in two coordinate systems, more general solution methods such as the Kabsch algorithm, Davenport's Q method, and other computational algorithms are used to determine the best rotation between the coordinate systems. The details of these methods vary, but the essence of all of them is to solve for the rotation that minimizes the least squares error when rotating from one coordinate system to another. Most methods involve a singular value decomposition of the 3D covariance matrix of the corresponding 3D vector pairs.

[0051] like Figure 1A As shown, the two common directions on the left side of the vehicle 30 are the vehicle drive direction DDL in the left camera 10L coordinate system and the gravity direction GDL in the left camera 10L coordinate system (initially measured in the left inclinometer coordinate system, but transformed to the left camera coordinate system using the previously described inclinometer to camera calibration). The vehicle drive direction and gravity direction are similarly measured in the right camera 10R coordinate system. These vector pairs DDL, GDL and DDR, GDR on each side of the vehicle 30 are then input to a least squares rotation estimation algorithm, and the 3D rotation of the vector from the left side of the vehicle to the right side (or vice versa) is the output. The 3D rotation can be represented as a 3x 3 rotation matrix, a 3-element vector of Euler angles, a 4-element unit quaternion, or other rotation representation. The specific format is not important, and the various representations can be interchangeable.

[0052] At least two unique directions are required

[0053] It must be emphasized that two or more unique common directions are required to compute a unique 3D rotation between two coordinate systems. In the case where there is no common direction between the two coordinate systems, there is simply no information to constrain the rotation between them. In the case where there is only one common direction between the two coordinate systems, there is not enough information to determine a unique rotation between the coordinate systems.

[0054] It is also important to emphasize that the two or more common directions used to determine the best rotation between coordinate systems must be unique directions. If two directions are parallel, they will actually point in the same direction. The more unique the directions, the better. Ideally, the common directions are orthogonal or nearly orthogonal. The more orthogonal the directions are to each other, the more unique information is incorporated into the calculation of the best left-right rotation solution.

[0055] Alternative Embodiments of the Invention

[0056] The above embodiment uses cameras 10L, 10R and inclinometers 102L, 102R to measure the vehicle driving direction and gravity direction, respectively. However, the basic principle of associating two coordinate systems based on the measurement of two or more common directions can be extended in various ways.

[0057] The operating principle is not limited to the vehicle driving direction and gravity direction

[0058] The disclosed "drive direction" aligner uses the drive direction of the vehicle and the direction of gravity as common directions to measure on both sides of the vehicle. However, the core concept of determining the relative left and right sides of the vehicle does not require these two directions. Alignment measurements may be performed in the manner described using any two or more common directions. For example, a magnetometer may be used to measure the direction to the magnetic north pole as the common direction, which is the same on both sides of the vehicle (for all practical purposes). Another sensor that may be employed is a gyroscope, where a left gyroscope and a right gyroscope are configured to measure a common direction. These are just some examples of other ways a common direction may be measured on both sides of a vehicle.

[0059] Using more than two common directions

[0060] In the described measurement system, two corresponding directions are measured on both sides of the vehicle to determine the transformation from left to right. However, the number of corresponding directions does not have to be limited to two. Any number of corresponding directions can be used to determine the left-to-right orientation. The calculation algorithm employed is not limited to two common directions as long as the additional common directions are not parallel and thus provide supplementary information to restrict the optimal solution.

[0061] Use only one common direction in a reduced-function system

[0062] As mentioned above, at least two 3D common directions are required to determine a unique 3D rotation between the left and right sides of the vehicle. However, if only one corresponding direction is measured on the left and right sides of the vehicle, some functionality of the system can be retained. For example, a 2D rotation can be determined from only one common measurement direction. This may be useful, for example, in situations where wheel alignment measurements need to be made in a strictly 2D operating mode.

[0063] Using alternative gravity measurement sensors and methods

[0064] As described above, measurements of the direction of gravity are made on both sides of the vehicle using conventional inclinometers. However, there are a variety of other ways in which the direction of gravity may be measured without using an inclinometer. An accelerometer may be used in place of an inclinometer to measure the direction of gravity. A plumb line or similar free hanging mass may also be used to provide a measure of the direction of gravity. If the camera itself can be fixed so that it does not rotate relative to the plane of the vehicle's running surface, an a priori calibration step may be performed to determine the normal of the running surface in each of the left and right camera coordinate systems. This normal direction may then be used to provide a common reference direction for both sides of the vehicle.

[0065] Use pre-existing vehicle features instead of or in addition to targets

[0066] Feature Points

[0067] In the embodiments described herein, a target of a predetermined geometry is affixed to the vehicle and measured with a camera to determine the vehicle driving direction. However, since there are many ways in which a reference target is not required to determine the 3D driving direction, a target is not required. One example is the use of stereo vision techniques. For example, if stereo cameras are used on each side of the vehicle, texture feature points can be matched across all cameras in each stereo camera array to provide a 3D position measurement. These feature points can then be tracked and used in a manner similar to a target of a predetermined geometry while the vehicle is in motion.

[0068] Other techniques besides stereo vision can be used to measure vehicle driving direction without using a target with a predetermined geometry. Structured light projection techniques can be used to determine the 3D position of feature points throughout the vehicle's motion and then use these feature points in a manner similar to a target with a predetermined geometry.

[0069] It is also possible to determine the 3D geometry of textured vehicle feature points from a single camera using a "motion recovery mechanism" technique, but providing some additional constraints on the camera motion. Using this technique, a single camera effectively becomes a stereo camera array.

[0070] Camera calibration using "real-time" inclinometer

[0071] exist Figure 1AIn the embodiment of -B, the relative transformation between the two sensors of interest (camera 10L, 10R and inclinometer 102L, 102R) is known at the start of alignment due to one or more previous calibration processes. It is assumed that the relative pose of the sensors does not change between the time when the relative sensor orientation calibration process and the alignment measurement process are started. However, for some situations it may be advantageous not to rigidly couple the camera and inclinometer, in which case the relative sensor orientation must be determined by the alignment process. This distinction (relative sensor orientation performed before or during the alignment process) is not germane to the core novelty of the present invention.

[0072] Calculate the alignment angle

[0073] Given the above measurements, calibrations, and intermediate transformations, how can the wheel alignment angle of interest be calculated from such a measurement system?Once the key equivalences are established, the basic geometric quantities of interest are very similar to those in a conventional wheel alignment measurement system where the right-to-left transformation is measured directly.

[0074] Measurement of runout compensation

[0075] Compensation for runout of the targets mounted to the wheels is performed in the same manner as specified in conventional wheel alignment systems. The concept and calculation of runout is discussed, for example, in U.S. Patent 5,535,522. The core concept is to observe the change in orientation of a coordinate system rigidly mounted to the vehicle wheel. The change in orientation of this coordinate system allows the calculation of the optimal wheel rotation axis as the wheel rotates. The only addition to this concept in a "drive direction" aligner is a downstream step in the processing chain where all wheel axles are transformed to a common coordinate system (i.e., from the right side of the vehicle to the left side) using an optimal right side to left side transformation.

[0076] Establishing the Vehicle Coordinate System (VCS)

[0077] The concept of a vehicle coordinate system (VCS) is a common concept in wheel alignment. See, for example, U.S. Patent Application Publication 2017 / 0097229. The VCS is used as a reference frame in which alignment angles can be expressed. In the prior art, camber angles are typically defined relative to the VCS (X, Y) plane and individual toe angles are typically defined relative to the GCL (geometric center line) or thrust line of the vehicle.

[0078] Calculating GCL in the "Drive Direction" Aligner

[0079] In the prior art, the geometric center line (GCL) is calculated as the direction from the middle of the rear wheels to the middle of the front wheels. This is shown in Figures 2a-c. In a measurement system that lacks a direct measurement from the left side to the right side of the vehicle, it may be considered impossible to measure the GCL. However, a deeper study of the vehicle geometry shows that a mathematically equivalent GCL can be measured using some additional information about the vehicle geometry.

[0080] A typical GCL measurement process is depicted in Figures 2a-c when measuring directly between the left and right sides. By measuring directly from left to right to the center of the vehicle, the middle of the rear axle and the middle of the front axle can be directly calculated. The middle of the rear axle is calculated by averaging the 3D positions of the left rear wheel 212 and the right rear wheel 213 using conventional vision techniques. The middle of the front axle is calculated by averaging the 3D positions of the left front wheel 210 and the right front wheel 211. The vector from the middle rear axle to the middle front axle is shown in 216, which is the GCL.

[0081] In the driving direction aligner described herein, although the left-to-right transformation is not measured directly, a mathematically equivalent GCL direction can be measured. The vector from the center of the left rear wheel 212 to the left front wheel 210 is represented by 214. The vector from the center of the right rear wheel 213 to the right front wheel 211 is represented by 215. When the back-to-front wheel vectors 214 and 215 are averaged, this vector is mathematically equivalent to the previously described GCL vector 216.

[0082] The thrust direction 217 is calculated based on the toe angle relative to the GCL 216. The toe angle is calculated relative to the thrust direction 217.

[0083] Calibrating the vehicle's running surface relative to gravity

[0084] In order to measure camber relative to gravity in a manner that is independent of the tilt of the running surface, the tilt of the running surface (e.g., alignment lift) relative to gravity must be measured. Once this calibration is complete, the orientation of the plane of the alignment lift can be characterized in the inclinometer coordinate system and then the normal to the alignment lift transformed from it (using other calibrations and real-time measurements) to other coordinate systems.

[0085] There are a number of ways in which such an orientation of a lifting device can be performed relative to gravity. Figure 3A method is depicted in FIG. The essence of the method is to view targets 300a-d of known geometry from a static camera / inclinometer assembly 310, where the calibration between the inclinometer and the camera is known from the calibration described previously. The targets 300a-d are positioned on an alignment lift 320 at the locations of the vehicle tires. The positions of the targets 300a-d in the camera coordinate system are measured at all four wheel locations. A best fit plane 330 is calculated from these four points, and a normal to the best fit plane 330 is calculated in the camera coordinate system. Using the known calibration between the camera and the inclinometer, the orientation of the best fit plane normal in the inclinometer coordinate system can be determined. The tilt angle between the alignment lift best fit normal direction and the gravity direction can then be determined using standard trigonometric functions and stored for later use.

[0086] Define the basis vectors of VCS

[0087] Three mutually orthogonal 3D Cartesian basis vectors defining the orientation of the VCS are defined from previously defined geometric quantities. The Y axis of the VCS, corresponding to the longitudinal axis of the vehicle, is defined as the GCL. The Z axis of the VCS corresponds to the vertical dimension of the vehicle and is roughly aligned with the direction of gravity. The transformation from the measured gravity vector to the orientation of the alignment lift normal in the inclinometer coordinate system is determined using the previously performed calibration of the alignment lift relative to gravity. The alignment lift normal is transformed from the inclinometer coordinate system to the left camera coordinate system - this transformation vector constitutes the Z axis of the VCS. The alignment lift normal is further orthogonalized to remove the component parallel to the drive direction of the measured vehicle. The VCS X axis is then defined as the cross product of the VCS Y axis and the VCS Z axis.

[0088] Calculate the basic alignment angle

[0089] Once the VCS has been determined and all wheel axles have been measured and transformed into the VCS, the alignment angles can be determined in a known manner. The wheel axles are projected onto various 2D planes of the vehicle coordinate system. The camber angle is defined in terms of the pitch angle of the wheel axles relative to the VCS (X, Y) plane. The tilt angle of the alignment lift described previously relative to gravity must also be incorporated and subtracted from the calculated camber angle. The rear toe angle is calculated relative to the geometric centerline 216, as described above. The front wheel toe angle is defined relative to the vehicle thrust line 217, as described above.

[0090] Aligner Architecture

[0091] Figure 1B is a schematic top view of certain elements of a conventional computer-aided 3D motor vehicle wheel alignment system ("aligner"). As will be discussed in detail below, Figure 1BThe aligner of is an example of an aligner that can be used to implement the disclosed technology. In particular, Figure 1B The aligner comprises a left camera compartment 2 and a right camera compartment 4 for aligning the wheels of a motor vehicle. The terms "left" and "right" are used for convenience and are not intended to require that a particular element be placed in a particular position or relationship relative to another element.

[0092] Figure 1B Arrow 30 schematically shows a motor vehicle for alignment. The vehicle includes left and right front wheels 22L, 22R and left and right rear wheels 24L, 24R. Alignment targets 80a, 80b, 80c, 80d are fixed to each of the wheels 22L, 24L, 22R, 24R, respectively. Each alignment target generally includes a plate 82 with target information printed thereon and a clamping mechanism 88 for fixing the target to the wheel. The left camera compartment 2 includes a left alignment camera 10L. The left alignment camera 10L faces the vehicle and observes the left side targets 80a, 80b along the axis 42. The right camera compartment 4 includes a right camera 10R facing the vehicle and observing the right side targets 80c, 80d along the axis 44. The left camera compartment 2 also includes a calibration camera 20 mounted perpendicular to the camera 10L. The calibration camera 20 views the calibration target 16 attached to the right camera compartment 4 to determine the position of the alignment cameras 10L, 10R relative to each other. However, as described above in the "Drive Direction Operation Principles" section, when calculating and using the drive direction, the cameras and targets do not need to be calibrated as long as the second common direction is determined (eg, using inclinometers 102L, 102R attached to cameras 10L, 10R).

[0093] The disclosed aligner also includes a data processor (not shown), such as a conventional personal computer (PC), having software with instructions that cause the data processor to electronically perform the calculations described herein.

[0094] Fast measurement process

[0095] A technique that can be used to implement the aligner of the present disclosure and the functionality described herein is a very fast measurement process. This known process is described in U.S. Patent 10,072,926. The core metric of interest is the pose (position and orientation) of a target rigidly mounted to the wheel of the vehicle. Therefore, performing a fast measurement process is equivalent to performing measurements of the target pose very quickly. In an imaging aligner, quickly calculating the pose involves performing optimized image processing and applying an optimized iterative algorithm to estimate the position and orientation of the reference target. The high-speed measurement process provides for many updates and checks to be performed during the wheel turning process, which takes only a few seconds. In order to obtain a measurement from, for example, Figure 1BEach of the cameras 10L and 10R measures the attitude of the target mounted to the wheel, and must have a calibration camera. Calibration cameras are those cameras that have had their internal geometry (focal length, camera center point, lens distortion) subjected to a camera calibration process.

[0096] Examples of well-known camera calibration procedures are the Heikkila method; the Zhang method; the Faugeras-Luong method; the Hartley-Zisserman method; and the Triggs method. In order to compute the 3D pose of a target from a single 2D camera, it is also necessary to know the geometry of the observed target. Given the target geometry and suppressing the internal camera geometry, the 3D pose of the target can be computed based on a single 2D image. An example of a commercially available "fast" camera that can be used to implement the disclosed aligner is the VC Z series camera available from Vision Components GMBH of Ettlingen, Germany.

[0097] The process of computing the target pose is conventional and begins by acquiring an image of the target mounted to the wheel. This image is then processed to identify image feature points that correspond to reference fiducials in the target. These reference image points are then associated with reference fiducials in the target. Finally, an iterative pose estimation process is performed. Pose estimation answers the question: "Where do I need to place the targets and how do I orient them in order to generate the pattern of measured image feature points?".

[0098] In certain embodiments of the disclosed aligner, the entire measurement process (image acquisition, image processing, image feature point to target reference point correspondence, and pose estimation) is cyclically repeated for all targets used until the target pose information is no longer needed for the alignment process. In order to obtain measurements very quickly (e.g., greater than 20 poses per second), all four steps must be performed quickly. The data processing steps must be implemented using efficient algorithms and the data processing steps must also be implemented on a processor with an architecture optimized for the image processing and numerical linear algebra steps used in the data processing algorithm. Examples of well-known processors with architectures optimized for image processing and numerical linear algebra include DSPs (digital signal processors); GPUs (graphics processing units); FPGAs (field programmable gate arrays); and ASICs (application-specific integrated circuits). Examples of efficient data processing algorithms include Gaussian filtering; gradient descent optimization; Sobel edge detection; Canny edge detection; SURF feature point detection; and optical flow point tracking. An example of a commercially available processor having an architecture optimized for application of the present disclosure is the TDA2x ADAS DSP / ARM-type system-on-chip processor, available from Texas Instruments of Dallas, Texas.

[0099] In certain embodiments, the data collection process and advanced analysis described below are performed using a "smart camera" that is fast and performs all data processing for the aligner, including servicing the user interface, preprocessing image data, and calculating alignment parameters. More specifically, a smart camera (e.g., Figure 1B The cameras 10L, 10R shown in the figure acquire images and preprocess the image data. Preprocessing generally includes processing the image data to prepare it for use in calculating alignment parameters (such as alignment angles). Examples of well-known preprocessing operations include background subtraction, gradient calculations, position data derivation, and data compression for reducing bandwidth. For example, the preprocessing of image data is described in detail in U.S. Patent 7,069,660. The preprocessed image data from the two cameras is used to, for example, calculate an alignment angle, which is used for a display device. In other embodiments, only one of the two cameras performs processing and calculation of the alignment parameters.

[0100] Wheel alignment system and method using driving direction calculation and active sensors mounted on the vehicle

[0101] A typical conventional vehicle aligner uses sensors such as cameras to measure the position and angle of each wheel of the vehicle. One embodiment of such a wheel aligner uses a passive head with an optical target on each of the wheels of one axle (e.g., on the front wheels) and an active sensing head including a camera on each of the wheels of the other axle (e.g., on the rear wheels) to measure the alignment angles. The active head also includes sensors to spatially orient the sensors on each wheel. Figure 4A Such an aligner is shown in FIG. Targets 421, 423 and active heads 425, 427 mounted in association with vehicle wheels 422, 424, 426, 428 are used to independently measure the alignment angles of the left and right wheels. A second set of targets 451 and 452 not attached to vehicle 420 are used to spatially orient the left sensor to the right sensor. Passive target 451 is fixed to 2D camera 460. Target 451 is visible to camera 429 of active sensor head 425. Passive sensor target 452 is visible to 2D camera 460 attached to target 451 and camera 431 of active sensor head 427. In this embodiment, spatial orientation of the left sensor and target to the right sensor and target requires a complex calibration target / sensor assembly 451 / 460 and a clear view of the passive target 452. The area required for a clear view of target 452 may be blocked by the vehicle being measured, by equipment, or by personnel.

[0102] The disclosed vehicle alignment technique using drive direction calculation improves upon current and previous alignment systems by eliminating the need for direct measurements to spatially correlate left wheel measurements with right wheel measurements. Instead, the left-to-right spatial relationship can be known by combining two direction vectors that are common between the left and right sides, as discussed above in the "Drive Direction Operation Principles" section. Some direction vectors that are common between the left and right sides include, but are not limited to, the direction of vehicle motion (i.e., drive direction), the direction of gravity, the earth's magnetic field, the direction toward a common radio signal, the direction toward a common light source, the direction toward a common passive sensor, or the direction toward a floor or lift surface.

[0103] In certain disclosed embodiments, a wheel alignment system includes a pair of passive heads and a pair of active sensing heads. The passive heads are adapted to be mounted in association with a first pair of wheels in a vehicle to be measured by operation of the wheel alignment system. The active sensing heads are adapted to be mounted in association with a second pair of wheels of the vehicle. Each of the passive heads includes a target, such as can be observed by an image sensor. Each active sensing head includes an image sensor for generating image data, which is expected to include an image of the passive target when the various heads are mounted on or in association with the corresponding wheels of the vehicle. The newly disclosed system does not require additional sensor modules to determine the spatial relationship of the active sensing heads. The image sensor can be used alone or in combination with other sensors (e.g., a gravity sensor) to determine the direction of motion of the vehicle on each side of the vehicle. The relationship between the direction of the vehicle and the direction of gravity can be known through the well-known calibration of the image sensor to the gravity sensor. If the vehicle is assumed to be a rigid body, it can be assumed that the direction of movement of the wheel on the left side of the vehicle is the same as the direction of movement of the wheel on the right side of the vehicle. It can also be reasonably assumed that the direction of gravity is the same for the left side of the vehicle and the right side of the vehicle. Using only these assumptions and in the absence of direct measurements between the left and right sides of the vehicle, the left wheel measurements can be correlated with the right wheel measurements. The system also includes a processor. The processor processes image data related to the observed target and relationship data from the sensor module. The data processing enables calculation of at least one measurement of the vehicle.

[0104] According to another aspect of the present disclosure, a sensing head for a wheel alignment system includes a housing or mounting portion located on a wheel of a vehicle to be measured by operation of the wheel alignment system and an image sensor mounted to the housing. The image sensor generates image data. In a measurement operation, the image data typically includes an image of a target associated with another wheel of the vehicle. The sensing head also includes at least one tilt sensor mounted to the housing for sensing the tilt angle of the active sensing head when the active sensing head is mounted on a wheel of the vehicle. The processor responds to the image data, the sensed tilt angle, and the relationship to another head mounted on the vehicle. A communication interface coupled to the processor allows the wheel alignment measurement data to be transmitted from the active sensing head to a user device of the wheel alignment system.

[0105] A method of performing at least one measurement of a vehicle according to the principles taught herein includes capturing an image of a target associated with a first wheel of the vehicle using an image sensor in a first head mounted in association with a second wheel of the vehicle to produce first image data. Capturing an image of a target associated with a third wheel of the vehicle using an image sensor in a second head mounted in association with a fourth wheel of the vehicle to produce second image data. The method also requires deriving a relationship of the first head and the second head relative to at least one reference. Processing the first image data and the second image data and the reference relationship measurement to calculate at least one measurement of the vehicle.

[0106] Figure 4B The target and active sensing head according to some embodiments of the disclosed aligner are schematically shown relative to the wheels of a vehicle 420 under test, for example to measure one or more wheel alignment parameters. For ease of illustration, elements of the vehicle 420 are omitted except for the wheels 422, 424, 426, 428.

[0107] The wheel alignment system includes a pair of passive heads 421 and 423 mounted on respective wheels 422 and 424 of the vehicle, which in this example are front steering wheels. The system also includes a pair of fixed passive targets 433 and 434, which are placed within the detection range of active sensing heads 425 and 427. Active sensing heads 425 and 427 are adapted to be mounted in association with other respective wheels 426 and 428 of the wheels (in this case, rear wheels). Each active sensing head includes an image sensor 429 or 431 for generating image data, which is expected to include an image of a passive target located on one side of the vehicle 420, as shown. In this example, the image sensors 429 and 431 in the active sensing heads 425 and 427 are two-dimensional (2D) imaging devices, such as cameras.

[0108] Targets 421, 423, 433, and 434 are passive in that they do not include any sensing elements. Each of passive targets 421, 423, 433, and 434 is viewed by one of the image sensors 429 or 431 in active heads 425 and 427. Targets 421, 423, 433, or 434 used for image sensing by a sensor on the other head can be active or passive. An active target, such as a light emitting diode (LED), is a source that is driven by electricity to emit energy (e.g., IR or visible light) that can be detected by a sensor. A passive target is an element that is not driven by power and does not emit energy for detection by a sensor. Assuming an image sensor in head 425 or 427, a passive target would be an object that reflects (or does not reflect) light or other energy in a manner that can be detected by the corresponding image sensor. In this example, although the target may include one or more light emitting elements, the target includes light and dark areas that may be detected when illuminated by other sources and imaged by cameras or the like in active sensing heads 425 and 427 .

[0109] Active heads 425 and 427 also contain gravity sensors or the like to measure the tilt of the head, typically yaw and pitch. In this first example, head 425 includes one or more tilt sensors 437; and head 427 includes one or more tilt 430 sensors 439.

[0110] The system also includes a derived spatial relationship between reference targets 433 and 434. The derived spatial relationship enables the spatial relationship between active and passive sensing devices to be known when the sensing devices are mounted on wheels on opposite sides of the vehicle. For example, image data of reference targets 433, 434 are used to calculate multiple poses of each of these targets when the vehicle is in operation, thereby reflecting the change in the position of the axis of rotation of the rear wheels 426, 428 when the wheels are in operation, which in turn can be used to calculate the vehicle drive direction and wheel alignment parameters. See the "Drive Direction Operation Principle" section above. The spatial relationship between the left sensing device and the right sensing device is derived by knowing two or more coordinate directions that are common between the left and right sides of the vehicle. Figure 4B The embodiment shown in includes using the vehicle direction and gravity as a common coordinate direction. According to the "Driving Direction Operation Principle" section above, the driving direction is used to calculate the alignment parameters of the vehicle 420.

[0111] It will be apparent to those skilled in the art that the wheel alignment system discussed herein can be implemented with a variety of different types of common coordinate directions. Some other examples of common coordinate directions include, but are not limited to, the earth's magnetic field, a direction toward a common radio signal, a direction toward a common light source, a direction toward a common passive sensor, or a direction toward a floor or lift surface.

[0112] The disclosed system also includes one or more devices for processing image and sensor data. One or more processors process image data related to observation of the target and tilt data from the active sensing head. The data processing enables calculation of at least one measurement of the vehicle. The precise location of the fixed passive targets 433 and 434 is not critical to the operation of the aligner, as long as these fixed passive targets are visible to the one or more sensing devices of the aligner and do not move during alignment.

[0113] Other configurations of targets and sensors on and around the vehicle are also suitable for this new process. For example, passive targets 421 and 423 can be mounted to rear wheels 426 and 428, while active sensors 425 and 427 are mounted on front wheels 422 and 424. There can also be configurations where active sensors 425 and 427 are mounted on the vehicle in such a way that targets can be seen on each of the vehicle's wheels. In other embodiments, active sensors 425 and 427 are attached to each wheel and each has a field of view of fixed targets 433 and 434. In a further configuration, a gravimeter is attached to active targets 421 and 423, and active sensors 425 and 427 include only cameras. There are many configurations for various locations / positionings of sensors, and more are not mentioned here.

[0114] Wheel alignment system and method using driving direction calculation and off-vehicle mounted active sensors

[0115] As described above, some conventional camera-based aligners require that the left and right cameras be rigidly mounted to a beam and their relative positions need to be known. Other types of conventional camera-based aligners use an additional camera to view a target between the measurement cameras (i.e., across the vehicle) to correlate the relative positions of the measurement cameras. All of these aligners either require expensive components, have structures that hinder use, or both. Therefore, there is a need for an aligner that does not require additional hardware to calibrate the calibration camera to calibrate the relative positions of the cameras; and / or an aligner that does not require a rigid beam between the two cameras to free up space for shops and technicians.

[0116] exist Figure 5AIn the embodiment of the disclosed aligner shown, two cameras 500, 510 with on-board gravimeters are placed in front and to the side of the vehicle to be measured. Two targets (for two-wheel alignment) or four targets 541, 542, 543, 544 are attached to each of the four vehicle wheels. The cameras 500, 510 measure the attitude of the targets 541, 542, 543, 544 and the vehicle is moved to another position (e.g., approximately 8") and the cameras re-measure the attitude of the wheel targets. Of course, more than two attitudes of the targets can be measured. The axis of rotation and drive direction of each wheel are measured based on the target attitudes, and the gravimeter readings are measured. Using the drive direction vector and the gravimeter measurements, the toe angle and camber are calculated for the vehicle under test according to the "Drive Direction Operation Principle" section above.

[0117] The camera can be positioned in many different locations to be able to perform this alignment. For example, in some embodiments, the camera is placed behind the vehicle. In other embodiments, the camera is placed to the side or even above the vehicle instead. Alternatively, the camera can be mounted off the floor on a tripod, mounted on a wall stand, or attached directly to the alignment stand or its support.

[0118] exist Figure 5B In other embodiments shown, more than two cameras may be used. For example, four cameras 550, 560, 570, 580 (one camera per wheel) may be placed between the front and rear wheels to achieve a more compact setup. For larger vehicles such as trucks, multiple cameras may be set up to view targets attached to a single wheel or a combination of more than one wheel, depending on the area and size of the truck being aligned.

[0119] Performance enhancement using driving direction calculation for visual wheel aligner

[0120] As described above, conventional camera-based aligners require that the relative positions of the cameras be known. If the relative positions of the cameras change, the aligner will incorrectly measure the alignment of the vehicle. Likewise, a vehicle that is in perfect alignment may still exhibit behavior that is considered to be poorly aligned; for example, "tire pull" will cause the vehicle to drift and require steering input to keep the car straight. In this case, the customer may believe that the alignment was not completed correctly. Therefore, there is a need for a method to diagnose and recalibrate the aligner if the relative positions of the cameras change over time. In addition, there is a need for an aligner to measure the difference between the measured thrust angle and the actual driving direction to assist in diagnosing and / or adjusting the vehicle to compensate for it.

[0121] In this embodiment, during the positioning sequence in which the vehicle is running and the camera measures one or more attitudes of the wheel targets to measure the regular alignment angle, the measurement of the driving direction is also calculated according to the "driving direction operation principle" section above. Comparing the driving directions of each of the four wheels with each other will indicate whether the calibration of the relative camera position is still good. Since the driving directions of all wheels should be the same, the amount by which these driving directions differ from each other is an indication of the relative camera position to be calibrated or other camera calibrations. In many cases, a warning may be issued to allow the technician to recalibrate the aligner. In other cases, the calibration may be changed to make all driving directions parallel and thus the relative camera position to be recalibrated. The driving direction errors between the wheels can be averaged over many vehicle measurements to ensure that a single poor measurement result does not change the accuracy of the aligner. In addition, if the alignment measurements disagree on the driving direction, a bad measurement sequence can be indicated and a message can be conveyed to the operator to re-measure. This helps to identify measurement sequence problems before the alignment is performed and the vehicle is incorrectly adjusted.

[0122] It is generally accepted that a vehicle will travel along its thrust line (also called thrust direction). For this reason, in conventional aligners, the toe angle is displayed on the user interface and is therefore aligned with the vehicle's thrust angle. Figure 6 , the toe direction 600, 602, 604, 606 is defined as the direction that a single tire will travel, and is therefore 90 degrees to the wheel axis 610, 612, 614, 616. The thrust direction 620 is the average of the two toe directions. It is generally accepted that the vehicle will travel in the thrust direction. The drive direction 630 is the measured direction in which the vehicle actually travels.

[0123] If the vehicle is not traveling along the defined and measured thrust line 620, the vehicle will be correctly adjusted to the thrust line 620, but the vehicle will not actually be driven along the road in this way and the operator will be shown that the vehicle is not correctly aligned. This can be a significant problem because the technician has aligned correctly, but in subsequent drives, the vehicle behaves as if it was incorrectly aligned. By comparing the drive direction 630 with the thrust direction 620, a measure of this problem and its significance in test drives can be obtained. If the difference is small (e.g., less than 0.1 degrees), there is no need to worry. However, if the difference is large (e.g., greater than 0.1 degrees), there are still problems with the drivability of the vehicle even after the vehicle is aligned to the industry standard. In these cases, it is advantageous for the technician to check the vehicle for the cause of the difference and, in other cases, align the front wheels to the drive direction or a combination of the thrust direction and the drive direction.

[0124] Some causes of deviation between the thrust line and the driving direction are, but are not limited to, different tire pressures, different tire sizes, tire rubber with greater running resistance on one or more wheels, the interaction between toe angle and camber affecting the contact patch of each wheel.

[0125] Drive-through aligner using drive direction calculation

[0126] According to another embodiment, a vehicle to be aligned has targets mounted on its wheels and is driven or operated between a pair of camera pods, each having a plurality of cameras positioned to view the targets on one side of the vehicle. The cameras each collect image data of the targets on a respective side of the vehicle, and the system combines the images from the plurality of cameras to determine a series of poses for each target, which are then used to calculate alignment parameters for the vehicle, such as toe and camber. The system also calculates the driving direction of the vehicle as the vehicle moves and uses the driving direction calculation as a reference for the alignment parameters. The system displays the calculated alignment parameters to a user on a display or other display in a conventional manner (e.g., via a conventional user interface).

[0127] The wheel alignment system according to the present disclosure has a pair of elongated beam-shaped camera cabins, each having a plurality of cameras and calibration targets mounted along the horizontal axis of the cabins and facing the same direction. The two cabins are mounted with the cameras and calibration targets facing each other on opposite sides of the vehicle to be aligned. For example, on opposite sides of a calibration frame. The targets to be viewed by the cameras of the cabins are mounted on vehicle wheels, and the vehicle is driven or operated between the two cabins and passes all cabin cameras without stopping.

[0128] Each camera in one of the cabins is used to view all wheel targets disposed on a respective side of the vehicle and to capture image data of the targets as the wheels and targets rotate continuously through a plurality of rotational degrees without pausing, i.e., as the vehicle moves through the cabin. The image data is used to calculate poses of a minimum number of the viewed targets; for example, at least one pose is calculated for every five degrees of rotation as the wheels and targets rotate continuously through a plurality of rotational degrees without pausing. After the vehicle moves through the cabin, the system processes the image data of each target captured from the plurality of cameras in each respective cabin (i.e., the system combines the image data from the plurality of cameras in the cabin) to generate, for each of the targets, a target pose covering approximately 360 degrees of rotation of the target. In one embodiment, the three cameras in the cabin capture poses covering 330 degrees of rotation of each of the targets viewed by the three cameras. In certain embodiments, each of the cabins includes a data processor for performing the following steps: preprocessing the image data, and calculating alignment parameters of the vehicle based on the preprocessed image data. The calculated alignment parameters are displayed to a user on a display or other display device in a conventional manner (e.g., via a conventional user interface).

[0129] The key technology to enable the aligner disclosed herein and the functionality described herein is the very fast measurement process explained in detail in the "Drive Direction Operation Principle" section.

[0130] Now refer to Fig. 7A -D describes an exemplary embodiment of the disclosed aligner. Fig. 7A - As shown in C, the aligner 700 includes a pair of camera cabins 710, each cabin having three cameras 711 and two calibration targets 712. Each of the cabins 710 also includes a data processor 713 for performing steps of preprocessing image data and calculating alignment parameters of the vehicle based on the preprocessed image data. The cabins 710 are arranged on opposite sides of a vehicle 720 to be aligned, such as a tractor in a tractor-trailer combination, with their cameras 711 and calibration targets 712 facing each other. In this embodiment, the vehicle 720 is located on a rack 730 between the cabins 710. Figure 7C One of the compartments 710 is shown, which in this embodiment is mounted on a base 714 that is attached to a floor or other suitable surface.

[0131] Reference again Fig. 7A -B, mounting a target 740 to be viewed by the camera 711 of the cabin 710 on a wheel of the vehicle 720 . Fig.7D An embodiment of a target 740 is shown in which each target 740 has a tubular support. The tubular support has a magnet at its distal end for attachment to a lug nut or lug bolt of a vehicle 720. The tubular support varies in length between targets to accommodate the shape and / or offset of the wheel to which a particular target 740 is to be attached. For example, the front wheels of a truck typically require a target with a short support, while the rear wheels of a truck may require a relatively longer support. Each target 740 has a conventional target head with a circular reference point that is mounted to its tubular support so that it is visible to all cameras 711 of the corresponding cabin 710 when it is mounted to a wheel of a vehicle 720.

[0132] Before performing alignment, calibration targets 712 and cameras 711 are used to position the cabins 710 in a conventional manner. In an exemplary embodiment, the attitude measurement of the calibration target 712 of one of the cabins 710 is calculated based on the image data from the camera 711 of the other cabin. The geometry of all cameras 711 and calibration targets 712 of the two cabins 710 is known, and the relative positions of all calibration targets 712 and cameras 711 are calibrated either based on manufacturing tolerances or by using the well-known process of camera-to-camera relative position (RCCP) and camera-to-target relative position. This information is used to calculate the measurement of the relative attitude of the two beams relative to each other and display the measurement on the display. The user interface of the system guides the technician to place the cabin 710 in the optimal position by displaying the attitude measurement and the acceptable range. The distance, parallelism and offset between the cabins are some of the controlled setting standards.

[0133] During alignment, the vehicle 720 drives or operates between the two cabins 710 in the direction of arrow A and passes all cabin cameras 711 without pausing, thereby capturing image data of the target 740. The image data is used to calculate a predetermined minimum number of poses of the viewed target 40; for example, at least one pose is calculated for every five degrees of rotation of the wheels of the vehicle 720.

[0134] After the vehicle 720 is driven or operated through the cabin 710, the processor 713 processes the captured image data of each target 740 captured by the multiple cameras 711 of each corresponding cabin 710 (i.e., the system combines the image data from the multiple cameras 711 of the cabin 710) to generate a series of poses for each target 740 covering a 360-degree rotation of the approach target. The data processor 713 performs the following steps: pre-processing the image data and calculating the alignment parameters of the vehicle, such as toe angle and camber, based on the pre-processed image data. Because multiple cameras are used to determine the target pose, the disclosed system has significantly improved accuracy in measuring alignment parameters compared to existing systems. For example, while conventional systems are accurate to within 0.05 degrees, the disclosed system is accurate to within 0.01 degrees or better.

[0135] To achieve these high accuracies, large rotational angles of the vehicle wheels are required; for example, more than 45 degrees of rotation. However, for any given camera, the rotational angle that the camera can see is limited by the camera's field of view (FOV). Cameras with large FOVs require lenses with high distortion and low pixel density, thus losing accuracy; however, if standard lens cameras with FOVs in the range of 30 degrees are used, the rotational angle that each camera can see is limited to 20 degrees of rotation (depending on the diameter of the vehicle wheel).

[0136] To achieve the desired accuracy in the disclosed system, a running angle of approximately 90 degrees is acceptable, but larger angles are desired. By arranging the separate cameras 711 of the disclosed system along the direction in which the measured vehicle 720 is running or driving, multiple cameras 711 can each see a piece of the vehicle wheel running, thereby covering 90 degrees or 180 degrees or even 330 degrees when these pieces are combined, depending on how far apart the cameras 711 are positioned and how many cameras 711 are used in each cabin 710. Of course, the RCCP must be known so that all cameras 711 measure the running of the wheels in a common coordinate system. It is advantageous to have more than two cameras so that there can be more coverage. Likewise, two independent measurements allow the system to check and warn or compensate for errors.

[0137] In conventional camera-based alignment systems, where the vehicle to be aligned is virtually static, a vehicle coordinate system (VCS) is determined based on the two front wheel positions and the two rear wheel positions. A reference plane (i.e., the plane on which the vehicle lies) is determined, and camber is measured relative to it. The vehicle's geometric center line (GCL) is measured based on the center of the front wheels and the center point between the rear wheels, and the toe angle is referenced to this GCL as the zero line. See, e.g., U.S. Patent 5,724,743.

[0138] In contrast, in the disclosed system, when a vehicle 720, such as a truck, moves past camera 711, the system sees only one axle at a time. To define each axle and determine its coordinate system, the wheel centers of the axle are determined in a conventional manner when the axle enters the range of camera 711 and then again when it leaves the camera range. A vehicle coordinate system (VCS) is constructed as in a conventional alignment system; however, the VCS is constructed separately for each axle. Note that this VCS is a virtual coordinate system, since there are not two axles in its determination: only one axle is measured in two locations. A virtual VCS is created by treating the wheel centers of the axle as a set of rear wheels when entering and treating the wheel centers of the axle as a set of front wheels when leaving. The virtual VCS is then constructed as in a conventional alignment system.

[0139] The virtual VCS is used to calculate the toe and camber of each axle. In addition, since the vehicle 720 is a rigid body and each axle is driven in the same direction, the system can compare the toe of each axle with the toe of the other axles by equating the drive direction of each axle. Therefore, the relative toe of the two axles (i.e., the axle offset) can be measured without the camera 711 seeing both axes at the same time.

[0140] Further with respect to the driving direction of the vehicle, the direction in which the center of each wheel of the vehicle 720 moves is measured during the time that the vehicle 720 moves past the camera 711, which can be referred to as a "positioning sequence." As explained in detail above in this document, because the vehicle is a rigid body, the centers of the wheels are connected to each other (although they rotate independently). Therefore, the driving direction; for example, the direction of rotation of the two wheels of the vehicle is the same. The location of the center of the wheel can be plotted at the beginning and end of the positioning sequence, or at many discrete points through the positioning of the wheel (also known as "jump"). This information can be used to obtain the overall driving direction from the beginning to the end of the positioning sequence, or to obtain smaller driving directions as the vehicle moves through the positioning sequence. The system can compare the entire driving direction or smaller discrete driving directions.

[0141] In this embodiment, the driving direction calculation has many uses. After the two sides of the aligner (the two compartments 710) are calibrated to each other, the driving direction measurements from each side can be determined and compared to confirm that the aligner is still in calibration. If the driving directions are different, the aligner is not calibrated. In addition, in an embodiment where the system has two cameras, one on each side of the vehicle, but these cameras are not rigidly attached to each other and are not pre-calibrated, the driving direction of each side of the vehicle is measured. Since the driving directions should be the same, the position of each camera can be calibrated. This is a significant cost and complexity savings.

[0142] Another embodiment eliminates the need for RCTP calibration, which is used to find where each beam is located relative to the other beam. Each beam measures the driving direction of the vehicle, and when the driving direction is coupled with one or more directions or vectors measured by the two beams, the position of one beam relative to the other can be found. Once the position of the two beams relative to each other is known, the alignment angle can be calculated in a similar manner to the previous embodiment.

[0143] As described above, the other common direction or directions or vectors may come from a variety of sources. The direction of gravity, the plane between the center camera of the beam and the other beam targets, etc. may be used. If the embodiment does not use beam targets, then these beam targets may be omitted from the system. If the embodiment uses the direction of gravity, then each camera cabin 710 includes a conventional inclinometer, etc.

[0144] In another embodiment, the vehicle is driven or driven between the camera beams and stops while the wheel targets are still visible to the camera. The aligner continuously checks whether the wheels have stopped translating and / or rotating to determine when the vehicle has stopped. Once the vehicle stops, the user interface automatically displays the alignment angle updated in real time, allowing the user to adjust the vehicle alignment.

[0145] In another embodiment, the vehicle is driven or driven between the camera beams and the aligner detects if the alignment of the vehicle needs to be adjusted and tells the user to stop inside the camera beams. The aligner continuously checks whether the wheels have stopped translating and / or rotating to determine when the vehicle has stopped. Once the vehicle stops, the user interface displays the alignment angle updated in real time, allowing the user to adjust the vehicle alignment.

[0146] In another embodiment, the pose of the wheel is determined without a target attached to the wheel. There are several conventional methods to accomplish this, such as by using stereo cameras (ie, two cameras with overlapping fields of view), structured light, lasers, and the like.

[0147] Use driving direction calculation to aim at ADAS targets

[0148] Advanced Driver Assistance Systems (ADAS) in vehicles assist the driver and provide an additional level of vehicle safety. This includes systems such as adaptive cruise control, lane departure warning, collision avoidance, and rear collision warning. These systems use a combination of sensors such as cameras, sonar, and radar that may require calibration to operate correctly. Calibration is performed by placing elements in a known relationship to the vehicle. Calibration elements are known sources or reflectors, such as lasers, mirrors, targets, or heat sources. Fig. 8A The bracket 871 is shown to hold one or more calibration elements 861 in a known position relative to the base of the fixture 860. The bracket 860 is oriented to the vehicle 800 relative to a desired line of the vehicle 800, such as Figure 8B Thrust line 810 is shown. US Patents 7,121,011 and 6,823,601 describe types of ADAS, adaptive cruise control, and the structure, use, and operation of the bracket.

[0149] Conventional systems exist that perform vehicle / carriage orientation using simple manual measurements by drawing a grid on the floor or using laser lines to line up the elements. There are also systems that use a computerized wheel alignment system where the elements are placed on a carriage to orient and position the calibration elements relative to the vehicle. There are many advantages to using a wheel alignment system, primarily that the process is guided by the aligner user interface, making it faster and requiring less skill from the operator. In addition, the wheel alignment system can measure and record whether the procedure was performed correctly. These systems all rely on direct measurement of the desired vehicle line, either by locating multiple points on the vehicle body or by directly measuring the combined orientation of the rear wheels. It would be desirable to create a low cost, simpler system that does not require a full wheel alignment system while still providing the user with guided operation to guide the position and orientation of the carriage and to measure and record the position of the calibration elements.

[0150] According to this embodiment, as described in the "Driving Direction Operation Principle" section above, one or more image sensors (e.g., cameras) are placed on a bracket and one or more targets are placed on the vehicle to measure the driving direction of the vehicle. Alternatively, the driving direction of the vehicle can be measured using one or more cameras on the vehicle and one or more targets on the bracket (including targets for calibrating ADAS systems). It is assumed that the driving direction is parallel to the vehicle thrust line and can be used as a line to orient the bracket to the vehicle. In some embodiments, it is advantageous to use a single camera to use the driving direction for direct measurement, and then align the ADAS target with the driving direction.

[0151] Additionally, as described above, the driving direction can be used to correlate the orientation of multiple cameras for measuring vehicle alignment angles such as toe and camber. As in conventional imaging wheel alignment systems, these angles can be used to determine the thrust or centerline of the vehicle to orient the carriage. The wheel alignment angles can also be used to verify that the wheel alignment angles meet manufacturing requirements before initiating ADAS calibration.

[0152] Example of two cameras mounted to a fixture

[0153] Reference again Fig. 8A In one embodiment, two cameras 879, 881 are rigidly mounted or mounted using pivot points 883, 885 that can be locked to either side of a bracket 871 carrying an ADAS calibration element 861 (e.g., a target). The relationship between the cameras 879, 881 does not have to be controlled or calibrated. For at least one camera, the relationship between the line created by the intersection of the plane of the calibration element 861 and the gravity plane must be calibrated or controlled.

[0154] The cameras 879 , 881 are connected to a display device such as a computer or tablet 898 for presenting a user interface via a wired or wireless communication interface.

[0155] Traditional targets, such as Figure 1B Targets 80a-d are shown attached to all vehicle wheels using a mechanical device 82 that attaches to the vehicle wheel or tire. The mechanical device 82 establishes a known relationship between the target position and the center of the vehicle wheel, or infers the center of the wheel from the movement of the target as the wheel rotates. For clarity, the target position is shown in FIG. Figure 8B These wheel targets are omitted.

[0156] The vehicle is driven forward, backward, or forward and backward. The translation of the wheel targets is measured and a driving direction is established. The driving direction and an additional known common direction that is not parallel to the driving direction are used to relate the orientation of the left and right cameras. In this case, the additional direction used is the direction of gravity, called the gravity vector, as described above in the "Drive Direction Operation Principle" section. Therefore, each of the cameras 879, 881 includes a gravity sensor, such as a conventional inclinometer 879a, 881a.

[0157] Since the orientation between the cameras 879, 881 and the bracket 871 is known, the bracket can be oriented (i.e., rotated) about its centerline 890 to the driving direction when guided by a supported user interface. However, since the position and orientation of the ADAS calibration element 861 needs to be set, and the left-to-right orientation of the vehicle 800 relative to the cameras 879, 881 is unknown, a point or line laser 896 (mounted on the bracket 871 perpendicular to the calibration element 861 and on the centerline 890 of the bracket) is used to align the position of the bracket and therefore align the calibration element 861 relative to the center of the front of the vehicle 800. In other words, the laser 896 is used to align the fixture 860 with a point on the centerline of the vehicle 800 (at Figure 8C 2 and described in detail below), and cameras 879, 881 are used to set the rotation of the fixture 860 about its centerline 890. Alternatively, an approximate calculation can be made by using the pre-measured distance between the camera mounts on the fixture.

[0158] This solution is somewhat compromised, but since the angle between the driving direction of the vehicle 800 and its geometric centerline should be smaller, the error will be smaller. Figure 8C , which shows the ADAS calibration element 861 at an ideal position 1 on the thrust line, which is assumed to be the same as the driving direction of the vehicle; and at position 2 using a laser 896 set parallel to the driving direction directed toward the front of the vehicle 800. It can be seen that position 2 is slightly compromised, but reduces the error caused by placement directly on the vehicle centerline position 3. Currently available ADAS elements are placed on the body centerline, so this embodiment is an improvement over the prior art.

[0159] Additionally, because the relative orientations of cameras 879, 881 are known, wheel alignment values ​​may be calculated as described above for verification against manufacturer specifications.

[0160] In an alternative embodiment, a target 894 is placed on the centerline of the vehicle in the field of view of camera 879 or 881. Based on the wheel alignment measurements, the direction of the vehicle centerline is known. The position of the centerline in the gravity plane is determined by the placed target. The user interface is then used to guide the operator to place the bracket on the vehicle centerline or vehicle thrust line based on manufacturer requirements. Alternatively, the camera can use a known structure such as a decorative feature or logo on the front of the vehicle instead of target 894 to locate a point on the vehicle centerline.

[0161] In another embodiment, a mechanical device is used to measure the width of the car from the outer surface of the rear wheels. This width can be input into a user interface, and the center of the vehicle relative to the wheel targets can be calculated and used to determine the centerline of the vehicle.

[0162] In an alternative embodiment, the distance between the cameras 879, 881 on the carriage 871 is mechanically measured or established. This will provide sufficient information to calculate the width of the vehicle and establish the position of the vehicle's thrust line.

[0163] The distance between the bracket 871 and the vehicle 800 may be set using a belt measurement, where the required distance is derived from the vehicle body and the bracket.

[0164] With a defined distance from the bracket 871 to the vehicle 800 measured from the center of the front wheels, cameras 879, 881 on the bracket 871 can be used to measure the distance between the vehicle 800 and the calibration element 861 and a user interface can guide the operator in setting up the bracket.

[0165] In other embodiments, the target 892 is placed on the surface on which the vehicle is located in the field of view of one of the cameras 879, 881, and the relative heights of the camera and the crossbar 877 are then controlled or calibrated. The user interface then instructs the operator to raise the crossbar 877 to the correct height to position the calibration element 861.

[0166] After the ADAS calibration element 861 is placed in the correct position using one of the above disclosed techniques, the diagnostic device is connected to the car's OBDII port and used to initiate the calibration process, which is executed by the car's ECU.

[0167] Single Camera Mounted to Fixture Embodiment

[0168] According to another embodiment of the disclosed technology, a single camera 891 is mounted to a fixture 860 and is located in the field of view of the camera 891, and one or more targets 893 are fixed to the vehicle at any position. The camera 891 is mounted at a known position and orientation relative to the centerline 890 of the bracket 871 by calibration or mechanical design. As in the previous embodiment, a display device such as a user interface 898 is provided. The vehicle 800 is driven forward, backward, or forward and backward. As described above, the translation of the target is measured and the driving direction is established. Since the camera 891 is in a known relationship with the bracket fixture 860, the user interface can guide the user to orient the centerline of the fixture to the driving direction. As described in the previous embodiment, the bracket 871 is set to be a certain distance from the car 800 and centered on the car 800.

[0169] In a variation of this embodiment, a target (i.e., target 894) is mounted at the center of the front bumper of vehicle 800. When the position of target 894 is measured, the centerline of vehicle 800 and the distance to the bumper are known. The user interface guides the operator to position bracket 860 in an orientation to the driving direction and in a relative position from the front bumper relative to the center and centerline of vehicle 800.

[0170] In another variation of this embodiment, an additional target is placed to the side of the vehicle just above the front wheel center in the field of view of camera 891; for example, target 893 can be a floor height target. Since camera 891 can measure the distance to this target 893, the distance between bracket 860 and vehicle 800 relative to the wheel center can be controlled and the placement can be guided by the user interface.

[0171] In another variation of this embodiment, a camera (i.e., camera 895) is placed on the crossbar 877 in a known or calibrated position relative to the centerline 890 of the crossbar and fixture 860. A target 892 is placed on the same surface as the vehicle in the field of view of the camera 895. As described above, the height of the target 892 from the floor is determined, and the user interface guides the user in setting the height of the crossbar 877.

[0172] Those skilled in the art will appreciate that there are many possible positions for the target and the camera, for example, depending on the location that needs to be measured. Therefore, the above positions are exemplary and not exclusive.

[0173] Camera viewing fixture embodiment

[0174] In another embodiment, the camera 897 is placed on the vehicle so that it can see the bracket 871, and the target is placed on the bracket 871. The target can be placed on the bracket 871, on the crossbar 877, embedded in the calibration element 861, or in the case of forward camera calibration, the ADAS calibration element 861 is used as a target and can be viewed by the camera 897. Operation and relative position are determined as described in the previous embodiment, where the positions of the camera and the target are interchanged. If the camera 897 is mounted to the front bumper of the car 800, the distance between the fixture and the front bumper can be controlled. If the target is placed on the floor in the field of view of the camera and the vehicle is located on the floor or at a known height from the floor and the target on the bracket is located on the crossbar, the height of the crossbar and the calibration element can be controlled.

[0175] In the above embodiment, all relative positions of the support and vehicle may be stored in the measurement system and retrieved to show that the process has been performed correctly. In the first embodiment, the wheel alignment may also be stored.

[0176] In all embodiments, the processing of the camera image and the derivation of relative distances, wheel alignments and drive lines may be performed in the camera, in a display device or on a remotely connected device.

[0177] ADAS calibration elements, targets or fixtures themselves may be used as targets and the measurement cameras used to position them measure the driving direction as previously described herein and the fixtures adjusted appropriately.

[0178] Driving assistance using driving direction calculation

[0179] The calculated driving direction can be used to guide the operator to drive the vehicle onto a conventional alignment stand or lift so that the vehicle is straight relative to the stand and the front wheels of the vehicle are centered on the turntable of the stand. This embodiment requires an aligner, two conventional floor height targets (optionally four floor height targets), and a target for the turntable of the stand (or alternatively, the turntable includes the target). The technique also requires that the turntable target be calibrated to the center of the turntable before using it.

[0180] Reference now Fig. 9, before driving the vehicle 920 onto the stand 900, the operator attaches the chassis height target 910 at the apex of each front fender wheel of the vehicle 920. The aligner ensures that the turn plate targets 930 and the chassis height targets 910 are visible to its camera 930 and proceeds to the setup screen on its user interface. The width between the front chassis height targets 910 is measured relative to the width between the turn plate targets 930, and the operator is instructed to adjust the distance between the turn plates 950 as needed to match the wheel track of the vehicle 920, which is approximately determined from the distance between the chassis height targets 910. In addition, the wheel track can be read from the vehicle specification database, if present. Otherwise, the aligner can measure the distance between the turn plates 950 at the end of the drive-on process and save it in the specification database.

[0181] The aligner advances to the guide screen of its user interface. The operator then begins to slowly drive the vehicle 920 onto the stand 900. As the vehicle 920 begins to move, the aligner processes the motion of the chassis height target 910 into an instantaneous driving direction and calculates a trajectory of the position of the vehicle 920 relative to the turn plate 950 in accordance with the "driving direction fundamentals" above. The aligner uses this trajectory to guide the operator so that the vehicle 920 is positioned straight on the stand 900 and parked with the front wheels 960 centered on the turn plate 950. Optionally, a rear chassis height target 970 is also attached and used to provide feedback on how well the vehicle 920 is coordinated with the aligner and stand 900 as the vehicle 920 is driven onto the stand 900. The disclosed embodiments of the method and system for performing vehicle alignment using driving direction calculations can be readily implemented in whole or in part in software using, for example, an object or object-oriented software development environment that provides portable source code that can be used on a variety of computer platforms. Alternatively, embodiments of the disclosed methods and systems may be implemented in part or in whole in hardware using, for example, standard logic circuits or VLSI designs. Other hardware or software may be used to implement embodiments, depending on the speed and / or efficiency requirements of the system, the specific functions and / or the specific software or hardware systems, microprocessors or microcomputer systems used. Those skilled in the art may implement embodiments of the disclosed methods and systems in hardware and / or software using any known or later developed systems or structures, devices and / or software, based on the functional descriptions and general descriptions provided herein and using basic knowledge of computers, marine barriers and gates, and / or existing automatic door technologies.

[0182] Furthermore, embodiments of the disclosed method and system for performing vehicle alignment using driving direction calculations may be implemented in software executed on a programmed general purpose computer, a special purpose computer, a microprocessor, etc. Moreover, the disclosed method may be implemented as a program embedded on a personal computer, such as or CGI scripts, implemented as resources residing on a server or graphics workstation, implemented as routines embedded in a dedicated processing system, etc.

[0183] Although the present disclosure has been described in conjunction with a number of embodiments, it is apparent that many substitutions, modifications and variations will be or will be apparent to those skilled in the art. Therefore, the applicant intends to include all such substitutions, modifications, equivalents and variations within the spirit and scope of the present disclosure.

Claims

1. A vehicle alignment system, comprising: a first image sensor for viewing a first target disposed on a first wheel of the vehicle on a first side of the vehicle and for capturing image data of the first target while the vehicle is in motion and the first wheel and target are rotating; a second image sensor for viewing a second target disposed on a second wheel of the vehicle on a second side of the vehicle opposite the first side of the vehicle and for capturing image data of the second target while the vehicle is in motion and the second wheel and target are rotating; a first gravity sensor attached to the first image sensor, the first gravity sensor for measuring a sensed orientation relative to gravity on the first side of the vehicle; a second gravity sensor attached to the second image sensor, the second gravity sensor for measuring a sensed orientation relative to gravity on the second side of the vehicle; as well as A data processor configured to perform the following steps: using the image data to calculate a plurality of poses of each of the first target and the second target as the first wheel and the second wheel and the targets rotate; using the calculated pose of the first target to calculate a driving direction of the first side of the vehicle, wherein the driving direction of the first side is the direction in which the vehicle moves in the coordinate system of the first image sensor, and using the calculated pose of the second target to calculate a driving direction of the second side of the vehicle, wherein the driving direction of the second side is the direction in which the vehicle moves in the coordinate system of the second image sensor; calculating the direction of gravity on the first side of the vehicle using the measured orientation of the first gravity sensor relative to gravity, and calculating the direction of gravity on the second side of the vehicle using the measured orientation of the second gravity sensor relative to gravity; calculating a vehicle driving direction by transforming the driving direction and the gravity direction of the first side of the vehicle into a common coordinate system common to the driving direction and the gravity direction of the second side of the vehicle; and Wheel alignment parameters are calculated using the vehicle drive direction.

2. The system according to claim 1, wherein: the first image sensor and the first gravity sensor being mounted to a third wheel of the vehicle on the first side of the vehicle such that the first image sensor views the first target and a first reference target not mounted on the vehicle for capturing image data of the first reference target while the vehicle is in motion; wherein a second image sensor and a second gravity sensor are mounted to a fourth wheel of the vehicle on the second side of the vehicle such that the second image sensor views the second target and a second reference target not mounted on the vehicle to capture image data of the second reference target while the vehicle is in operation; and The data processor is used to calculate the posture of each of the first reference target and the second reference target, calculate the vehicle driving direction, and calculate the wheel alignment parameter using the image data of the first reference target and the second reference target.

3. The system according to claim 1, wherein: the first image sensor being for viewing the first target and for viewing a third target disposed on a third wheel of the vehicle on the first side of the vehicle, and for capturing image data of the first target and the third target while the vehicle is in operation; wherein a second image sensor is used to view the second target and to view a fourth target disposed on a fourth wheel of the vehicle on the second side of the vehicle, and to capture image data of the second target and the fourth target while the vehicle is in operation; and The data processor is used to perform the following steps: using the image data to calculate a plurality of poses of each of the first target, the second target, the third target, and the fourth target as the wheel and the target rotate; and The driving direction of the first side of the vehicle is calculated using the calculated poses of the first target and the third target, and the driving direction of the second side of the vehicle is calculated using the calculated poses of the second target and the fourth target.

4. The system according to claim 1, wherein: The first image sensor is mounted on the first side of the vehicle beside the vehicle side, and the second image sensor is mounted on the second side of the vehicle beside the vehicle side, such that the vehicle can be operated between the first image sensor and the second image sensor; wherein the first image sensor and the second image sensor each include a plurality of image sensors, the plurality of first image sensors being used to view the first target when the first wheel and target rotate when the vehicle is in operation, and the plurality of second image sensors being used to view the second target when the first wheel and target rotate when the vehicle is in operation; and The data processor is used to combine image data from the plurality of first image sensors and the plurality of second image sensors to respectively calculate a plurality of postures of each of the first target and the second target.

5. The system according to claim 3, wherein: The first image sensor and the second image sensor are mounted to a fixture, the fixture is used to carry an ADAS calibration element, the orientation between the image sensor and the ADAS calibration element is known, and the fixture is located in front of the vehicle; The data processor is used to use a user interface to guide a user to orient the ADAS calibration element to the vehicle driving direction.

6. The system of claim 5 , comprising a laser mounted on the fixture on the centerline of the ADAS calibration element and located in a field of view of the centerline of the vehicle so that the user can visually associate the ADAS calibration element with a point on the centerline of the vehicle.

7. The system according to claim 5, wherein: The calculated wheel alignment parameters include a vehicle centerline, the system further comprising a centerline target placed on the vehicle centerline in the field of view of one of the first image sensor and the second image sensor, and the data processor is used to calculate the position of the vehicle centerline in the gravity plane using the centerline target and guide a user to place the fixture on the vehicle centerline using the user interface.

8. The system according to claim 5, wherein: The data processor is used to measure the distance between the center of the vehicle's front wheel and the ADAS calibration element using the first image sensor and the second image sensor and the first target and the second target, and to guide a user to place the fixture at a predetermined distance from the center of the front wheel using the user interface.

9. The system according to claim 5, wherein: The fixture comprises a height-adjustable crossbar for carrying the ADAS calibration element and the first image sensor and the second image sensor; wherein the system further comprises a second target located in a field of view of one of the image sensors, the second target being disposed on a surface on which the vehicle is located; and The data processor is used to determine the height of the crossbar relative to the surface using a second target and one of the image sensors, and is used to guide a user to adjust the height of the crossbar to a predetermined height using the user interface.

Images