A hand-eye calibration method based on rotation scanning of a line laser three-dimensional profile scanner

Abstract

The application provides a hand-eye calibration method based on rotary scanning of a line laser three-dimensional profile scanner, a customized calibration piece with obvious height features is used to calculate the depth value of an image unwrapping diagram collected by the line laser three-dimensional profile scanner corresponding to the center of a rotating disc, then the length of the unwrapping diagram is converted into an angle value of 0-360 degrees by using a normalization principle, the polar radius is calculated by the difference between the image depth value and the center depth value of the rotating disc, the two-dimensional unwrapping diagram is converted into a three-dimensional space coordinate so as to use the N-point calibration method to calibrate the hand-eye calibration of a mechanical arm and the line laser profile scanner, the principle of the application is simple, the converted result can be directly calibrated by using the N-point calibration method, and the accuracy of the calibration result is improved.

Description

Technical Field

[0001] This invention relates to the field of robot hand-eye calibration technology, and in particular to a hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner. Background Technology

[0002] Hand-eye calibration is a commonly used technique in robotics and computer vision. It primarily addresses the coordinate transformation between a robot's end effector (hand) and a camera (eye), i.e., the relative pose and position of the hand and eye. In many robotic applications, robots need to perform vision-related tasks such as picking, assembly, and navigation. Hand-eye calibration ensures accurate coordinate transformations between the robot's end effector and the camera, enabling coordinated operation of the robot's hand and vision system. In some applications, multiple cameras or other sensors may be used to acquire more information. Hand-eye calibration can be used to fuse coordinate transformations between multiple sensors, allowing components within the system to work collaboratively.

[0003] Robot vision systems are categorized into two types: eye-in-hand and eye-to-hand. The calibration method for eye-to-hand systems typically employs the "N-point calibration method," which involves finding N (N>=3) identical points in two 3D coordinate systems, calculating the transformation matrix between them. Nine points are commonly used in projects, hence the term "nine-point calibration." However, directly using the "N-point calibration method" for eye-to-hand calibration has limitations. For example, when the robot's "eye" is a line-laser contour scanner or similar line-scanning camera, and the task involves scanning the unfolded image of an object on a turntable, the traditional "N-point calibration method" is not directly applicable. This is because the unfolded image essentially converts a 3D object into 2D planar information. Directly using the "N-point calibration method" in this case will result in an inaccurate and significantly biased transformation matrix. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides a hand-eye calibration method based on rotational scanning of a line laser three-dimensional contour scanner, which converts the planar two-dimensional image coordinate system into a three-dimensional spatial coordinate system. The converted three-dimensional spatial coordinate system can then be used normally for hand-eye calibration using the "N-point calibration method".

[0005] To achieve the above-mentioned technical objectives, the technical solution adopted by this invention is as follows:

[0006] A hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner includes:

[0007] 1) The calibration component is coaxially mounted on a custom turntable, with its axis centered on the turntable, ensuring that the image depth values ​​at the center of the turntable are the same. The calibration component has distinct height characteristics and multiple feature points, which can be used for the converted "N-point calibration method." The calibration component has a radial dimensional difference along its axis to facilitate the calculation of the depth value represented by a unit length.

[0008] 2) Start the turntable to rotate the calibration part one revolution, scan the side of the calibration part with a line laser 3D contour scanner, and unfold the obtained side scan image of the calibration part in a planar form;

[0009] 3) Establish a two-dimensional Cartesian coordinate system on the side unfolded view of the calibration part. Under the same vertical coordinate, select two points with different depth values ​​on the unfolded view and define them as a and b. Calculate the depth difference between the two points a and b. Map the actual distance between the two points a and b on the top view of the calibration part to the calculated depth difference between the two points a and b to obtain the depth value Lu represented by the unit length.

[0010] 4) Select any one of points a and b, obtain the depth value Ls of that point, and measure the actual distance L from that point to the axis of the calibration component. Using the depth value Lu represented by unit length, the image depth value Ls of the selected point, and the actual distance L from the selected point to the axis of the calibration component, calculate the depth value Lcenter of the turntable center on the image. The expression for Lcenter is as follows:

[0011] L center =L×L u +Ls;

[0012] 6) Using the normalization principle, convert the length value of the side unfolded view of the calibration part into an angle value α of 0°-360°. Select a feature point on the calibration part and read the depth value Lsurface of that point and the depth value Lcenter of the turntable to obtain the polar radius r. The expression for r is as follows:

[0013] r = |L surface -L center |;

[0014] 6) Establish a polar coordinate system with the center of the turntable as the origin and the line connecting the starting position of the line laser 3D contour scanner and the origin as the polar axis. Use the polar coordinate transformation formula to convert the obtained angle value α and polar radius r into the X-axis coordinates and Y-axis coordinates in the three-dimensional Cartesian coordinate system. Use the vertical axis of the side unfolded view of the calibration part as the Z-axis of the three-dimensional Cartesian coordinate system to convert the coordinate points into spatial coordinate points.

[0015] 7) Calculate the position of the required feature points in space through steps 5) to 6), and finally use the "N-point calibration method" to complete the hand-eye calibration.

[0016] As one possible implementation, the calibration element is further formed by coaxially connecting a regular prism and a straight cylinder.

[0017] As one possible implementation, the customized turntable is further provided with a slot that matches the shape of the bottom of the calibration part, for positioning the calibration part so that the axis of the calibration part is located at the center of the turntable.

[0018] As a preferred implementation method, the feature points are preferably selected from the corner points of a regular prism.

[0019] As a possible implementation, further, in step 5), the length value of the side unfolded view of the calibration part (3) is converted into an angle value α of 0°-360° using the normalization principle; wherein, the expression for the angle value α is as follows:

[0020]

[0021] Where x is the abscissa of the selected point in the two-dimensional Cartesian coordinate system established in step 3); x min The x-coordinate of the starting point of the unfolded side view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3); max The x-coordinate of the end point of the side unfolded view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3).

[0022] As a possible implementation, further, in step 6), the obtained angle value α and polar radius r are converted into X-axis and Y-axis coordinates in a three-dimensional Cartesian coordinate system using a polar coordinate transformation formula; wherein, the polar coordinate transformation formula is:

[0023]

[0024] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0025] 1) The calibration method mentioned in this invention is applicable to hand-eye calibration (eye outside the hand) of all line laser depth scanners in the case of object unfolding on the scanning turntable.

[0026] 2) The customized calibration model required by this invention is simple, easy to process, and has obvious features, which can quickly and accurately extract the feature points required for hand-eye calibration.

[0027] 3) This invention utilizes the characteristics of planar images to establish a polar coordinate system in space, and then transforms the polar coordinate system into a spatial Cartesian coordinate system. The principle is simple, and after the transformation, it can be directly calibrated using the "N-point calibration method", which improves the accuracy of the calibration results. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a schematic diagram of the hardware structure implemented in this invention;

[0030] Figure 2 This is a side view of the calibration component.

[0031] Figure 3 A schematic diagram of the customized turntable planar structure;

[0032] Figure 4 This is a schematic diagram of the calibration component structure;

[0033] Figure 5 A schematic diagram for establishing a two-dimensional Cartesian coordinate system on the unfolded side view of the calibration part;

[0034] Figure 6 This is a schematic diagram of points a and b on the top view;

[0035] Figure 7 A schematic diagram of the polar axis of the established polar coordinate system (the dashed line T in the diagram represents the polar axis).

[0036] Line laser 3D contour scanner-1; Custom turntable-2; Calibration part-3; Turntable-4; Six-axis robotic arm-5. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the invention. Similarly, the following embodiments are only some, not all, embodiments of the present invention, and all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] See attached document Figure 1-4 As shown, this embodiment provides a hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner, including:

[0039] (1) Install the custom turntable 2 onto the turntable 4, and embed the calibration component 3 into the groove of the custom turntable 2, ensuring that the calibration component 3 is positioned at the center of the turntable 4. In this embodiment, the calibration component 3 is composed of a cube and a cylinder (as shown in the attached figure). Figure 4As shown), the corner points on the cube can be used as feature points in the "N-point calibration method"; the custom turntable 2 is provided with a slot that matches the bottom of the calibration part 3 (as shown in the attached diagram). Figure 3 As shown), it is used to position the calibration component 3 so that the axis of the calibration component 3 is located at the center of the turntable 4.

[0040] (2) Start the turntable 4 to rotate the calibration part 3 one revolution, scan the side of the calibration part 3 using the line laser 3D contour scanner 1, and unfold the obtained side scan image of the calibration part 3 in a planar form (as shown in the attached image). Figure 2 (As shown).

[0041] (3) Establish a two-dimensional Cartesian coordinate system on the unfolded side view of calibration part 3. In this embodiment, the lower left corner of the unfolded side view of calibration part 3 is used as the origin to establish a two-dimensional Cartesian coordinate system. Then, under the same vertical coordinate, two points with different depth values ​​on the unfolded view are selected and defined as a and b (as shown in the attached figure). Figure 5 (as shown); where the depth value can be approximated as the distance value obtained by the line laser 3D contour scanner 1, and the depth difference between points a and b is calculated. The actual distance between points a and b on the top view calibration piece is mapped to the depth value Lu represented by a unit length by the calculated depth difference between points a and b (as shown in the attached figure). Figure 6 (As shown). The expression for the depth value Lu represented by a unit length is as follows:

[0042]

[0043] Among them, L cy L represents the depth value at point a on the surface of the cylinder. cu L represents the depth value at point b on the surface of the cube. A Top view (with appendix) Figure 6 The actual distance between points a and b in ().

[0044] (4) Select any one of points a and b, obtain the depth value Ls of that point, and measure the actual distance L from that point to the axis of calibration component 3. Calculate the depth value Lcenter of the turntable center in the image using the depth value Lu represented by the unit length, the image depth value Ls of the selected point, and the actual distance L from the selected point to the axis of calibration component 3. In this embodiment, point a is selected to calculate the depth value Lcenter of the turntable center in the image. The actual distance from point a to the axis of calibration component 3 is defined as L (as shown in the attached diagram). Figure 6 As shown in the figure, the depth value of point a is defined as Ls. The expression for Lcenter is as follows:

[0045] L center =L×L u +Ls;

[0046] (5) Using the normalization principle, the length value of the side unfolded view of the calibration part is converted into an angle value α of 0°-360°, where the expression for the angle value α is as follows:

[0047]

[0048] Where x is the abscissa of the selected point in the two-dimensional Cartesian coordinate system established in step (3); x min The x-coordinate of the starting point of the unfolded side view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3); max The x-coordinate of the end point of the side unfolded view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3).

[0049] Next, select the feature point on calibration component 3, i.e., the corner point of the cube, and read the depth value Lsurface of this point and the depth value Lcenter of the turntable to obtain the polar radius r. The expression for r is as follows:

[0050] r = |L surface -L center |;

[0051] (6) Establish a polar coordinate system with the center of turntable 4 as the origin and the line connecting the starting position of the line laser 3D contour scanner and the origin as the polar axis (as shown in the attached diagram). Figure 7 As shown), the obtained angle value α and polar radius r are converted into X-axis and Y-axis coordinates in a three-dimensional Cartesian coordinate system using the polar coordinate transformation formula; where the polar coordinate transformation formula is:

[0052]

[0053] Using the vertical axis of the unfolded side view of the calibration part as the Z-axis of the three-dimensional Cartesian coordinate system, the coordinate points are transformed into spatial coordinate points (X, Y, Z).

[0054] (7) Calculate the position of the required feature points in space through steps (5) to (6), and finally use the "N-point calibration method" to complete the hand-eye calibration.

[0055] The above description is only a part of the embodiments of the present invention and does not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made based on the content of the present invention specification and drawings, or direct or indirect application in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner, characterized in that, include: 1) The calibration part (3) is coaxially mounted on the turntable (4) by a custom turntable (2) so that the axis of the calibration part is at the center of the turntable; the calibration part has obvious height characteristics and multiple feature points, and the calibration part has a radial dimension difference along the axis direction; 2) Start the turntable (4) to rotate the calibration part (3) one revolution, scan the side of the calibration part (3) by the line laser three-dimensional contour scanner (1), and unfold the obtained side scan image of the calibration part (3) in a planar form; 3) Establish a two-dimensional Cartesian coordinate system on the side unfolded view of the calibration part (3). Under the same vertical coordinate, select two points with different depth values ​​on the unfolded view and define them as a and b. Calculate the depth difference between the two points a and b. Map the depth value Lu represented by the unit length to the actual distance between the two points a and b on the top view and the depth difference between the two points a and b. 4) Select any one of points a and b, obtain the depth value Ls of that point, and measure the actual distance L from that point to the axis of the calibration component. Using the depth value Lu represented by unit length, the image depth value Ls of the selected point, and the actual distance L from the selected point to the axis of the calibration component, calculate the depth value Lcenter of the turntable center on the image. The expression for Lcenter is as follows: L center =L×L u +Ls; 5) Using the normalization principle, the length value of the side unfolded view of the calibration part (3) is converted into an angle value α of 0°-360°. A feature point on the calibration part (3) is selected, and the depth value Lsurface of the point is read. The difference between the depth value Lcenter of the turntable is used to obtain the polar radius r, where the expression for r is as follows: r=|L surface -L center |; 6) Using the center of the turntable as the origin, the line connecting the starting position of the line laser 3D contour scanner and the origin as the polar axis to establish a polar coordinate system. The obtained angle value α and polar radius r are converted into X-axis coordinates and Y-axis coordinates in the three-dimensional Cartesian coordinate system using the polar coordinate transformation formula. The vertical axis of the side unfolded view of the calibration part (3) is used as the Z-axis of the three-dimensional Cartesian coordinate system to convert the coordinate points into spatial coordinate points. 7) Calculate the position of the required feature points in space through steps 5) to 6), and finally use the "N-point calibration method" to complete the hand-eye calibration.

2. The hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner according to claim 1, characterized in that, The calibration component (3) is formed by coaxially connecting a regular prism and a straight cylinder.

3. The hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner according to claim 1, characterized in that, The custom turntable (2) is provided with a slot that matches the bottom shape of the calibration part (3) for positioning the calibration part (3) so that the axis of the calibration part (3) is located at the center of the turntable (4).

4. The hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner according to claim 2, characterized in that, The feature points are selected from the corner points of a regular prism.

5. The hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner according to claim 1, characterized in that, In step 5), the length value of the side unfolded view of the calibration part (3) is converted into an angle value α of 0°-360° using the normalization principle; where the expression for the angle value α is as follows: Where x is the abscissa of the selected point in the two-dimensional Cartesian coordinate system established in step 3); x min The x-coordinate of the starting point of the unfolded side view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3); max The x-coordinate of the end point of the side unfolded view of the calibration component in the two-dimensional Cartesian coordinate system established in step 3).

6. The hand-eye calibration method based on rotational scanning of a line laser 3D contour scanner according to claim 1, characterized in that, In step 6), the obtained angle value α and polar radius r are converted into X-axis and Y-axis coordinates in a three-dimensional Cartesian coordinate system using the polar coordinate transformation formula; the polar coordinate transformation formula is:

Images