Hand-eye calibration method based on line structured light contourgraph

Abstract

The invention relates to the technical field of gluing application, in particular to a hand-eye calibration method based on a line structured light contourgraph, which comprises the following steps of: 1, calibrating a translation relation # imgabs0 # between a tool coordinate system Otool at the tail end of a probe and a coordinate system Oend at the tail end of a robot flange by a TCP sharp point touch four-point method, and 2, calibrating a translation relation # imgabs0 # between a tool coordinate system Otool at the tail end of the probe and a coordinate system Oend at the tail end of a robot flange by a TCP sharp point touch three-point method; step 3, manually aligning the calibration plate coordinate system Ocalobj and the line structured light contourgraph coordinate system Ocam, and carrying out rigid body transformation on the calibration plate coordinate system Ocalobj and the line structured light contourgraph coordinate system Ocam according to a rigid body transformation relation # imgabs1 # of the calibration plate coordinate system Ocalobj and the line structured light contourgraph coordinate system Ocam; 4, calculating to obtain a rigid body transformation relation # imgabs3 # of the coordinate system Ocam of the line structured light contourgraph and the coordinate system Oend of the flange tail end of the robot. According to the hand-eye calibration method based on the line structured light contourgraph, the hand-eye calibration efficiency of the line structured light contourgraph can be improved, and the detection precision can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of glue application, and particularly relates to a hand-eye calibration method based on a line structured light profiler. Background Art

[0002] In the automotive and lithium battery industries, there are a large number of glue application scenarios. The size, shape, position, and quality of the glue are extremely important for preventing leakage and ensuring the structural strength of the product. The measurement requirements for 3D metrics such as glue height have led to the application of 3D vision inspection in glue inspection. Among them, due to its high precision, line structured light is often used to scan the glue profile with a motion mechanism (robot or servo, etc.) carrying a line structured light profiler after the glue application is completed. However, for some glue inspection stations with high beat requirements, the above method cannot be adopted, and real-time inspection during the glue application process is required, that is, the line structured light profiler is installed around the glue gun and moves with the glue gun to perform real-time data acquisition and processing.

[0003] Generally, to obtain a depth map using a line structured light profiler and ensure that the scanned object does not distort, the motion direction of the motion mechanism must be parallel to the Y-axis of the sensor's light plane, and the sampling frequency and interval set by the camera must match the speed of the motion mechanism. This is because the line structured light profiler can only obtain a single contour on the light plane at a given moment and has no Y-axis data. For the image coordinate system, the X-axis and Y-axis must be perpendicular. Therefore, other conditions must be used to restrict the Y-axis motion direction of the line structured light profiler so that the collected contour data can be stitched together to form the true contour of the object.

[0004] During real-time glue inspection, the glue profile changes according to project requirements, and there may be a situation where the motion direction is not parallel to the sensor's Y-axis. Moreover, due to physical limitations such as tooling and pipelines, the glue gun is generally not rotated during glue application, so it is difficult to rotate the glue gun to cooperate with vision inspection. If the method of installing multiple line structured light profilers around the glue gun is adopted, to obtain real glue profile data, the glue profile must be unified in a fixed coordinate system for reconstruction. For a single profiler, the profile can be unified into the robot base coordinate system through real-time robot pose transformation. However, for a system with multiple line structured light profilers, hand-eye calibration must be performed to clarify the relationship between each profiler and the robot, which requires the robot to take multiple photos, resulting in low efficiency.

[0005] Therefore, the present invention proposes a hand-eye calibration method based on a line structured light profiler. Summary of the Invention

[0006] To solve the problem of low hand-eye calibration efficiency in existing systems with line structured light profilers, the present invention proposes a hand-eye calibration method based on a line structured light profiler, which can improve the hand-eye calibration efficiency and the detection accuracy.

[0007] The technical solution of the present invention is as follows:

[0008] A hand-eye calibration method based on a line-structured light profiler, including a target,

[0009] The target is fixed at a position reachable by the robot in space, so that the coordinate system O of the calibration plate calobj has a fixed relative position relationship with the robot base coordinate system O base Install the probe at a certain position close to the end of the robot flange, so that the coordinate system O of the end of the robot flange end and the tool coordinate system O at the end of the probe tool have a fixed relative position relationship; the method includes the following steps:

[0010] Step 1: TCP tip touch four-point method to calibrate the translation relationship between the tool coordinate system O at the end of the probe tool and the coordinate system O of the end of the robot flange end

[0011] Step 2: TCP tip touch three-point method to calibrate the rigid body transformation relationship between the coordinate system O of the calibration plate calobj and the robot base coordinate system O base

[0012] Step 3: Manually align the coordinate system O of the calibration plate calobj and the coordinate system O of the line-structured light profiler cam to obtain the rigid body transformation relationship between them

[0013] Step 4: Calculate the rigid body transformation relationship between the coordinate system O of the line-structured light profiler cam and the coordinate system O of the end of the robot flange end

[0014] Furthermore, step 1 is specifically:

[0015] Move the robot, select n postures with large joint differences, so that the origin of the tool coordinate system O at the end of the probe tool always coincides with the origin of the coordinate system O of the calibration plate calobj to obtain n homogeneous transformation matrices of the coordinate system O of the end of the robot flange end relative to the robot base coordinate system O base ;

[0016] The transformation relationship between the tool coordinate system O at the end of the probe tool and the robot base coordinate system O base is: For:

[0017]

[0018] where represents the transformation relationship from the coordinate system O at the end of the robot flange end to the coordinate system O of the robot base base , represents the transformation relationship between the tool coordinate system O at the end of the probe tool and the coordinate system O at the end of the robot flange end .

[0019] The block form is:

[0020]

[0021] Finally, we get:

[0022] is the rotation matrix of the coordinate system O at the end of the robot flange end relative to the coordinate system O of the robot base base . Since the origin of the tool coordinate system O at the end of the probe tool coincides with the origin of the calibration plate coordinate system O calobj , and the relative position relationship between the calibration plate coordinate system O calobj and the coordinate system O of the robot base base is fixed, then the translation relationship between the tool coordinate system O at the end of the probe tool and the coordinate system O of the robot base base has a fixed value of . For the robot in n different postures, there are

[0023]

[0024] Furthermore, we get:

[0025]

[0026] where [·] T represents the matrix transpose

[0027] Furthermore, step 2 is specifically: control the robot to make the probe touch m points on the target. The coordinates of the touched points are known in the calibration plate coordinate system Q calobj . Obtain the m homogeneous transformation matrices of the coordinate system O at the end of the robot flange end relative to the coordinate system O of the robot base base when the probe touches these points respectively

[0028] The coordinates of the points on the target in the coordinate system O of the robot base base are:

[0029]

[0030] The coordinates of the points on the target in the calibration plate coordinate system O calobj are Then, based on the least squares method, the rigid body transformation relationship between the calibration plate coordinate system O calobj and the robot base coordinate system O base is The formula is as follows:

[0031]

[0032] Furthermore, step 3 is specifically: Using the target shape, the calibration plate coordinate system O calobj and the line structured light profilometer coordinate system O cam are aligned to eliminate the rotation relationship of the rigid body transformation between the calibration plate coordinate system O calobj and the line structured light profilometer coordinate system O cam and the translation relationship is obtained based on the contour;

[0033] The rotation of the rigid body transformation can be described using Euler angles and is represented by the following formula:

[0034] R gba = R x (ψ)R y (θ)R z (φ)R x (ψ), R y (θ), R z (φ) respectively represent the 3×3 rotation matrices about the X-axis, Y-axis, and Z-axis;

[0035] According to different installation angles, there are 4 cases for the rotation angle, which are: φ = 0, φ = 90°, φ = 180°, φ = 270°;

[0036] Therefore, the rotation matrix in the transformation relationship calobj from the calibration plate coordinate system O cam to the line structured light profilometer coordinate system O is:

[0037]

[0038] where, E3 represents the 3×3 identity matrix;

[0039] Determine the translation relationship That is, determine the origin of the calibration plate coordinate system O calobj in the line structured light profilometer coordinate system O camThe translation coordinates below. For a line-structured light profiler, the y-value of the coordinates of the points falling on its light plane must be 0 in its coordinate system. Therefore, from the profile data collected by the line-structured light profiler, it is obtained that:

[0040]

[0041] Therefore, the rigid body transformation relationship can be expressed as:

[0042]

[0043] And the rigid body transformation relationship of the coordinate system O at the end of the robot flange end relative to the coordinate system O of the robot base base is obtained

[0044] Furthermore, step 4 is specifically as follows:

[0045] The expression of the rigid body transformation relationship is as follows:

[0046]

[0047] Furthermore, after step 4, multiple groups of rigid body transformation relationships

[0048] At time t0, the point set collected by the i-th profiler in the coordinate system O of the robot base base has the coordinates:

[0049]

[0050] where refers to the rigid body transformation relationship of the coordinate system O at the end of the robot flange relative to the coordinate system O of the robot base at time t0 end relative to the coordinate system O of the robot base base ;

[0051] After merging the point sets of all profilers from time t0 to t n , a point cloud is reconstructed and point cloud processing is performed.

[0052] Furthermore, it also includes the inspection of the hand-eye calibration accuracy, and the inspection method is as follows:

[0053] Control the robot to make the origin of the calibration plate coordinate system O calobj fall on the light plane, and obtain the coordinates of the origin of the calibration plate coordinate system O calobj in the coordinate system O of the line-structured light profiler cam at this time Record the coordinate system O at the end of the robot flange at this time end Relative to the robot base coordinate system O base The rigid body transformation relationship Through the hand-eye relationship matrix, transform the origin coordinates of the calibration plate coordinate system O calobj To the robot base coordinate system O base There is:

[0054]

[0055] Where Refers to the origin coordinates of the reconstructed calibration plate coordinate system O calobj Origin coordinates.

[0056] Change the robot pose multiple times to make the origin of the calibration plate coordinate system O calobj Always on the light plane. Since in the ideal case, the origin of the reconstructed calibration plate coordinate system O calobj The coordinates in the robot base coordinate system O base Should be the same as those in the results of the three-point method of TCP tip touch. Therefore, the calibration error is calculated based on this. Same, so the calibration error calculation is carried out based on this.

[0057] The error calculation formula can be obtained:

[0058]

[0059] Where represents the modulus of the vector, Refers to the coordinates of the i-th reconstruction.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] (1) A hand-eye calibration method based on a line structured light profiler designed by the present invention has a simple calibration method. By using the unique laser line of the line structured light profiler and a special target for alignment, the calibration time is effectively reduced, the calibration accuracy is high, and it is also helpful for recovery after impact.

[0062] (2) In the present invention, by designing a special target, it can have enough features under a single contour, and the pose angle of the robot with respect to the light plane can be corrected by fine-tuning. Therefore, first, the relative position between the target and the robot is determined by TCP tip calibration, and then through correction, the angle of the pose transformation relationship between the profiler and the target is eliminated, so that there is only a translational relationship between the pose of the profiler coordinate system and the pose of the target coordinate system, and there is no rotational relationship. In this way, the six degrees of freedom of the pose are reduced to three degrees of freedom, which can simplify the operation and calculation method. Description of the Drawings

[0063] Figure 1 Schematic diagram of the position relationship of each coordinate system;

[0064] Figure 2 is the three-dimensional view of the target;

[0065] Figure 3 is the top view of the target;

[0066] Figure 4 is the schematic diagram of the light plane and coordinate system of the line structured light profiler;

[0067] Figure 5 is the schematic diagram of the intersection of the laser line and the target plane;

[0068] Figure 6 is the schematic diagram of the intersection of the laser line and the top of the target cone;

[0069] Figure 7 is the schematic diagram of the sensor profile data after aligning the cone tip. Specific embodiments

[0070] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0071] See Figures 1-3 , a hand-eye calibration method based on a line structured light profiler, the method comprising the following steps:

[0072] Step 1: The TCP tip touches four points method to calibrate the translational relationship between the tool coordinate system O tool at the end of the probe and the coordinate system O end at the end of the robot flange

[0073] Step 2: The TCP tip touches three points method to calibrate the rigid body transformation relationship between the calibration plate coordinate system P calobj and the robot base coordinate system P base

[0074] Step 3: Manually align the calibration plate coordinate system O calobj and the line structured light profiler coordinate system O cam to obtain the rigid body transformation relationship between them

[0075] Step 4: Calculate the rigid body transformation relationship between the line structured light profiler coordinate system O cam and the robot flange end coordinate system O end

[0076] ​​In the present invention, the target has sufficient features under a single contour, and the pose angle of the light plane is corrected by finely adjusting the robot; the robot is installed on the machine table, and the positional relationship of the robot base relative to the world coordinate system remains fixed; the profiler and the glue gun are both installed at the end of the robot manipulator, and their positional relationships with the robot flange remain fixed, and the positional relationship between the profiler and the glue gun remains fixed; the target is placed at a certain position in space, and its positional relationship with the robot base remains fixed.

[0077] The target is fixed at a position in space that can be reached by the robot, so that the calibration plate coordinate system O calobj and the robot base coordinate system O base have a fixed relative positional relationship; the probe is installed at a certain position close to the end of the robot flange, so that the coordinate system O end of the end of the robot flange and the tool coordinate system O tool of the end of the probe have a fixed relative positional relationship.

[0078] The present invention will be further described below with a specific embodiment:

[0079] The target is fixed at a position in space that can be reached by the robot, so that the calibration plate coordinate system O calobj and the robot base coordinate system O base have a fixed relative positional relationship; the probe is installed at a certain position close to the end of the robot flange, so that the coordinate system O end of the end of the robot flange and the tool coordinate system O tool of the end of the probe have a fixed relative positional relationship.

[0080] Table 1 Coordinates of the vertex of the target cone

[0081]

[0082] Step 1:

[0083] Move the robot and select n poses (n≥4) with large joint differences, so that the origin of the tool coordinate system O tool at the end of the probe always coincides with the origin of the calibration plate coordinate system O calobj to obtain n homogeneous transformation matrices of the coordinate system O end of the end of the robot flange relative to the robot base coordinate system O base ;

[0084] The transformation relationship between the tool coordinate system O tool at the end of the probe and the robot base coordinate system O base is: is:

[0085]

[0086] Where Denote the coordinate system O at the end of the robot flange end to the coordinate system O of the robot base base transformation relationship, Denote the tool coordinate system O at the end of the probe tool and the coordinate system O at the end of the robot flange end transformation relationship.

[0087] Written in block form as:

[0088]

[0089] According to matrix multiplication, expanding gives finally:

[0090]

[0091] is the rotation matrix of the coordinate system O at the end of the robot flange end relative to the coordinate system O of the robot base base ;

[0092] Since the origin of the tool coordinate system O at the end of the probe tool coincides with the origin of the calibration plate coordinate system O calobj and the calibration plate coordinate system O calobj has a fixed relative position relationship with the coordinate system O of the robot base base then the translation relationship between the tool coordinate system O at the end of the probe tool and the coordinate system O of the robot base base The value of is fixed and unchanged, then for the robot in n different postures, there is

[0093]

[0094] Written in matrix form as:

[0095]

[0096] When n≥3, the rank of the coefficient matrix is generally equal to 3, and it is a column full-rank matrix. The equation is an inconsistent system of equations, and only the least squares solution can be obtained, that is:

[0097]

[0098] Expanding gives:

[0099]

[0100] where [·] T represents the matrix transpose.

[0101] Step 2:

[0102] Control the robot to make the probe touch m points (m≥3) on the target, and the coordinates of the touched points are known in the calibration board coordinate system O calobj Obtain the m homogeneous transformation matrices of the robot flange end coordinate system O end relative to the robot base coordinate system O base when the probe touches these points respectively

[0103] The coordinates of the points on the target in the robot base coordinate system O base are as follows:

[0104]

[0105] The coordinates of the points on the target in the calibration board coordinate system O calobj are Then, based on the least squares method, the rigid body transformation relationship between the calibration board coordinate system O calobj and the robot base coordinate system O base is obtained The formula is as follows:

[0106]

[0107] Step 3:

[0108] Since only one contour can be obtained in one frame by the line structured light profiler, the rigid body transformation relationship between the calibration board coordinate system O calobj and the line structured light profiler coordinate system O cam cannot be obtained by methods such as photographing a checkerboard calibration board like a area array camera;

[0109] Therefore, the shape of the target can be used to align the calibration board coordinate system O calobj and the line structured light profiler coordinate system O cam to eliminate the rotation relationship of the rigid body transformation between the calibration board coordinate system O calobj and the line structured light profiler coordinate system O cam and obtain the translation relationship based on the contour;

[0110] The rotation of the rigid body transformation can be described by Euler angles and is expressed by the following formula:

[0111] R gba =R x (ψ)R y (θ)R z (φ)

[0112] R x (ψ), R y (θ), R z (φ) represent the 3×3 rotation matrices around the X-axis, Y-axis, and Z-axis respectively;

[0113] The optical plane of the line structured light profiler is the XZ plane, as Figure 4 shown. Since the line structured light profiler cannot rotate at the end of the robot flange, there are 4 cases for the rotation angle according to different installation angles. There are 4 cases for the rotation angle according to different installation angles, which are: φ = 0, φ = 90°, φ = 180°, φ = 270°;

[0114] The following takes φ = 0 as an example to illustrate how to align:

[0115] First, make θ = 0, and then control the robot to make the laser line emitted by the profiler intersect with part of the target plane, as Figure 5 shown. Then keep the profiler stationary, trigger the sensor to acquire an image 40 mm (more or less) long at this position, obtain the depth map and measure the height value (Z value) along the laser line direction, and finely adjust the robot so that the change in the height value at both ends of the laser line ≤ 0.1 mm. This determines the rotation angle of the sensor and the target around the Y axis.

[0116] Then make ψ = 0, φ = 0. Control the robot to make the laser line emitted by the profiler intersect with the top of the target cone, as Figure 6 shown. Then keep the sensor stationary and finely adjust the robot so that the laser line can hit the tops of three collinear cones with different heights at the same time, which means that the rotation angle ψ = 0 around the Z axis and the rotation angle ψ = 0 around the X axis between the profiler and the target.

[0117] After the above adjustments are completed, it is necessary to confirm again whether the above three aspects meet the conditions to avoid affecting the results of the previous adjustment during the adjustment process.

[0118] The above realizes the complete alignment in attitude from the calibration plate coordinate system O calobj to the line structured light profiler coordinate system O cam Therefore, the conversion relationship calobj from the calibration plate coordinate system O cam to the line structured light profiler coordinate system O The rotation matrix in is:

[0119]

[0120] where E3 represents the 3×3 identity matrix;

[0121] Determine the translation relationship That is, determine the translation coordinates of the origin of the calibration plate coordinate system Q calobj in the line structured light profiler coordinate system O cam For the line structured light profiler, the y value of the coordinates of the points falling on its optical plane must be 0 in its coordinate system; therefore, through the contour data collected by the line structured light profiler, obtain:

[0122]

[0123] Therefore, the rigid body transformation relationship can be expressed as:

[0124]

[0125] And obtain the rigid body transformation relationship of the end coordinate system O of the robot flange end relative to the base coordinate system O of the robot base

[0126] Step 4 is specifically as follows:

[0127] Substitute the result of Step 2

[0128]

[0129] and the results obtained in Step 3 and Step 3

[0130]

[0131] into the formula

[0132]

[0133] to obtain the rigid body transformation relationship between the current line structured light profiler coordinate system O cam and the end coordinate system O of the robot flange end

[0134] During operation, the usage method of the hand-eye calibration result is as follows:

[0135] Assume that there are n profilers in the system, and it is necessary to transform all the data collected by the profilers to the unified coordinate system for processing. After performing hand-eye calibration, multiple groups of rigid body transformation relationships are obtained

[0136]

[0137] At time t0, the point set collected by the i-th profiler in the base coordinate system O of the robot base has the following coordinates:

[0138]

[0139] where refers to the rigid body transformation relationship of the end coordinate system O of the robot flange at time t0 relative to the base coordinate system O of the robot end relative to the base coordinate system O of the robot base ; ​​

[0140] Merge all the point sets of the profilometers from time t0 to t n After that, reconstruct the point cloud and perform subsequent operations such as point cloud processing.

[0141] In an embodiment of the present invention,

[0142] Control the robot to make the origin of the calibration board coordinate system O calobj fall on the light plane. Similar to the corresponding operation, obtain the coordinate of the origin of the calibration board coordinate system O Figure 6 at this time in the line structured light profilometer coordinate system O calobj cam below Record the rigid body transformation relationship of the coordinate system O at the end of the robot flange end relative to the robot base coordinate system O base Transform the origin coordinate of the calibration board coordinate system O calobj to the robot base coordinate system O base through the hand-eye relationship matrix, and we have:

[0143]

[0144] where refers to the origin coordinate of the reconstructed calibration board coordinate system O calobj

[0145] Change the pose of the robot multiple times to make the origin of the calibration board coordinate system O calobj always on the light plane. In the ideal case, the coordinate of the origin of the reconstructed calibration board coordinate system O calobj in the robot base coordinate system O base should be the same as that in the result of the TCP tip touching three-point method. Therefore, calculate the calibration error based on this.

[0146] The error calculation formula can be obtained:

[0147]

[0148] where represents the modulus of the vector, refers to the coordinate of the i-th reconstruction.

[0149] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, shall be equally included in the patent protection scope of the present invention.​​​​

Claims

1. A hand-eye calibration method based on line structured light profilometer, characterized in that: The target includes a target that has enough features under a single profile, and the position angle of the light plane is corrected by fine-tuning the robot; the robot is installed on the machine platform, and the position relationship of the robot base relative to the world coordinate system remains unchanged; the profiler and the glue gun are installed at the end of the robot arm, and the position relationship with the robot flange remains unchanged, and the position relationship between the profiler and the glue gun remains unchanged; the target is placed somewhere in space, and the position relationship with the robot base remains unchanged; The target is fixed at a position in space that the robot can reach, so that the calibration plate coordinate system O calobj With the robot base coordinate system O base The relative position relationship is fixed; Install the probe somewhere near the end of the robot flange so that the robot flange end coordinate system O end and the tool coordinate system O at the end of the probe tool The relative position relationship is fixed; the method comprises the following steps: Step 1: TCP tip touch four-point method, calibrate the tool coordinate system O at the end of the probe tool and the robot flange end coordinate system O end Translational relationship Step 2: TCP point touch three-point method, calibrate the calibration plate coordinate system O calobj and the robot base coordinate system O base The rigid body transformation relationship Step 3: Manually align the calibration plate coordinate system O calobj and line structured light profiler coordinate system O cam , get the rigid body transformation relationship between them Step 4: Calculate the line structured light profiler coordinate system O cam and the robot flange end coordinate system O end The rigid body transformation relationship 2. The hand-eye calibration method based on line structured light profilometer according to claim 1, characterized in that: Step 1 is as follows: The mobile robot selects n postures with large joint differences so that the tool coordinate system O at the end of the probe tool The origin is always the same as the calibration plate coordinate system O calobj The origin coincides, and the robot flange end coordinate system O is obtained under these n postures. end Relative to the robot base coordinate system O base n homogeneous transformation matrices of ; Tool coordinate system O at the end of the probe tool and the robot base coordinate system O base The transformation relationship for: in Indicates the robot flange end coordinate system O end To the robot base coordinate system O base The transformation relationship, The tool coordinate system O represents the end of the probe tool and the robot flange end coordinate system O end The transformation relationship of The block format is: Finally we get: is the robot flange end coordinate system O end Relative to the robot base coordinate system O base The rotation matrix of the tool coordinate system O at the end of the probe tool Origin and calibration plate coordinate system O calobj The origin coincides, and the calibration plate coordinate system O calobj With the robot base coordinate system O base The relative position relationship is fixed, then the tool coordinate system O at the end of the probe tool With the robot base coordinate system O base Translational relationship The value of is fixed, then for the robot in n different postures, there is Further we get: in[·] T Represents matrix transpose.

3. The hand-eye calibration method based on line structured light profilometer according to claim 1, characterized in that: Step 2 is as follows: Control the robot to make the probe touch m points on the target, and the touching points are in the calibration plate coordinate system O calobj The coordinates of the robot flange end are known, and the coordinate system O of the robot flange end when the probe touches these points is obtained. end Relative to the robot base coordinate system O base The m homogeneous transformation matrices of The point on the target is in the robot base coordinate system O base The coordinates below are: … The point on the target is in the calibration plate coordinate system O calobj The coordinates below are Based on the least squares method, the calibration plate coordinate system O is obtained calobj and the robot base coordinate system O base The rigid body transformation relationship The formula is as follows:

4. The hand-eye calibration method based on line structured light profilometer according to claim 2, characterized in that: Step 3 is as follows: The target shape can be used to convert the calibration plate coordinate system O calobj and line structured light profiler coordinate system O cam Align and eliminate the calibration plate coordinate system O calobj and line structured light profiler coordinate system O cam The rotation relationship of the rigid body transformation between them is obtained, and the translation relationship is obtained based on the contour; The rotation of a rigid body transformation can be described using Euler angles, expressed as follows: R gba =R x (ψ)R y (i)R z (f) R x (ψ), R y (θ), R z(φ) represents the 3×3 rotation matrix around the X-axis, Y-axis, and Z-axis respectively; According to the different installation angles of the camera relative to the robot flange end, there are four rotation angles: φ = 0°, φ = 90°, φ = 180°, and φ = 270°; Therefore, the calibration plate coordinate system O calobj To the line structured light profiler coordinate system O cam The conversion relationship The rotation matrix in is: φ=0,90°,180°,270° Where, E3 represents the 3×3 identity matrix; Determine the translation relationship That is to determine the calibration plate coordinate system O calobj Origin online structured light profiler coordinate system O cam For the line structured light profiler, the coordinates of the points falling on its light plane must have a y value of 0 in its coordinate system; therefore, the profile data collected by the line structured light profiler is obtained: Therefore, the rigid body transformation relationship It can be expressed as: And obtain the robot flange end coordinate system O during alignment end Relative to the robot base coordinate system O base The rigid body transformation relationship 5. The hand-eye calibration method based on line structured light profilometer according to claim 1, characterized in that: Step 4 is as follows: Rigid body transformation relationship The expression is as follows:

6. The hand-eye calibration method based on line structured light profilometer according to claim 1, characterized in that: After step 4, multiple sets of rigid body transformation relationships are obtained At time t0, the point set collected by the i-th profiler is In the robot base coordinate system O base The coordinates below are: in Refers to the robot flange end coordinate system O at time t0 end Relative to the robot base coordinate system O base The rigid body transformation relationship; Translate the time from t0 to t n After merging the point sets of all the profilometers, the point cloud is reconstructed and point cloud processing is performed.

7. The hand-eye calibration method based on line structured light profilometer according to claim 1, characterized in that: It also includes a hand-eye calibration accuracy test, and the test method is as follows: Control the robot to make the calibration plate coordinate system O calobj The origin falls on the light plane, and the coordinate system O of the calibration plate is obtained at this time calobj Origin online structured light profiler coordinate system O cam The coordinates below Record the robot flange end coordinate system O at this time end Relative to the robot base coordinate system O base The rigid body transformation relationship The hand-eye relationship matrix obtained in step 4 The calibration plate coordinate system O calobj The origin coordinates are transformed to the robot base coordinate system O base Next, we have: in Refers to the reconstructed calibration plate coordinate system O calobj Origin coordinates. Change the robot's position several times to make the calibration plate coordinate system O calobj The origin is always on the light plane. Ideally, the reconstructed calibration plate coordinate system O calobj The origin is in the robot base coordinate system O base The coordinates below should match those in the TCP cusp touch three-point method result. The calibration error is calculated based on this. The error calculation formula is: where it represents the magnitude of the vector, Refers to the coordinates of the i-th reconstruction.

Images