External vision-based dual-wheel differential robot chassis speed calibration method and device

Abstract

The application relates to a kind of double-wheel differential robot chassis speed calibration method and device based on external vision, comprising the following steps: system calibration;Data acquisition;Vision data processing;Motion trajectory analysis;Parameter calibration.The beneficial effects of the present application are: the circular calibration runway serves as the absolute physical reference of the vision system;The stress sensing sheet is embedded in the radial position scale corresponding to the angle of the runway, when the robot wheel presses the stress sensing sheet, a trigger signal will be generated and simultaneously marked to the time axis of the vision video stream, forming the corresponding relationship between physical trigger and vision frame, and reducing the error caused by vision detection jitter;Non-collinear feature points are selected on the circular calibration runway, the pixel coordinates of the feature points in the image are collected, the homography matrix between the image coordinate system and the physical coordinate system is solved using the direct linear transformation algorithm, and the pixel coordinates are converted into physical coordinates on the runway through the inverse matrix of the homography matrix, which improves the coordinate conversion accuracy.

Description

Technical Field

[0001] This invention belongs to the field of robotics technology, and in particular relates to a method and device for calibrating the chassis speed of a two-wheeled differential robot based on external vision. Background Technology

[0002] Two-wheeled differential robots are widely used in warehousing and logistics, service robots, and other fields due to their simple structure and convenient control. However, due to problems such as "uncertainty of ground contact friction coefficient", "unequal wheel diameters", "uncertainty of wheel track", and "discrepancy between the actual average and nominal average diameters of the two wheels", the actual motion performance of the robot deviates significantly from the theoretical model.

[0003] Currently, the calibration methods for two-wheel differential robots mainly rely on:

[0004] Encoder / odometer calibration method: Calibration is performed by comparing the robot's encoder feedback with the actual distance traveled. For example, the authorized patent: A chassis calibration method for a two-wheel differential warehousing and logistics robot CN119126816A calibrates by listening to the distance between the robot's base coordinate system and the odometer coordinate system (odom), and measuring the difference between the actual distance traveled and the distance detected by the system. This method relies on the accuracy of the robot's own sensors, and the calibration results are greatly affected by factors such as wheel slippage and ground friction.

[0005] LiDAR-assisted calibration method: The overdetermined equations are established by combining the pose information collected by the lidar with the pose calculated by the odometer to solve for the wheel diameter and wheel spacing. This method requires additional lidar equipment and depends on environmental features. The calibration effect will be affected in environments with sparse features.

[0006] The aforementioned methods all share a common limitation: they rely on the robot's own sensors (encoders, odometers) or environmentally sensitive external sensors (LiDAR) for calibration, failing to completely eliminate the influence of sensor system errors and ground friction. Therefore, this invention proposes a calibration scheme based on an external visual reference, enabling absolute measurement of the robot's motion trajectory, thereby eliminating reliance on the robot's own sensors and improving calibration accuracy and reliability. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and device for calibrating the chassis speed of a two-wheeled differential robot based on external vision.

[0008] This method for calibrating the chassis speed of a two-wheeled differential robot based on external vision includes the following steps:

[0009] Step 1, System Calibration: Establish a mapping relationship between the image coordinate system and the physical coordinate system of the circular calibration runway using a camera above the runway.

[0010] Step 2, Data Acquisition: Place the robot on the circular calibration track and control its constant speed movement; acquire motion video through the camera, and the stress sensing plate on the circular calibration track generates a trigger signal for the robot and marks it on the video stream timeline;

[0011] Step 3: Visual data processing: Convert the robot's position from pixel coordinates to physical coordinates to obtain a position sequence;

[0012] Step 4: Motion trajectory analysis: Calculate the robot's instantaneous velocity and trajectory fitting circle parameters based on the position sequence, separate the linear velocity and angular velocity components based on the fitting circle parameters, and optimize the robot's instantaneous velocity;

[0013] Step 5, Parameter Calibration: Compare the optimized robot instantaneous speed with the theoretical model to obtain the wheel diameter and wheelbase.

[0014] Preferably, in step one, the circular calibration track has angular scales and corresponding radial position scales on its surface facing the camera. Using the radial position scales, at least four non-collinear feature points with known physical coordinates are selected on the circular calibration track, and the pixel coordinates of the feature points in the image are collected. The homography matrix between the image coordinate system and the physical coordinate system is solved using a direct linear transformation algorithm.

[0015] Preferably, in step two, the robot moves at least one revolution, and the stress-sensing plate is embedded below the radial position scale. When the robot presses over the stress-sensing plate, the stress-sensing plate generates a trigger signal.

[0016] Preferably, in step three, a target is attached to the top of the robot. The pixel coordinates of the four corner points of the target in the image are extracted using a target detection algorithm. The pixel coordinates of the geometric center of the four corner points are taken as the position of the robot, and the target center coordinates of consecutive frames are filtered.

[0017] Preferably, in step four, the actual linear velocity and angular velocity of the robot are calculated by substituting the linear velocity and angular velocity calculation formulas into the position sequence; the robot's motion trajectory is fitted using the least squares method to obtain the fitting circle parameters; the average residual between the robot trajectory points and the fitting circle is calculated to verify the trajectory quality; the tangential angle of the trajectory points is represented by the fitting circle parameters, and the angular velocity of the robot is optimized using the tangential angle expression.

[0018] Preferably, in step five, a wheel diameter correction coefficient and a wheel spacing correction coefficient are introduced into the robot's theoretical motion model to establish a calibration model. The actual linear velocity and the optimized angular velocity are compared with the calibration model to obtain the wheel diameter correction coefficient and the wheel spacing correction coefficient, thereby obtaining the calibrated actual wheel diameter and wheel spacing.

[0019] As a preferred option, step six follows step five: perform repeatability tests by calibrating multiple times under the same conditions; use the calibrated parameters to control the robot's movement and verify the degree of conformity between the actual trajectory and the expected trajectory; and compare the calibration results with those obtained using traditional methods.

[0020] The device used for the aforementioned external vision-based dual-wheel differential robot chassis speed calibration method includes: a circular calibration track, a camera, and a stress-sensing plate; the camera is mounted on the circular calibration track via a mounting bracket; the surface of the circular calibration track facing the camera is provided with angular scales and corresponding radial position scales, and the stress-sensing plate is embedded below the radial position scales; the stress-sensing plate is connected to a data collection device via a sensor data acquisition line.

[0021] The beneficial effects of this invention are:

[0022] 1) The circular calibration track of the present invention serves as both an absolute physical reference for the vision system: the concentric rings and angle scales on the track surface provide known physical coordinate points for establishing the mapping from the image coordinate system to the physical coordinate system, and assisting in camera extrinsic parameter calibration and distortion correction, eliminating systematic errors in visual measurement; and a motion constraint and trajectory guidance for the robot: the circular structure matches the circular motion model of the two-wheeled differential robot. When the robot moves along the track, its theoretical trajectory is highly coupled with the geometric features of the track, which can reduce human intervention and allow the motion trajectory to naturally meet the geometric constraints required for calibration, realizing the linkage of track features, visual perception, and motion constraints.

[0023] 2) In this invention, a stress-sensing plate is embedded under the radial position scale corresponding to the angle of the runway. When the robot wheel passes over the stress-sensing plate, a trigger signal is generated and synchronously marked on the time axis of the visual video stream, forming a correspondence between physical trigger and visual frame. During the data acquisition stage, if the vision detects that the robot is located near a certain scale and the corresponding stress-sensing plate has a trigger signal, then that frame can be directly used as a high-confidence position anchor point. During the trajectory analysis stage, trajectory segments near these anchor points are preferentially selected to participate in the calibration parameter calculation, reducing the error caused by visual detection jitter.

[0024] 3) This invention selects at least four non-collinear feature points with known physical coordinates on a circular calibration track, collects the pixel coordinates of the feature points in the image, uses the direct linear transformation algorithm to solve the homography matrix between the image coordinate system and the physical coordinate system, and converts the pixel coordinates into physical coordinates on the track through the inverse matrix of the homography matrix. This model also compensates for the rotation, translation and perspective distortion of the camera, significantly improving the coordinate transformation accuracy. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the overall structure of the device;

[0026] Figure 2 This is a schematic diagram of the structure of a circular calibration track;

[0027] Figure 3 This is a flowchart of a vision-based motion parameter calibration method;

[0028] Figure 4 This is a schematic diagram of robot motion trajectory analysis.

[0029] Figure labeling: 1. Robot, 2. Circular calibration track, 3. Camera, 4. Fixture, 5. Stress sensor plate, 6. Angle scale, 7. Sensor data acquisition line, 8. Data collection device. Detailed Implementation

[0030] The present invention will be further described below with reference to embodiments. The description of the embodiments below is only for the purpose of helping to understand the present invention. It should be noted that those skilled in the art can make several modifications to the present invention without departing from the principle of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

[0031] Example 1

[0032] As one embodiment, a method for calibrating the chassis speed of a two-wheeled differential robot based on external vision is proposed, such as... Figure 1-4 As shown, it includes the following steps:

[0033] Step 1, System Calibration: Establish a mapping relationship between the image coordinate system and the physical coordinate system of the circular calibration track 2 using the camera 3 above the circular calibration track 2. Specifically, the surface of the circular calibration track 2 facing the camera 3 is provided with angle scale 6 and corresponding radial position scale. Using the radial position scale, at least four non-collinear feature points with known physical coordinates are selected on the circular calibration track 2, and the pixel coordinates of the feature points in the image are collected. The homography matrix between the image coordinate system and the physical coordinate system is solved using the direct linear transformation algorithm.

[0034] More specifically, let X be the homogeneous coordinate of a point in the physical coordinate system. world=[X,Y,1] T The homogeneous coordinates of the image pixels are x img =[u,v,1] T The transformation relationship is then: λx img =HX world ;

[0035] Where H is a 3×3 homography matrix, For physical coordinate systems, λ is the image pixel coordinate, and λ is the scale factor of the homogeneous coordinate. By collecting at least 4 sets of non-collinear known physical coordinate-pixel coordinate point pairs on the radial position scale of the runway, the model is solved using the Direct Linear Transform (DLT) algorithm. This model can simultaneously compensate for camera rotation, translation and perspective distortion, and improve coordinate transformation accuracy.

[0036] More specifically, the homography matrix transformation process is as follows: Stage 1: Select at least four non-collinear feature points with known physical coordinates on the circular calibration track 2, record their physical coordinates (X,Y) and image pixel coordinates (u,v), and establish the homogeneous equation: λ[u,v,1] T =H[X,Y,1] T The homography matrix H is solved linearly using the DLT algorithm;

[0037] Phase 2: Given the pixel coordinates (u,v) of the center of the top target of Robot 1, substitute this value into the calibrated homography matrix H to solve for its corresponding physical coordinates (X,Y).

[0038] Step 2, Data Acquisition: Place robot 1 on the circular calibration track 2 and control its constant speed movement; acquire motion video through camera 3, and the stress sensing plate 5 on the circular calibration track 2 generates a trigger signal for robot 1 and marks it on the video stream timeline;

[0039] Step 3: Visual data processing: Convert the position of robot 1 from pixel coordinates to physical coordinates to obtain the position sequence;

[0040] Step 4: Motion trajectory analysis: Calculate the instantaneous velocity and trajectory fitting circle parameters of robot 1 based on the position sequence, separate the linear velocity and angular velocity components based on the fitting circle parameters, and optimize the instantaneous velocity of robot 1.

[0041] Step 5, Parameter Calibration: Compare the optimized instantaneous speed of Robot 1 with the theoretical model to obtain the wheel diameter and wheelbase.

[0042] Example 2

[0043] As another embodiment, this second embodiment proposes, based on the first embodiment, a more specific method for calibrating the chassis speed of a two-wheeled differential robot based on external vision, such as... Figure 1-4 As shown, it includes the following steps:

[0044] Step 1, System Calibration: Establish the mapping relationship between the image coordinate system and the physical coordinate system of the circular calibration track 2 through the camera 3 above the circular calibration track 2;

[0045] Step 2, Data Collection: (e.g.) Figure 1 and Figure 2 As shown, robot 1 is placed on the circular calibration track 2 and its movement is controlled at a constant speed; motion video is captured by camera 3, and stress sensing plate 5 on the circular calibration track 2 generates a trigger signal for robot 1 and marks it on the video stream timeline; specifically, the two-wheel differential robot is placed at the starting point of the track, for example, at 0° position, and robot 1 is controlled to move on the circular calibration track 2 at a constant speed. The nominal angular velocity of the left and right wheels is set, and robot 1 moves along the circular calibration track 2. The top camera 3 continuously captures motion video of robot 1. Robot 1 moves at least one revolution to obtain a complete motion video. The stress sensing plate 5 is embedded below the radial position scale. When robot 1 passes over the stress sensing plate 5, the stress sensing plate 5 generates a trigger signal;

[0046] More specifically, a micro stress-sensing plate 5 is embedded below the radial position scale at angles such as 0° starting point, 90°, and 180°, and connected to an independent data collection device 8. When the wheel of robot 1 runs over the stress-sensing plate 5, a hardware trigger signal with millisecond precision is generated. This signal is synchronously marked on the time axis of the visual video stream, forming a correspondence between physical trigger and visual frame.

[0047] During the data acquisition phase, if the vision detects that the robot 1 is near a certain scale and the corresponding stress sensing plate 5 has a trigger signal, then the frame can be directly used as a high-confidence position anchor point.

[0048] During the trajectory analysis phase, trajectory segments near these anchor points are prioritized for calibration parameter calculation to reduce errors caused by visual detection jitter.

[0049] Step 3: Visual data processing: Convert the position of robot 1 from pixel coordinates to physical coordinates to obtain the position sequence;

[0050] Step 4: Motion trajectory analysis: Calculate the instantaneous velocity and trajectory fitting circle parameters of robot 1 based on the position sequence, separate the linear velocity and angular velocity components based on the fitting circle parameters, and optimize the instantaneous velocity of robot 1.

[0051] Step 5, Parameter Calibration: Compare the optimized instantaneous speed of Robot 1 with the theoretical model to obtain the wheel diameter and wheelbase.

[0052] It should be noted that the parts in this embodiment that are the same as or similar to those in Embodiment 1 can be referred to each other, and will not be repeated in this application.

[0053] Example 3

[0054] As another embodiment, this third embodiment proposes, based on the second embodiment, a more specific method for calibrating the chassis speed of a two-wheeled differential robot based on external vision, such as... Figure 1-4 As shown, it includes the following steps:

[0055] Step 1, System Calibration: Establish the mapping relationship between the image coordinate system and the physical coordinate system of the circular calibration track 2 through the camera 3 above the circular calibration track 2;

[0056] Step 2, Data Acquisition: Place robot 1 on the circular calibration track 2 and control its constant speed movement; acquire motion video through camera 3, and the stress sensing plate 5 on the circular calibration track 2 generates a trigger signal for robot 1 and marks it on the video stream timeline;

[0057] Step 3, Visual Data Processing: The position of robot 1 is converted from pixel coordinates to physical coordinates to obtain a position sequence. Specifically, a target is attached to the top of robot 1. The pixel coordinates of the four corner points of the target in the image are extracted using a target detection algorithm. The pixel coordinates of the geometric center of the four corner points are taken as the position of robot 1, and the target center coordinates of consecutive frames are filtered. More specifically, the robot position in each frame is identified using target detection algorithms such as YOLO or SSD. A customized AprilTag or ArUco code is pasted on the top center of robot 1. The coordinates of its internal corner points are known. Instead of using the center of the detection box, the sub-pixel coordinates of the four corner points of the target in the image are extracted. The geometric center of the four corner points is taken as the position of robot 1, which can reduce the positioning error from tens of pixels to within 1-2 pixels. Kalman filtering or moving average is applied to the target center coordinates of consecutive frames to further suppress high-frequency jitter.

[0058] Step 4: Motion Trajectory Analysis: Based on the position sequence, calculate the instantaneous velocity and trajectory fitting circle parameters of robot 1. Separate the linear velocity and angular velocity components using the fitting circle parameters to optimize the instantaneous velocity of robot 1. Specifically, based on the position sequence, substitute the linear velocity and angular velocity calculation formulas to calculate the actual linear velocity and angular velocity of robot 1. Use the least squares method to fit the motion trajectory of robot 1 and solve for the fitting circle parameters. Calculate the average residual between the trajectory points of robot 1 and the fitting circle to verify the trajectory quality. Represent the tangential angle of the trajectory points using the fitting circle parameters, and optimize the angular velocity of the robot using the tangential angle expression.

[0059] More specifically, extract the robot's position coordinates in the image sequence. , forming a trajectory, in which ; The frame number; the position sequence is: ; Calculate the actual linear velocity based on timestamps and position changes and angular velocity The calculation formula is:

[0060] ,

[0061] ;

[0062] like Figure 4 As shown, the motion trajectory parameters are optimized through curve fitting, and the least squares method is used to fit a circular trajectory. The objective function is:

[0063] ;

[0064] in, To fit the coordinates of the center of the circle, Let be the radius of the fitted circle;

[0065] Based on the obtained fitted circle parameters, perform the following steps:

[0066] Trajectory quality verification: Calculate the average residual between the trajectory points and the fitted circle. The calculation formula is:

[0067] ,

[0068] If the residual is too large, the motion is considered unstable and data needs to be collected again.

[0069] Smoothing optimization of angular velocity calculation: The tangential angle of discrete position points can be expressed using the fitted circle parameters. At this point, the angular velocity can be optimized to This formula is more stable than the difference formula and can effectively suppress fluctuations caused by position detection noise.

[0070] Step 5, Parameter Calibration: Compare the optimized instantaneous velocity of robot 1 with the theoretical model to obtain the wheel diameter and wheelbase; specifically, introduce wheel diameter correction coefficient and wheelbase correction coefficient into the robot's theoretical motion model, establish a calibration model, compare the actual linear velocity and optimized angular velocity with the calibration model to obtain the wheel diameter correction coefficient and wheelbase correction coefficient, and then obtain the calibrated actual wheel diameter and wheelbase.

[0071] More specifically, a theoretical motion model for a two-wheeled differential robot is established, with the following basic formula:

[0072] ,

[0073] ;

[0074] in, The theoretical linear velocity, Theoretical angular velocity, For nominal wheel diameter, , The nominal angular velocities of the left and right wheels are... The nominal wheelbase;

[0075] Introducing wheel diameter correction factor Wheel spacing correction factor Establish a calibration model:

[0076] ,

[0077] ;

[0078] Solve the simultaneous equations to find the correction coefficients:

[0079] ,

[0080] ;

[0081] Output the actual wheel diameter after calibration and actual wheelbase ;

[0082] Step 5 is followed by Step 6: Repeatability test: Calibrate multiple times under the same conditions and evaluate the consistency of the results; specifically, the same robot 1 is calibrated 5 times at the same speed, and the range of the wheel diameter calibration results is ≤0.12mm and the standard deviation is ≤0.05mm; the range of the wheel track calibration results is ≤0.25mm and the standard deviation is ≤0.09mm.

[0083] Cross-validation: The robot's motion was controlled using the calibrated parameters to verify the degree of conformity between the actual trajectory and the expected trajectory; specifically, the robot 1 was controlled to move along a circular trajectory with a diameter of 3m using the calibrated parameters, and the maximum radial deviation between the actual trajectory and the theoretical trajectory decreased from 42mm before calibration to 3.5mm.

[0084] Comparative experiment: The advantages of the method of the present invention are evaluated by comparing the calibration results with those of the traditional method. Specifically, compared with the traditional encoder calibration method, the wheel diameter difference measured by the present method on three types of ground surfaces, namely ceramic tile, epoxy flooring and carpet, is less than 0.08%, while the difference of the traditional method on different ground surfaces is 1.2%-3.7%.

[0085] It should be noted that the parts in this embodiment that are the same as or similar to those in Embodiment 2 can be referred to each other, and will not be repeated in this application.

[0086] Example 4

[0087] As another embodiment, this fourth embodiment, based on the third embodiment, proposes an apparatus for calibrating the chassis speed of a two-wheeled differential robot based on external vision, such as... Figure 1 and Figure 2 As shown, the system includes: a circular calibration track 2, a camera 3, and a stress-sensing plate 5; the camera 3 is mounted on the circular calibration track 2 via a mounting bracket 4, and the camera's field of view covers the entire circular track; the circular calibration track 2 is made of a hard, flat material, and the surface of the circular calibration track 2 facing the camera 3 has angle scales 6 and corresponding radial position scales; the stress-sensing plate 5 is embedded below the radial position scales and is used in conjunction with the camera to determine the current position of the robot 1, while no longer being affected by ground friction factors, allowing calibration on different road surfaces; the circular track design allows the robot 1 to move continuously without multiple starts, stops, and turns, greatly shortening the calibration time; the stress-sensing plate 5 is connected to a data collection device 8 via a sensor data acquisition line 7.

[0088] It should be noted that the parts in this embodiment that are the same as or similar to those in Embodiment 3 can be referred to each other, and will not be repeated in this application.

[0089] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

Claims

1. A method for calibrating the chassis speed of a two-wheeled differential robot based on external vision, characterized in that, Includes the following steps: Step 1, System Calibration: Establish the mapping relationship between the image coordinate system and the physical coordinate system of the circular calibration track (2) through the camera (3) above the circular calibration track (2); Step 2, Data Acquisition: Place the robot (1) on the circular calibration track (2) and control its constant speed movement; acquire motion video through the camera (3), and the stress sensing plate (5) on the circular calibration track (2) generates a trigger signal for the robot (1) and marks it on the video stream timeline; Step 3, Visual Data Processing: Convert the robot's (1) position from pixel coordinates to physical coordinates to obtain the position sequence; Step 4, Motion Trajectory Analysis: Calculate the instantaneous velocity and trajectory fitting circle parameters of the robot (1) based on the position sequence. Separate the linear velocity and angular velocity components based on the fitting circle parameters to optimize the instantaneous velocity of the robot (1). Substitute the linear velocity and angular velocity calculation formulas into the position sequence to calculate the actual linear velocity and angular velocity of the robot (1). Fit the motion trajectory of the robot (1) using the least squares method to obtain the fitting circle parameters. Calculate the average residual between the trajectory points of the robot (1) and the fitting circle to verify the trajectory quality. Represent the tangential angle of the trajectory points using the fitting circle parameters and optimize the angular velocity of the robot using the tangential angle expression. Step 5, parameter calibration: Compare the instantaneous velocity of the optimized robot (1) with the theoretical model to obtain the wheel diameter and wheel spacing; introduce wheel diameter correction coefficient and wheel spacing correction coefficient into the robot's theoretical motion model to establish a calibration model, compare the actual linear velocity and optimized angular velocity with the calibration model to obtain the wheel diameter correction coefficient and wheel spacing correction coefficient, and then obtain the calibrated actual wheel diameter and wheel spacing.

2. The method for calibrating the chassis speed of a two-wheeled differential robot based on external vision according to claim 1, characterized in that, In step one, the circular calibration track (2) facing the camera (3) has an angle scale (6) and a corresponding radial position scale. Using the radial position scale, at least four non-collinear feature points with known physical coordinates are selected on the circular calibration track (2), and the pixel coordinates of the feature points in the image are collected. The homography matrix between the image coordinate system and the physical coordinate system is solved using the direct linear transformation algorithm.

3. The method for calibrating the chassis speed of a two-wheeled differential robot based on external vision according to claim 2, characterized in that, In step two, the robot (1) moves at least one revolution, and the stress sensing plate (5) is embedded below the radial position scale. When the robot (1) presses over the stress sensing plate (5), the stress sensing plate (5) generates a trigger signal.

4. The method for calibrating the chassis speed of a two-wheeled differential robot based on external vision according to claim 1, characterized in that, In step three, a target is attached to the top of the robot (1). The target detection algorithm is used to extract the pixel coordinates of the four corner points of the target in the image. The pixel coordinates of the geometric center of the four corner points are taken as the position of the robot (1). The target center coordinates of consecutive frames are filtered.

5. The method for calibrating the chassis speed of a two-wheeled differential robot based on external vision according to claim 1, characterized in that, Step 5 is followed by step 6: Repeated calibration is performed under the same conditions; the calibrated parameters are used to control the robot (1) to move and verify the degree of conformity between the actual trajectory and the expected trajectory; and the calibration results are compared with those of the traditional method.

6. An apparatus for calibrating the chassis speed of a two-wheeled differential robot based on external vision as described in any one of claims 1 to 5, characterized in that, include: A circular calibration track (2), a camera (3), and a stress-sensing plate (5); the camera (3) is mounted on the circular calibration track (2) via a mounting bracket (4); the surface of the circular calibration track (2) facing the camera (3) is provided with an angle scale (6) and a corresponding radial position scale, and the stress-sensing plate (5) is embedded below the radial position scale; the stress-sensing plate (5) is connected to a data collection device (8) via a sensor data acquisition line (7).

Images