Hand-eye relationship and tool center calibration method for three-dimensional vision guided robotic machining

Abstract

The application discloses a hand-eye relationship and tool center calibration method for three-dimensional vision guided robot machining, and adopts a spherical ball as a calibration tool to realize calibration of a tool center. When calibration is performed, one ball is fixed in a camera field of view, and the other ball is installed at the end of a cylindrical simulation tool. The end posture of the robot is changed, the balls collide with each other for multiple times, and the ball center is fitted through acquisition of a spherical point cloud of the camera at the same time. Based on the robot posture parameters and the ball center distance parameters, the current tool tip, i.e. the ball center of the end of the cylindrical simulation tool, and the hand-eye matrix of the three-dimensional vision camera and the robot base coordinate system are calibrated at the same time. The method can avoid high calibration cost caused by easy wear of a needle tip in a traditional method, reduce tool tip calibration error caused by non-real alignment of the needle tip, and further efficiently calibrate a high-precision hand-eye relationship. In addition, considering the difference between the length of the cylindrical simulation tool and the length of the track test needle tip and the length of an actual machining tool, the tool axis calibration and tool center compensation method can be used to quickly calculate the current tool tip point.

Description

Technical Field

[0001] This invention relates to the fields of robot calibration and camera calibration, specifically to a method for calibrating the hand-eye relationship and tool center in 3D vision-guided robot machining. Background Technology

[0002] With the continuous improvement of automation levels in the manufacturing industry, the application scenarios of industrial robots have been further expanded, such as spraying, welding, grinding and polishing, deburring and other fields. With the addition of vision cameras to automated production lines, the requirements and efficiency of industrial robots have been further improved.

[0003] The tool center point refers to the geometric center of the cutting tool on the end effector of an industrial robot, which is the actual point where the robot performs its operations. Therefore, in the production process, to ensure the robot performs precise movements, the tool center point needs to be accurately located. Tool center point calibration aims to achieve this goal, namely, determining the position of the robot's tool center point so that it can perform more accurate operations. Tool center point calibration is crucial to ensuring the robot completes accurate operations. With the continuous improvement of industrial automation, the importance of tool center point calibration is becoming increasingly prominent.

[0004] The calibration of industrial robots and vision cameras, also known as eye-to-hand calibration, can be approached in two ways: eye-on-the-hand and eye-to-hand. Eye-to-hand calibration refers to determining the coordinate transformation relationship between the robot's end effector coordinate system and the camera coordinate system, or between the robot's base coordinate system and the camera coordinate system. There are two scenarios for eye-to-hand calibration: the first is where the camera (eye) is fixed to the end effector of the robot arm (hand), and moves with the robot arm; this type of calibration is called Eye-in-hand. The second is where the camera (eye) is separated from the robot arm (hand), and the camera is fixed relative to the robot's base; the robot arm's movement has no effect on the camera; this type of calibration is called Eye-to-hand.

[0005] Traditional manual teaching methods, such as four-point calibration and six-point calibration of tool center, and the hand-eye calibration method of AX=XB, can no longer meet the high-precision requirements of vision camera-guided robot processing. Existing methods suffer from problems such as insufficient calibration stability, poor efficiency, and low accuracy.

[0006] Therefore, it is indeed necessary to improve existing technologies to address their shortcomings. Summary of the Invention

[0007] 1. The technical problem to be solved:

[0008] To address the aforementioned technical problems, this invention provides a method for hand-eye relationship and tool center calibration in 3D vision-guided robot machining. Addressing the issues of low accuracy and large human observation errors in existing tool center calibration methods for scenarios where the eye is outside the hand, and the complex, inaccurate, and time-consuming hand-eye calibration procedures between the camera and the robot, this invention proposes a highly efficient, high-precision, and tool center-compensated method for hand-eye relationship and tool center calibration in 3D vision-guided robot machining. This effectively solves the problems of large errors, low accuracy, long time consumption, increased errors after tool replacement, and inefficient repetitive calibration caused by manual operation in existing tool center calibration and hand-eye calibration processes.

[0009] 2. Technical Solution:

[0010] A method for hand-eye relationship and tool center calibration in 3D vision-guided robot machining, characterized by the following steps:

[0011] Step 1: Install the workpiece; the workpiece includes: a cylinder, two spheres of different sizes and known radii, and a 3D vision camera; the end effector of the industrial robot is equipped with a robot flange, an electric spindle-flange adapter plate, and an electric spindle connected in sequence; the electric spindle is used to fix and clamp the cutting tool / simulated cutting tool; the tail end of the cylinder is rigidly connected to sphere A to form a simulated cutting tool; sphere B is fixed in space, and the 3D vision camera is fixed in space; during calibration, both the simulated cutting tool and sphere B are within the field of view of the 3D vision camera; use a ruler to roughly measure the initial values ​​of the three coordinate components corresponding to the center of sphere A in the coordinate system of the robot flange center; the coordinate system of the robot flange center is a rectangular coordinate system constructed with the center point of the flange as the origin;

[0012] Step 2: Control the robot to make sphere A, which is fixed at the end of the robot, come into contact with sphere B. Repeat this process multiple times. The position of contact is different each time, and the orientation of the robot's flange is different. Record the six-position parameters of the flange in the robot's base coordinate system in the robot teach pendant at each contact time, and take a picture with the 3D vision camera to obtain the point cloud information of sphere A.

[0013] Step 3: Based on the initial values ​​of the three coordinate axes of the center of sphere A in the robot flange center coordinate system obtained in Step 1 and the six pose parameter data recorded in multiple sets of robot teach pendants in Step 2, calculate the projection distance of the vector formed by the center of sphere A to the center point of the robot flange on the three axes of the flange center coordinate system, i.e. the tool center offset value. Use this tool center offset value to establish the tool coordinate system.

[0014] Step 4: Calculate the new six-position parameters corresponding to the six-position parameters of the flange in the multiple robot teach pendants in Step 2 in the tool coordinate system, that is, the position parameters of the center of the sphere A in the robot base coordinate system.

[0015] Step 5: Based on the point cloud of sphere A obtained by the 3D vision camera in Step 2, fit the point cloud of sphere A corresponding to each contact, and obtain the center of sphere A corresponding to each contact in the 3D vision camera coordinate system.

[0016] Step Six: Based on the point cloud of the center of sphere A in the robot base coordinate system obtained in Step Four and the point cloud of the center of sphere A in the 3D vision camera coordinate system obtained in Step Five, use the point set matching method to register the point cloud of the center of sphere A in the 3D vision camera coordinate system corresponding to each contact to the point cloud of the center of sphere A in the robot base coordinate system, and obtain the transformation matrix between the 3D vision camera coordinate system and the robot base coordinate system, i.e., the hand-eye matrix;

[0017] Step 7: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with a rigid connection between the needle tip and the cylinder. Determine the axial direction of the needle tip tool, calculate the center point of the needle tip tool, i.e. the tip point, and establish a new needle tip tool coordinate system.

[0018] Furthermore, step three specifically includes:

[0019] S31: Let the initial values ​​of the coordinate components of the center of sphere A on the three axes of the robot flange center coordinate system be x0, y0, and z0, respectively. Let the six pose parameters (x, y, z, w, p, r) of the flange in the robot teach pendant be as follows:

[0020]

[0021] In the above formula, n is the number of times sphere A and sphere B come into contact;

[0022] The transformation matrix from the flange center coordinate system to the robot base coordinate system is:

[0023]

[0024] In the above formula, i ≤ n;

[0025] S32: Use a nonlinear optimization method to find the tool center offset value x. tcp y tcp z tcp That is, the vector formed by the center of sphere A to the center point of the robot flange corresponds to the true distance on the three coordinate axis components in the coordinate system of the robot flange center; so that the distance between the center of the robot end sphere A, obtained according to the tool center offset value, and the center of the sphere B fixed in space is the sum of the radii of the two spheres; the objective equation of this nonlinear optimization is expressed as:

[0026]

[0027] (3) In the formula, x tcpy tcp z tcp This represents the offset of the tool center along the three coordinate axes of the robot flange center coordinate system; with the midpoint (x) of the flange center coordinate system as the reference point. tcp y tcp z tcp The tool coordinate system is constructed using ) as the origin; r is the sum of the radii of spheres A and B; x ti y ti z ti The position data of the center of the robot's end effector sphere in the robot's base coordinate system is obtained by transforming the initial distance values ​​(x0, y0, z0) in the i-th data set. The specific transformation process is as follows:

[0028]

[0029] Furthermore, the specific process for calculating the pose parameters of the center of sphere A in the robot's base coordinate system in step four is as follows:

[0030] Based on the tool center offset value x tcp y tcp z tcp The tool center offset vector TCP is then denoted as:

[0031]

[0032] Let P be the center point of the flange in the robot teach pendant under the robot base coordinate system. i F (x i y i z i The center point of sphere A in the robot's base coordinate system The pose parameters of the center of sphere A in the robot's base coordinate system are:

[0033]

[0034] Furthermore, step four also includes calculating the accuracy of the tool coordinate system; specifically, using the pose parameters of the center of sphere A in the robot's base coordinate system, a point cloud of the center point of sphere A is created, a circle is fitted using the point cloud of the center point of sphere A, the radius R of this fitted circle is measured, and compared with the sum r of the radii of sphere A and sphere B to verify the accuracy of the tool coordinate system obtained in step three; the accuracy formula is:

[0035] Err = Rr (7);

[0036] If the accuracy Err of the tool coordinate system is lower than the preset value, starting from step one, remeasure the initial values ​​of the three coordinate components corresponding to the center of the sphere in the robot flange center coordinate system, repeat steps two and three, re-establish the tool coordinate system, and re-verify the accuracy of the newly established tool coordinate system obtained in step three until the preset accuracy requirements are met.

[0037] Furthermore, step seven specifically includes:

[0038] S71: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with the rigid connection between the needle tip and the cylinder. Calculate the theoretical offset distance l between the tip of the needle and the center of the spherical shell based on the radius of the spherical shell and the length of the needle tip.

[0039] S72: Control the robot to place the new cylindrical needle tip simulated tool in the field of view of the 3D vision camera. The robot moves again to bring the cylindrical needle tip simulated tool within the field of view of the 3D vision camera and takes a picture. Record the six pose parameters in the tool coordinate system that have been updated in the robot teach pendant. The first three X, Y, and Z of these six pose parameters are also the coordinates of the center of the sphere A in the robot base coordinate system. The 3D vision camera takes a picture to obtain the point cloud information of the cylinder.

[0040] S73: Based on the point cloud of the cylinder obtained in S72, fit the cylinder to obtain the direction of the cylinder's central axis. Compare this direction with the three-axis direction of the robot flange center coordinate system to obtain the angle (α, β, γ) between the central axis of the cylinder and the three axes of the flange center coordinate system. This angle is the actual tool installation angle. Project the center point of the sphere A obtained in S72 in the robot base coordinate system onto the cylinder axis to obtain the center projection point of the sphere A. Offset this projection point along the cylinder axis direction by a certain distance. This distance is the offset distance between the theoretical needle tip obtained in S71 and the center of the sphere A. A new point, i.e., the needle tip, is obtained along the cylinder axis direction.

[0041] S74: Based on the theoretical offset distance l between the tip of the cutting tool and the center of the sphere A obtained in S71, and the installation angles α, β, and γ in the three axes of the robot flange coordinate system obtained in S73, calculate the offset distance between the new cutting tool tip and the original cutting tool tip in the three axes of the robot flange coordinate system. Establish a new needle tip tool coordinate system:

[0042]

[0043] 3. Beneficial effects:

[0044] (1) In this calibration method, the tool center calibration method of ball-to-ball contact is adopted to replace the traditional manual teaching method such as four-point calibration, six-point calibration and needle-to-needle tool center calibration method. The six pose parameters of the flange are obtained by ball-to-ball contact. The distance between the tool ball connected to the end of the flange and the fixed ball in space during the ball-to-ball contact experiment is a constant value. The tool center point of the ball tool is calculated by using this constant distance as a constraint. Since the curvature of the ball is large, it is not easy to wear during the contact experiment. Through multiple sets of experiments, the interference of random errors is eliminated. The experimental results are highly accurate and the tool center point calibration accuracy is high. It reduces the error caused by manual observation and operation and the cost of easy needle tip wear, and improves the accuracy of tool center calibration.

[0045] (2) This method replaces the traditional needle-to-needle-point calibration method, avoids the high calibration cost caused by easy wear of the needle tip, reduces the calibration error caused by the needle tip not being truly aligned, and can simultaneously and efficiently calibrate a high-precision hand-eye relationship.

[0046] (3) In this invention, the difference between the cylindrical simulated tool and the trajectory test needle tip and the actual machining tool length can be considered. The current tool tip point can be quickly calculated by the tool axis calibration and tool center compensation method in this invention.

[0047] In summary, this calibration method, while meeting the accuracy requirements of tool center calibration, uses point clouds obtained from images taken by a 3D vision camera whose tool center calibration accuracy has been verified for hand-eye calibration using a point set matching method. This ensures the reliability of the hand-eye calibration data, reduces the time required for re-collecting hand-eye calibration data after tool center calibration, and improves the accuracy and speed of hand-eye calibration. Furthermore, it enables rapid calibration of tool tips of different lengths and accuracy verification by replacing needle-tip tools through tool axis calibration and tool center compensation updates. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the hand-eye relationship and tool center calibration process in the 3D vision-guided robot machining of this invention.

[0049] Figure 2 This is a schematic diagram of the hand-eye relationship and tool center calibration in the 3D vision-guided robot machining process of this invention;

[0050] Figure 3 This is a schematic diagram of the tool axis calibration and correction in this invention;

[0051] Figure 4 This is a schematic diagram of the tool center compensation update in this invention;

[0052] Figure 5 This is a scene diagram showing the hand-eye relationship and tool center calibration in the 3D vision-guided robot processing of this invention;

[0053] Figure 6 This is a chart verifying the accuracy of the hand-eye relationship in the 3D vision-guided robot processing of this invention.

[0054] Figure reference numerals: 1. Industrial robot; 2. 3D vision camera; 3. Robot flange; 4. Electric spindle-flange adapter plate; 5. Electric spindle; 6. Cylinder; 7. Sphere A; 8. Sphere B; 9. Needle tip; 10. Cone. Detailed Implementation

[0055] The present invention will now be described in detail with reference to the accompanying drawings.

[0056] As attached Figure 1 As shown, a method for hand-eye relationship and tool center calibration in 3D vision-guided robot machining is characterized by the following steps:

[0057] Step 1: Install the workpiece; the workpiece includes: a cylinder, two spheres of different sizes and known radii, and a 3D vision camera; the end effector of the industrial robot is equipped with a robot flange, an electric spindle-flange adapter plate, and an electric spindle connected in sequence; the electric spindle is used to fix and clamp the cutting tool / simulated cutting tool; the tail end of the cylinder is rigidly connected to sphere A to form a simulated cutting tool; sphere B is fixed in space, and the 3D vision camera is fixed in space; during calibration, both the simulated cutting tool and sphere B are within the field of view of the 3D vision camera; use a ruler to roughly measure the initial values ​​of the three coordinate components corresponding to the center of sphere A in the coordinate system of the robot flange center; the coordinate system of the robot flange center is a rectangular coordinate system constructed with the center point of the flange as the origin;

[0058] Step 2: Control the robot to make sphere A, which is fixed at the end of the robot, come into contact with sphere B. Repeat this process multiple times. The position of contact is different each time, and the orientation of the robot's flange is different. Record the six-position parameters of the flange in the robot's base coordinate system in the robot teach pendant at each contact time, and take a picture with the 3D vision camera to obtain the point cloud information of sphere A.

[0059] Step 3: Based on the initial values ​​of the three coordinate axes of the center of sphere A in the robot flange center coordinate system obtained in Step 1 and the six pose parameter data recorded in multiple sets of robot teach pendants in Step 2, calculate the projection distance of the vector formed by the center of sphere A to the center point of the robot flange on the three axes of the flange center coordinate system, i.e. the tool center offset value. Use this tool center offset value to establish the tool coordinate system.

[0060] Step 4: Calculate the new six-position parameters corresponding to the six-position parameters of the flange in the multiple robot teach pendants in Step 2 in the tool coordinate system, that is, the position parameters of the center of the sphere A in the robot base coordinate system.

[0061] Step 5: Based on the point cloud of sphere A obtained by the 3D vision camera in Step 2, fit the point cloud of sphere A corresponding to each contact, and obtain the center of sphere A corresponding to each contact in the 3D vision camera coordinate system.

[0062] Step Six: Based on the point cloud of the center of sphere A in the robot base coordinate system obtained in Step Four and the point cloud of the center of sphere A in the 3D vision camera coordinate system obtained in Step Five, use the point set matching method to register the point cloud of the center of sphere A in the 3D vision camera coordinate system corresponding to each contact to the point cloud of the center of sphere A in the robot base coordinate system, and obtain the transformation matrix between the 3D vision camera coordinate system and the robot base coordinate system, i.e., the hand-eye matrix;

[0063] Step 7: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with a rigid connection between the needle tip and the cylinder. Determine the axial direction of the needle tip tool, calculate the center point of the needle tip tool, i.e. the tip point, and establish a new needle tip tool coordinate system.

[0064] Furthermore, step three specifically includes:

[0065] S31: Let the initial values ​​of the coordinate components of the center of sphere A on the three axes of the robot flange center coordinate system be x0, y0, and z0, respectively. Let the six pose parameters (x, y, z, w, p, r) of the flange in the robot teach pendant be as follows:

[0066]

[0067] In the above formula, n is the number of times sphere A and sphere B come into contact;

[0068] The transformation matrix from the flange center coordinate system to the robot base coordinate system is:

[0069]

[0070] In the above formula, i ≤ n;

[0071] S32: Use a nonlinear optimization method to find the tool center offset value x. tcp y tcp z tcp That is, the vector formed by the center of sphere A to the center point of the robot flange corresponds to the true distance on the three coordinate axis components in the coordinate system of the robot flange center; so that the distance between the center of the robot end sphere A, obtained according to the tool center offset value, and the center of the sphere B fixed in space is the sum of the radii of the two spheres; the objective equation of this nonlinear optimization is expressed as:

[0072]

[0073] (3) In the formula, x tcpy tcp z tcp This represents the offset of the tool center along the three coordinate axes of the robot flange center coordinate system; with the midpoint (x) of the flange center coordinate system as the reference point. tcp y tcp z tcp The tool coordinate system is constructed using ) as the origin; r is the sum of the radii of spheres A and B; x ti y ti z ti The position data of the center of the robot's end effector sphere in the robot's base coordinate system is obtained by transforming the initial distance values ​​(x0, y0, z0) in the i-th data set. The specific transformation process is as follows:

[0074]

[0075] Furthermore, the specific process for calculating the pose parameters of the center of sphere A in the robot's base coordinate system in step four is as follows:

[0076] Based on the tool center offset value x tcp y tcp z tcp The tool center offset vector TCP is then denoted as:

[0077]

[0078] Let P be the center point of the flange in the robot teach pendant under the robot base coordinate system. i F (x i y i z i The center point of sphere A in the robot's base coordinate system The pose parameters of the center of sphere A in the robot's base coordinate system are:

[0079]

[0080] Furthermore, step four also includes calculating the accuracy of the tool coordinate system; specifically, using the pose parameters of the center of sphere A in the robot's base coordinate system, a point cloud of the center point of sphere A is created, a circle is fitted using the point cloud of the center point of sphere A, the radius R of this fitted circle is measured, and compared with the sum r of the radii of sphere A and sphere B to verify the accuracy of the tool coordinate system obtained in step three; the accuracy formula is:

[0081] Err = Rr (7);

[0082] If the accuracy Err of the tool coordinate system is lower than the preset value, starting from step one, remeasure the initial values ​​of the three coordinate components corresponding to the center of the sphere A in the coordinate system of the robot flange center. Repeat steps two and three to re-establish the tool coordinate system and re-verify the accuracy of the newly established tool coordinate system obtained in step three until the accuracy requirements are met.

[0083] Furthermore, step seven specifically includes:

[0084] S71: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with the rigid connection between the needle tip and the cylinder. Calculate the theoretical offset distance l between the tip of the needle and the center of the spherical shell based on the radius of the spherical shell and the length of the needle tip.

[0085] S72: Control the robot to place the new cylindrical needle tip simulated tool in the field of view of the 3D vision camera. The robot moves again to bring the cylindrical needle tip simulated tool within the field of view of the 3D vision camera and takes a picture. Record the six pose parameters in the tool coordinate system that have been updated in the robot teach pendant. The first three X, Y, and Z of these six pose parameters are also the coordinates of the center of the sphere A in the robot base coordinate system. The 3D vision camera takes a picture to obtain the point cloud information of the cylinder.

[0086] S73: Based on the point cloud of the cylinder obtained in S72, fit the cylinder to obtain the direction of the cylinder's central axis. Compare this direction with the three-axis direction of the robot flange center coordinate system to obtain the angle (α, β, γ) between the central axis of the cylinder and the three axes of the flange center coordinate system. This angle is the actual tool installation angle. Project the center point of the sphere A obtained in S72 in the robot base coordinate system onto the cylinder axis to obtain the center projection point of the sphere A. Offset this projection point along the cylinder axis direction by a certain distance. This distance is the offset distance between the theoretical needle tip obtained in S71 and the center of the sphere A. A new point, i.e., the needle tip, is obtained along the cylinder axis direction.

[0087] S74: Based on the theoretical offset distance l between the tip of the cutting tool and the center of the sphere A obtained in S71, and the installation angles α, β, and γ in the three axes of the robot flange coordinate system obtained in S73, calculate the offset distance between the new cutting tool tip and the original cutting tool tip in the three axes of the robot flange coordinate system. Establish a new needle tip tool coordinate system:

[0088] Specific implementation examples:

[0090] As attached Figure 1 The diagram shown is a flowchart illustrating the hand-eye relationship and tool center calibration process in this specific embodiment of 3D vision-guided robot machining. This solution is applicable to layouts where the eye is outside the hand, as shown in the attached diagram. Figure 2 ,5 6. The three-dimensional vision-guided robot processing scene, hand-eye relationship, tool center calibration diagram and tool involved in this invention. Figure 2 A schematic diagram of hand-eye relationship and tool center calibration; both the 3D vision camera 2 and sphere B 8 are fixed in space. The robot 1 is controlled to move, causing sphere A 7 to move within the field of view of the 3D vision camera 2. When sphere A 7 contacts sphere B 8, the pose parameters of the robot flange center are recorded. The 3D vision camera takes a picture of the target and records the point cloud data of sphere A 7. The contact experiment is repeated. The experimental data is attached. Figure 1 The process shown uses a point set matching method to calculate the hand-eye matrix. It matches the point cloud that has been verified for accuracy by the tool center calibration method with the point cloud obtained by fitting the image taken by the 3D vision camera, avoiding repeated experiments and improving accuracy and efficiency.

[0091] After calibration, the needle tip is changed to verify accuracy, including tool axis calibration and tool center compensation update, as shown in the attached document. Figure 3 This is a schematic diagram for tool axis calibration and correction. In the diagram, the robot flange coordinate system is {O}. F xyz}, direction as follows Figure 3 As shown, during the installation of the adapter plate 4, electric spindle 5, and cylinder 6 with the tool, installation angle errors inevitably occur, and the axis direction of the tool is not the axis direction of the robot flange coordinate system; as shown in the attached diagram. Figure 4 This is a schematic diagram of tool center compensation update. When the tool is replaced or the tool tip is changed, the magnitude of the tool center compensation value is linearly related to the tool axis offset angle and the tool center offset distance.

[0092] Accuracy is verified using a needle tip, such as Figure 6 As shown in the figure, 10 represents the needle tip. The figure shows the data from multiple experiments, the hand-eye matrix, and their corresponding precision Err.

[0093] Although the present invention has been disclosed above with reference to preferred embodiments, these are not intended to limit the invention. Any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention should be defined by the scope of the claims of this application.

Claims

1. A method for hand-eye relationship and tool center calibration in 3D vision-guided robot machining, characterized in that: Includes the following steps: Step 1: Install the workpiece; the workpiece includes: a cylinder, two spheres of different sizes and known radii, and a 3D vision camera; the end effector of the industrial robot is equipped with a robot flange, an electric spindle-flange adapter plate, and an electric spindle connected in sequence; the electric spindle is used to fix and clamp the cutting tool / simulated cutting tool; the tail end of the cylinder is rigidly connected to sphere A to form a simulated cutting tool; sphere B is fixed in space, and the 3D vision camera is fixed in space; during calibration, both the simulated cutting tool and sphere B are within the field of view of the 3D vision camera; use a ruler to roughly measure the initial values ​​of the three coordinate components corresponding to the center of sphere A in the coordinate system of the robot flange center; the coordinate system of the robot flange center is a rectangular coordinate system constructed with the center point of the flange as the origin; Step 2: Control the robot to make sphere A, which is fixed at the end of the robot, come into contact with sphere B. Repeat this process multiple times. The position of contact is different each time, and the orientation of the robot's flange is different. Record the six-position parameters of the flange in the robot's base coordinate system in the robot teach pendant at each contact time, and take a picture with the 3D vision camera to obtain the point cloud information of sphere A. Step 3: Based on the initial values ​​of the three coordinate axes of the center of sphere A in the robot flange center coordinate system obtained in Step 1 and the six pose parameter data recorded in multiple sets of robot teach pendants in Step 2, calculate the projection distance of the vector formed by the center of sphere A to the center point of the robot flange on the three axes of the flange center coordinate system, i.e. the tool center offset value. Use this tool center offset value to establish the tool coordinate system. Step 4: Calculate the new six-position parameters corresponding to the six-position parameters of the flange in the multiple robot teach pendants in Step 2 in the tool coordinate system, that is, the position parameters of the center of the sphere A in the robot base coordinate system. Step 5: Based on the point cloud of sphere A obtained by the 3D vision camera in Step 2, fit the point cloud of sphere A corresponding to each contact, and obtain the center of sphere A corresponding to each contact in the 3D vision camera coordinate system. Step Six: Based on the point cloud of the center of sphere A in the robot base coordinate system obtained in Step Four and the point cloud of the center of sphere A in the 3D vision camera coordinate system obtained in Step Five, use the point set matching method to register the point cloud of the center of sphere A in the 3D vision camera coordinate system corresponding to each contact to the point cloud of the center of sphere A in the robot base coordinate system, and obtain the transformation matrix between the 3D vision camera coordinate system and the robot base coordinate system, i.e., the hand-eye matrix; Step 7: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with a rigid connection between the needle tip and the cylinder. Determine the axial direction of the needle tip tool, calculate the center point of the needle tip tool, i.e. the tip point, and establish a new needle tip tool coordinate system.

2. The hand-eye relationship and tool center calibration method for three-dimensional vision-guided robot machining according to claim 1, characterized in that: Step three specifically includes: S31: Let the initial values ​​of the coordinate components of the center of sphere A on the three axes of the robot flange center coordinate system be x0, y0, and z0, respectively. Let the six pose parameters (x, y, z, w, p, r) of the flange in the robot teach pendant be as follows: In the above formula, n is the number of times sphere A and sphere B come into contact; The transformation matrix from the flange center coordinate system to the robot base coordinate system is: In the above formula, i ≤ n; S32: Use a nonlinear optimization method to find the tool center offset value x. tcp y tcp z tcp That is, the vector formed by the center of sphere A to the center point of the robot flange corresponds to the true distance on the three coordinate axis components in the coordinate system of the robot flange center; so that the distance between the center of the robot end sphere A, obtained according to the tool center offset value, and the center of the sphere B fixed in space is the sum of the radii of the two spheres; the objective equation of this nonlinear optimization is expressed as: (3) In the formula, x tcp y tcp z tcp This represents the offset of the tool center along the three coordinate axes of the robot flange center coordinate system; with the midpoint (x) of the flange center coordinate system as the reference point. tcp y tcp z tcp The tool coordinate system is constructed using ) as the origin; r is the sum of the radii of spheres A and B; x ti y ti z ti The position data of the center of the robot's end effector sphere in the robot's base coordinate system is obtained by transforming the initial distance values ​​(x0, y0, z0) in the i-th data set. The specific transformation process is as follows:

3. The hand-eye relationship and tool center calibration method for three-dimensional vision-guided robot machining according to claim 1, characterized in that: The specific process for calculating the pose parameters of the center of sphere A in the robot's base coordinate system in step four is as follows: Based on the tool center offset value x tcp y tcp z tcp The tool center offset vector TCP is then denoted as: Let P be the center point of the flange in the robot teach pendant under the robot base coordinate system. i F (x i y i z i The center point of sphere A in the robot's base coordinate system The pose parameters of the center of sphere A in the robot's base coordinate system are:

4. The hand-eye relationship and tool center calibration method for three-dimensional vision-guided robot machining according to claim 3, characterized in that: Step four also includes calculating the accuracy of the tool coordinate system; specifically, using the pose parameters of the center of sphere A in the robot's base coordinate system, a point cloud of the center point of sphere A is created. A circle is fitted using the point cloud of the center point of sphere A, and the radius R of this fitted circle is measured. This radius R is then compared with the sum r of the radii of sphere A and sphere B to verify the accuracy of the tool coordinate system obtained in step three. The accuracy formula is: Err = Rr (7); If the accuracy Err of the tool coordinate system is lower than the preset value, starting from step one, remeasure the initial values ​​of the three coordinate components corresponding to the center of the sphere in the robot flange center coordinate system, repeat steps two and three, re-establish the tool coordinate system, and re-verify the accuracy of the newly established tool coordinate system obtained in step three until the preset accuracy requirements are met.

5. The method for hand-eye relationship and tool center calibration in three-dimensional vision-guided robot machining according to claim 4, characterized in that: Step seven specifically includes: S71: Disconnect the rigid connection between the cylinder and the spherical shell in the simulated tool and replace it with the rigid connection between the needle tip and the cylinder. Calculate the theoretical offset distance l between the tip of the needle and the center of the spherical shell based on the radius of the spherical shell and the length of the needle tip. S72: Control the robot to place the new cylindrical needle tip simulated tool in the field of view of the 3D vision camera. The robot moves again to bring the cylindrical needle tip simulated tool within the field of view of the 3D vision camera and takes a picture. Record the six pose parameters in the tool coordinate system that have been updated in the robot teach pendant. The first three X, Y, and Z of these six pose parameters are also the coordinates of the center of the sphere A in the robot base coordinate system. The 3D vision camera takes a picture to obtain the point cloud information of the cylinder. S73: Based on the point cloud of the cylinder obtained in S72, fit the cylinder to obtain the direction of the cylinder's central axis. Compare this direction with the three-axis direction of the robot flange center coordinate system to obtain the angle (α, β, γ) between the central axis of the cylinder and the three axes of the flange center coordinate system. This angle is the actual tool installation angle. Project the center point of the sphere A obtained in S72 in the robot base coordinate system onto the cylinder axis to obtain the center projection point of the sphere A. Offset this projection point along the cylinder axis direction by a certain distance. This distance is the offset distance between the theoretical needle tip obtained in S71 and the center of the sphere A. A new point, i.e., the needle tip, is obtained along the cylinder axis direction. S74: Based on the theoretical offset distance l between the tip of the cutting tool and the center of the sphere A obtained in S71, and the installation angles α, β, and γ in the three axes of the robot flange coordinate system obtained in S73, calculate the offset distance between the new cutting tool tip and the original cutting tool tip in the three axes of the robot flange coordinate system. Establish a new needle tip tool coordinate system:

Images