CN103398660B - For obtaining the structured light vision sensor parameter calibration method of weld bead height information - Google Patents

Abstract

The method for calibrating the parameters of the structured light vision sensor used to obtain the height information of the welding seam in the present invention relates to the measurement of non-specific variables, and is a method for calibrating the parameters of the structured light vision system based on wedge-shaped targets. The offset of the structured light can be obtained by processing the three-dimensional image, and the height information of the weld can be calculated, and then the three-dimensional coordinates of the intersection point of the laser belt and the weld on the target can be calculated. The specific steps include: the first step, the horizontal deviation is calculated Identification; the second step is to identify the vertical deviation; the third step is to use the camera to calibrate the parameters of the structured light vision system; the fourth step is to obtain the weld height information. The calibration process of the method of the invention does not involve large-scale matrix transformation, the amount of calculation is small, there is no problem of occlusion of the target, and there is no problem of non-convergence to the global optimal solution that occurs in solving constrained nonlinear equations.

Description

Calibration method of structured light vision sensor parameters for obtaining weld height information

technical field

The technical solution of the present invention relates to the measurement of non-specific variables, in particular, a method for calibrating parameters of a structured light vision sensor for obtaining weld height information.

Background technique

The calibration of the vision system is an important issue in robot vision. The image collected by the vision system, the one-to-one correspondence between the image coordinate position of each point and a certain point on the surface of the space object needs to be determined by the imaging model of the vision system, and the imaging model Each parameter of , needs to be determined through the calibration of the vision system, so the calibration of the vision system is the key to realize the conversion from 2D image information to 3D environment information. In the calibration process of the vision system, the acquisition of three-dimensional height information has always been a difficult problem.

The structured light visual inspection method has the advantages of large range, non-contact, large field of view and good system flexibility. It has been widely used in industrial environments, especially in the field of welding in recent years. The calibration of the parameters of the structured light vision sensor mainly includes the calibration of the internal parameters of the camera and the calibration of the structural parameters of the line structured light vision sensor. The internal parameters of the camera mainly refer to the geometric and optical characteristics inside the camera, such as optical center, focal length, position, etc. The structural parameters of the sensor refer to the transformation relationship between the image coordinate system and the three-dimensional camera coordinate system. The purpose of sensor parameter calibration Reconstruct the 3D world coordinates from the 2D image coordinates to realize 3D measurement.

In the field of welding technology, the traditional structured light vision sensor parameter calibration method is to project point structured light or linear structured light onto the surface of the workpiece, and calculate the weld height information by triangular measurement method (see "Wu Lin, Chen Shanben, Intelligent Welding Technology, pp. 210-211, pp. 215-217, Beijing: National Defense Industry Press, 1999"; see "Xu De, Tan Min, Li Yuan, Robot Vision Measurement and Control, p. 133, Beijing: National Defense Industry Press, 2008"). The calibration calculation process is based on the fact that the intersection point of the laser and the horizontal plane just passes through the optical center of the lens. However, in actual operation, it is difficult for us to determine the position of the optical center axis of the camera, especially when it is used on site. It is difficult to accurately calibrate. Therefore, the above-mentioned triangle method needs to be further improved and perfected according to the actual situation.

CN03142658.1 discloses "a structured light vision sensor calibration method based on a plane target", which mainly uses a freely movable two-dimensional plane target distributed with black squares to calibrate the structured light vision sensor, and the calibration process In the method, the vertices of the black squares are used as the feature points, and the calibration of the internal parameters of the camera is completed by using the transformation between the image coordinate system of the feature points and the world coordinate system. This method involves large-scale matrix transformation, and the calculation and derivation are complicated. The precision two-dimensional planar target is difficult to process, and there is a problem of mutual occlusion.

CN200710121397.X discloses "a structural parameter calibration method of a structured light vision sensor", which mainly uses a plane target with multiple nonlinear feature points, and obtains the position of the plane target after each movement by moving the position of the plane target multiple times. The coordinates of four or more nonlinear feature points on the target image are transformed into the camera coordinate system to fit the high-order ternary structured light equation. The calibration process of this method is cumbersome and will affect the accuracy, and the coordinate transformation and fitting process is complicated.

CN200910089307.2 reported "Structured light vision sensor calibration method based on one-dimensional target", this method uses at least three feature points of one-dimensional target with known spatial constraints, combined with perspective projection equation, according to the length of feature points Constraints and direction constraints Calculate the camera system coordinates of feature points and perform fitting to obtain the light plane equation. This method needs to solve nonlinear equations with constraints, and introduces nonlinear optimization. The method is complicated, slow, sensitive to initial value selection and noise, and cannot guarantee that the parameters converge to the global optimum.

Contents of the invention

The technical problem to be solved by the present invention is to provide a structured light vision sensor parameter calibration method for obtaining weld height information, which is a structured light vision system parameter calibration method based on a wedge-shaped target. Carry out precise measurement in advance, so that the three-dimensional coordinates of the intersection point of the laser belt and the weld seam on the target can be calculated. During the calibration process, the parameters of the vision system can still be accurately and quickly calibrated when the position of the optical center axis of the camera is uncertain, and the calibration The process does not involve large-scale matrix transformation, the amount of calculation is small, there is no problem of occlusion of the target, and there is no problem of non-convergence to the global optimal solution that occurs in solving constrained nonlinear equations.

The technical solution adopted by the present invention to solve the technical problem is: the structured light vision sensor parameter calibration method for obtaining weld height information is a wedge-shaped target-based structured light vision system parameter calibration method, the steps are:

The wedge-shaped target adopts a triangular wedge, and the main equipment used includes a camera, a laser, an encoder in the motion mechanism, a triangular wedge with an acute angle of 20° to 60°, weldments, an electronic wedge vernier and a servo. The motor, the camera mainly involves the camera lens;

The first step is to identify the horizontal deviation

By controlling the servo motor, the camera is moved along the parallel direction of the structured light belt for a distance of 5mm to 25mm and measured and recorded, and then the pixel value of the welding seam moving in the up and down direction in the two-dimensional image is calculated, and then the actual value represented by each pixel is calculated. The distance can be used to obtain the proportional coefficient K between the pixel value and the actual distance value;

The second step is to identify the vertical deviation

Step 2-I, place the triangular wedge under the camera to ensure that the laser belt hits the edge of the triangular wedge;

Step 2-II, move the triangular wedge forward for a distance of y _x1 , the structured light strip produced by the laser will move upward along the hypotenuse of the triangular wedge until the structured light strip rises to The height above the horizontal plane AC=10mm stops, the moving distance of the structured light strip on the image plane is △y ₁ , the encoder in the motion mechanism will record the forward moving distance y _x1 of the triangular wedge during the moving process, in two dimensions Process and record the distance Δy ₁ of the structured light band moving on the image plane in the image;

Step 2-Ⅲ, repeat the operation process of step 2-Ⅱ, but continue to move the triangular wedge block forward for a distance of y _x2 until the structured light strip rises to the height A'C'=20mm above the horizontal plane and stops. At this time, the moving distance of the structured light strip on the image plane is △y ₂ , and the encoder in the motion mechanism will record the distance y _x2 of the forward movement of the triangular wedge during the moving process, and process and record the structured light strip in the two-dimensional image The moving distance △y ₂ on the image plane;

The third step is to use the camera to calibrate the parameters of the structured light vision system

Mark the position of the horizontal plane 0 mm where the acute angle apex of the triangular wedge in the position of the triangular wedge in the operation of step 2-I is D, and place it in the position of the triangular wedge in the operation of step 2-II The position of the acute-angle apex of the triangular wedge in the horizontal plane 0 mm is marked as D1, and the position of the acute-angle apex of the triangular wedge in the position of the triangular wedge in the operation process of step 2-Ⅲ is marked as D2, Then the distance between D and D1 is y _x1 , the distance between D1 and D2 is y _x2 ,

Let the intersection point of the line laser emitted by the laser and the hypotenuse of the triangular wedge in the position of the triangular wedge in step 2-Ⅱ be A, the center of the camera lens be marked as O, and the intersection of the extension line of the camera lens center O and A The horizontal plane with a height of 0mm is at point B, draw a vertical line from A to the horizontal plane with a height of 0mm, and intersect the horizontal plane at point C, then the three points form △ABC; draw a vertical line from the center point O of the camera lens to the horizontal plane with a height of 0mm, and intersect the horizontal plane at Point O', then three points O, B, O' form a triangle △OBO', record the distance between AC as h ₁ ,

According to △ABC∽△OBO’similarity:

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; BC</span>}}{{\mathrm{<span\; class="notranslate">\; BO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Let the intersection point of the line laser emitted by the laser and the hypotenuse of the triangular wedge at the position of the triangular wedge in the operation process of step 2-Ⅲ be A′, and the horizontal plane with a height of 0mm intersected by the extension line of the camera lens center O and A′ is at Point B', make a vertical line from A' to the horizontal plane with a height of 0mm, and intersect the horizontal plane at point C', then the three points form △A'B'C'; make a vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0mm Intersect the horizontal plane with the point O', then the three points O, B', O' form a triangle △OB'O'. Record the distance between A'C' as h ₂ .

According to the similarity of △A'B'C'∽△OB'O':

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; C</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; o</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Denote the angle between the central axis of the laser and the vertical direction as θ, record BD=y ₁ , B′B=y ₂ ,

Then BC=y ₁ -(y _x1 -h ₁ /tanθ), B′C′=y ₁ +y ₂ -(y _x1 +y _x2 -h ₁ /tanα), record the distance OO from the lens center O to the height 0mm ' is H, the distance from the acute angle vertex D of the triangular wedge block to the camera optical central axis OO' in step 2-I is Y, that is, OO'=H, DO'=Y, and BO'=Y+y ₁ , B 'O'=Y+y ₁ +y ₂ , put in (1) and (2) to get:

$\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, h ₁ =10mm, h ₂ =20mm, α=20°～60°, taking the known quantity into it, we can get:

The values of parameters H and Y can be obtained from formula (4). In addition, the geometric relationship in the figure can be obtained:

y _x1 -AC/tanα=AC*tanθ

Right now:

$\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; AC</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; AC</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

It is known that h ₁ =AC=10mm,α=20°～60°substitute into (5),

From the above formula, tanθ can be calculated.

So far, the camera is used to calibrate the parameters of the structured light vision system, and the values of the parameters K, H, Y, and tanθ are respectively obtained;

The fourth step, the acquisition of weld height information

In the case of seam tracking, what is to be measured is the seam height information at the intersection of the seam and the structured light belt, and the weldment to be measured for seam height information is moved until the weld point where the seam height is to be measured is in contact with the The structured light belts intersect, record the point of intersection as M, make the extension line of OM intersect the 0mm plane at point N, record the distance between point N and the acute angle vertex D of the triangular wedge in step 2-I as y ₃ =ND, it can be seen that y ₃ is The horizontal distance between the intersection point M of the structured light strip and the welding point of the weld to be measured for the height of the weld and the apex D of the acute angle of the triangular wedge in step 2-I is the perpendicular line from point M to the 0mm plane, and the perpendicular point is P, record The distance between point M and point P is h ₃ =MP,

According to △MNP∽△ONO’similarity:

$\frac{\mathrm{<span\; class="notranslate">\; MP</span>}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; NP</span>}}{{\mathrm{<span\; class="notranslate">\; NO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$

That is:

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Right now: $h_3=\frac{{\mathrm{the\; y}}_3*h}{{\mathrm{the\; y}}_3+Y+\mathrm{the\; tan}\mathrm{\&theta;}*h}=f\left({\mathrm{the\; y}}_3\right)---\left(7\right)$

It can be seen from the above formula that h ₃ is a function of y ₃ ,

Move the weld along the direction of the vertical structured light strip until the structured light strip intersects with the welding point where the height of the weld is to be measured, and record the distance between the intersection point of the weld seam and the structured light strip and the optical central axis OO' of the camera lens through a two-dimensional image The distance between △y3', then △y3=△y3'-Y/K, that is, the size of the offset △y3. Then the size of y3 is y3=K*△y3, and the horizontal height of the initial welding point is set to 0mm. When there is a vertical deviation between the camera and the weld, the laser stripes in the image will be in the left and right directions. At this time, the weld seam does not move. When such a change occurs in the image, the displacement offset △y3 of the intersection point of the weld seam and the structured light belt is recorded by image processing, and then it is known that y3=K*△y3, combined with (7 ) formula can identify the deviation between the welding torch and the actual welding seam in the vertical direction. So far, the welding seam height information at the intersection of the welding seam and the structured light belt is obtained, and the reconstruction of the three-dimensional camera coordinates is realized.

In the method for calibrating the parameters of the structured light vision sensor for obtaining the weld seam height information, the laser used is a line laser.

In the method for calibrating parameters of the structured light vision sensor for obtaining weld height information, one of the acute angles of the triangular wedge is 30°.

In the method for calibrating the parameters of the structured light vision sensor for obtaining weld height information, the equipment involved is all commercially available.

The beneficial effects of the present invention are:

The outstanding substantive features of the method of the present invention are: the method of the present invention is a method for calibrating the parameters of the structured light vision system based on a wedge-shaped target, using the parameters obtained from the calibration to process the two-dimensional image to obtain the offset of the structured light, namely The height information of the weld can be calculated, and then the three-dimensional coordinates of the intersection point of the laser band and the weld on the target can be calculated.

The remarkable progress of the inventive method is:

(1) Monocular vision has certain difficulties in obtaining weld height information. When binocular vision obtains weld height information, it needs to fuse the information obtained by two cameras, which has a large amount of calculation and affects the real-time performance of the system. The present invention The proposed algorithm improves this situation. It not only avoids the cumbersome matrix transformation when obtaining the weld height information, but also avoids the inaccuracy of determining the position of the camera's optical center axis, greatly saving calibration and calculation time, and significantly improving the real-time performance of the system.

(2) In the process of calibrating the parameters of the structured light vision sensor, the parameters of the vision system can still be accurately and quickly calibrated when the position of the optical central axis of the camera is uncertain, and the calibration process does not involve large-scale matrix transformation, the calculation amount is small, and it is not necessary to There is the problem of occlusion of the target, and there is no problem of non-convergence to the global optimal solution that occurs in solving constrained nonlinear equations.

(3) On the one hand, the method of the present invention avoids the establishment of complex camera imaging models, and on the other hand, it enhances the accuracy, robustness and adaptability of camera calibration.

Description of drawings

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

Figure 1(a) is a general schematic diagram of the position change of the triangular wedge during the calibration process of the method of the present invention.

Fig. 1(b) is a schematic diagram of the position of the triangular wedge in identifying the deviation in the vertical direction in step 2-I of calibration in the method of the present invention.

Fig. 1(c) is a schematic diagram of identifying the position of the triangular wedge for the deviation in the vertical direction in step 2-II of the method of the present invention.

Fig. 1(d) is a schematic diagram of the position of the triangular wedge in the process of repeating the 2-II step in the process of calibrating the 2-III step to identify the vertical deviation in the method of the present invention.

Fig. 1(e) is a schematic diagram of obtaining weld height information of a weldment by using the calibration result in the method of the present invention.

Fig. 2(a) is a schematic diagram of the relationship between the deviation between the camera and the weld seam in the horizontal direction and the change of the image in the method of the present invention.

Fig. 2(b) is a schematic diagram of the deviation between the camera and the welding seam in the vertical direction and the change of the image in the method of the present invention.

Fig. 3 is a curve diagram of the test result data fitting result of the method of the present invention.

In the figure, 1. Camera lens, 2. Laser, 3-1. The position of the triangular wedge in the second step of 2-I step operation in the second step of identifying the vertical deviation, 3-2. The second step of the vertical deviation The position of the triangular wedge during step 2-Ⅱ in the identification of the directional deviation, 3-3. The position of the triangular wedge during the 2-Ⅲ step in the second step of identifying the vertical deviation, 4 .Weldment, 5. Camera, 6. Laser stripe, 7. Weld seam.

Detailed ways

The embodiment shown in Fig. 1(a) shows the change of the position of the triangular wedge during the calibration process of the method of the present invention. The figure shows camera lens 1, laser 2, the position of the triangular wedge in the second step 2-I step 3-1 of identifying the vertical deviation, and the triangular wedge in the second step of identifying the vertical deviation The position 3-2 of the triangular wedge in the third step to identify the vertical deviation; the acute angle of the triangular wedge is α, and the second step is to identify the vertical deviation in the 2-I step operation The position of the triangular wedge in the process 3-1 The position where the acute angle apex of the triangular wedge is at 0mm on the horizontal plane is marked as D, the distance from point D to the optical central axis OO' of the camera lens 1 is marked as Y, and 2-II The position of the acute angle vertex of the triangular wedge in the position 3-2 of the triangular wedge in the step operation process is recorded as D1, and the position of the triangular wedge in the position 3-2 of the triangular wedge in the 2-Ⅲ step operation process is recorded as D1. The position where the apex of the acute angle is at 0mm on the horizontal plane is recorded as D2. When the distance of the triangular wedge moving forward is y _x1 , the distance △y ₁ that the structured light moves on the image plane is △y 1 . When the distance of the triangular wedge moving forward is y _x2 , the structure The distance △y ₂ that the light moves on the image plane, the hypotenuse of the triangular wedge at the position of the triangular wedge during the operation of step 3-2 in the line laser and the second step of identifying the deviation in the vertical direction The intersection point is A, the extension line of the line OA connecting the center O of the camera lens 1 and A intersects the horizontal plane with a height of 0 mm at point B, and the perpendicular line from A to the horizontal plane with a height of 0 mm intersects the horizontal plane at point C, then the three points form △ABC; Make the vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0mm intersect the horizontal plane at point O', then the three points O, B, O' form a triangle △OBO', and the line laser and the second step identify the vertical deviation The intersection point of the hypotenuse of the triangular wedge in the position of the triangle wedge in the 2-Ⅲ step 3-3 is A', and the horizontal plane with a height of 0mm intersected by the extension line of the center O of the camera lens 1 and A' is at point B ’, draw a vertical line from A’ to the horizontal plane with a height of 0mm, and intersect the horizontal plane at point C’, then the three points form △A’B’C’, BD=y ₁ , B’B=y ₂ .

For clarity of illustration, this figure is decomposed into the following Fig. 1(b), Fig. 1(c) and Fig. 1(d).

The embodiment shown in Fig. 1(b) shows that in the 2-I step operation process of identifying the deviation in the vertical direction, the triangular wedge block is first placed under the camera lens 1, which is the second step in the identification of the deviation in the vertical direction 2- The position of the triangular wedge in step Ⅰ is 3-1. Make sure that the laser belt of laser 2 is just on the edge of the triangular wedge, and record the position where the acute angle vertex of the triangular wedge is at 0 mm on the horizontal plane as D, and the camera lens The center of 1 is O, the intersection point of the vertical line from the center point O of camera lens 1 to the horizontal plane with a height of 0 mm intersects the horizontal plane is O', the acute angle of the triangular wedge is α, and the angle between the central axis of laser 2 and the vertical direction is θ.

The embodiment shown in Figure 1(c) shows that the angle between the central axis of the laser 2 and the vertical direction is θ, and the acute angle of the triangular wedge is α, and the 2-I step operation process in the second step of identifying the vertical deviation The position of the acute angle vertex of the triangular wedge in position 3-1 of the triangular wedge in the horizontal plane 0mm is recorded as D, and the acute angle vertex of the triangular wedge in position 3-2 of the triangular wedge in the operation process of step 2-II The position at 0 mm on the horizontal plane is denoted as D1, and the forward distance of the triangular wedge is y _x1 , the distance Δy ₁ that the structured light moves on the image plane. During the 2-II step operation process of identifying the vertical deviation, the line laser and the triangular wedge at the position 3-2 of the triangular wedge block during the 2-II step operation process of the second step identification of the vertical deviation The intersection point of the hypotenuse of the block is A, the extension line of the connection line OA between the center O of the camera lens 1 and A intersects the horizontal plane with a height of 0mm at point B, and the vertical line from A to the horizontal plane with a height of 0mm intersects with the horizontal plane at point C, then three The points form △ABC; as the vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0mm intersects the horizontal plane at point O', then the three points O, B, O' form a triangle △OBO'. According to the similarity of △ABC∽△OBO':

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; BC</span>}}{{\mathrm{<span\; class="notranslate">\; BO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, h ₁ is the distance between AC.

The embodiment shown in Figure 1(d) shows that the angle between the central axis of the laser 2 and the vertical direction is θ, and the acute angle of the triangular wedge is α, and the 2-I step operation process in the second step of identifying the vertical deviation The position of the acute angle vertex of the triangular wedge in position 3-1 of the triangular wedge in the horizontal plane is recorded as D, and the acute angle vertex of the triangular wedge in position 3-2 of the triangular wedge in the operation process of step 2-Ⅲ The position at 0 mm on the horizontal plane is denoted as D2, and the forward distance of the triangular wedge is y _x2 , the distance Δy ₂ that the structured light moves on the image plane. In the 2-Ⅲ step operation process of identifying the vertical deviation, the line laser and the triangular wedge at the position 3-3 of the triangular wedge block in the 2-Ⅲ step operation process of the second step identification of the vertical deviation The intersection point of the hypotenuse of the block is A', the extension line of the camera lens 1 center O and A' intersects the horizontal plane with a height of 0 mm at point B', makes a perpendicular line from A' to the horizontal plane with a height of 0 mm, and intersects the horizontal plane with point C', Then three points form △A′B′C′; as the vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0mm intersects the horizontal plane and point O', then the three points O, B′, O' form a triangle △OB′O '. According to the similarity of △A′B′C′∽△OB′O':

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; C</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; o</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, h2 is the distance between _A'C '.

BD=y ₁ , B′B=y ₂ , then BC=y ₁ -(y _x1 -h ₁ /tanθ), B′C′=y ₁ +y ₂ -(y _x1 +y _x2 -h ₁ /tanα ). Assuming that the distance OO′ from the center O of the camera lens 1 to the height 0 mm is H, and the distance from the acute angle vertex of the triangle calibration block 3 to the optical central axis of the camera in the first step is Y, then OO′=H, BO′=Y+y ₁ , B′O′=Y+y ₁ +y ₂ , put in (1) and (2) to get:

$\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, h ₁ =10mm, h ₂ =20mm, α=20°～60°, taking the known quantity into it, we can get:

The values of parameters H and Y can be obtained from formula (4). In addition, the geometric relationship in the figure can be obtained:

y _x1 -AC/tanα=AC*tanθ

Right now:

$\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; AC</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; AC</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

It is known that h ₁ =AC=10mm,α=20°～60°substitute into (5),

From the above formula, tanθ can be calculated.

So far, the camera parameter calibration is completed, and the parameters K, H, Y, and tanθ are obtained respectively.

In Fig. 1(e), the center of the camera lens 1 is O, the intersection point of the vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0 mm intersects the horizontal plane is O’, and the angle between the central axis of the laser 2 and the vertical direction is θ.

The embodiment shown in Figure 1(e) shows that in the case of seam tracking, what is to be measured is the seam height information at the intersection of the seam and the structured light strip, and the weldment 4 to be measured for the seam height information is moved until The welding point where the welding seam height is to be measured intersects with the structured light strip, record the intersection point as M, make the extension line of OM intersect the 0mm plane at point N, record N point and the acute angle vertex D of the triangular wedge in step 2-Ⅰ The distance is y ₃ =ND, it can be seen that y ₃ is the horizontal distance between the intersection point M of the structured light belt and the weld point to measure the weld height and the acute angle vertex D of the triangular wedge in step 2-I, as point M to The vertical line of the 0mm plane, the vertical point is P, and the distance between point M and point P is h ₃ =MP,

According to △MNP∽△ONO’similarity:

$\frac{\mathrm{<span\; class="notranslate">\; MP</span>}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; NP</span>}}{{\mathrm{<span\; class="notranslate">\; NO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$

That is:

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Right now: $h_3=\frac{{\mathrm{the\; y}}_3*h}{{\mathrm{the\; y}}_3+Y+\mathrm{the\; tan}\mathrm{\&theta;}*h}=f\left({\mathrm{the\; y}}_3\right)---\left(7\right)$

It can be seen from the above formula that h ₃ is a function of y ₃ .

Move the weld along the direction of the vertical structured light strip until the structured light strip intersects with the welding point where the height of the weld is to be measured, and record the distance between the intersection point of the weld seam and the structured light strip and the optical central axis OO' of the camera lens through a two-dimensional image The distance between △y3', △y3=△y3'-Y/K, that is, the size of the offset △y3. Taking y1 as an example, in Figure 1(a), the offset obtained through image processing is △y1, then the size of y1 is K*△y1. Similarly, it can be known that y3=K*△y3, so as long as the offset △y3 is obtained through image processing, the value of y3 can be obtained, and then the weld height information h ₃ of this point can be calculated. Set the horizontal height of the initial welding point to 0mm. When the camera has a vertical deviation relative to the weld, it will cause the laser stripes in the image to move in the left and right directions. At this time, the weld does not move, and the image appears When such a change occurs, the deviation between the welding torch and the actual weld seam in the vertical direction can be identified in combination with the above calibration method. So far, the height information of the weld seam at the intersection of the weld seam and the structured light strip has been obtained.

The embodiment shown in Figure 2(a) shows that the positional relationship between the weldment 4 and the camera 5 is shown in (1) in Figure 2(a), and the weld image is shown in (2) in Figure 2(a) . When the camera 5 deviates in the horizontal direction relative to the weld seam 7, the weld seam 7 in the image will move up and down as shown by the dotted line in the figure, and the laser stripe 6 does not move at this time.

The embodiment shown in Figure 2(b) shows that the positional relationship between the weldment 4 and the camera 5 is shown in (1) in Figure 2(b), and the weld seam image is shown in (2) in Figure 2(b) . When there is a vertical deviation between the camera 5 and the welding seam 7, the laser stripe 6 in the image will move in the left and right directions as shown by the dotted line in the figure, and the welding seam 7 does not move at this time.

Example 1

In this embodiment, the wedge-shaped target adopts a triangular wedge block with an acute angle of 30°, and the main equipment used is a camera, a laser, an encoder in the motion mechanism, a triangular wedge block with an acute angle of 30°, weldments, electronics, etc. Wedge-shaped vernier scale and servo motor, the camera mainly involves the camera lens;

The first step is to identify the horizontal deviation

By controlling the motor to move the camera along the parallel direction of the structured light belt for a distance of 10mm, and then calculating the pixel value of the welding seam moving in the up and down direction in the two-dimensional image, and then calculating the actual distance represented by each pixel, you can get The proportional coefficient K between the pixel value and the actual value, K=83;

The second step is to identify the vertical deviation

Step 2-I, operate as in the embodiment shown in Figure 1(b) above, place the triangular wedge under the camera to ensure that the laser belt is just on the edge of the triangular wedge; place the acute angle vertex of the triangular wedge at 0 mm in the horizontal plane The position of is denoted as D,

Step 2-II, perform operations and calculations as in the embodiment shown in Figure 1(c) above, in which the triangular wedge moves forward y _x1 to make the structured light band rise to a height of AC=10mm above the horizontal plane and stop, and the movement mechanism The encoder in the system records the forward movement distance of the triangular wedge block y _x1 =20.118mm, processes and records the moving distance of the structured light strip on the image plane in the two-dimensional image △y ₁ =0.0291mm;

Step 2-Ⅲ, repeat the operation process of step 2-Ⅱ, just continue to move the triangular wedge forward for a distance of y _x2 , so that the structured light belt rises to the height A'C'=20mm above the horizontal plane and stops. The encoder in the mechanism records the forward moving distance of the triangular wedge block y _x2 =20.118mm, processes and records the moving distance of the structured light strip on the image plane in the two-dimensional image △y ₂ =0.03497mm;

The third step is to use the camera to calibrate the parameters of the structured light vision system

As described in the embodiment shown in Figure 1(d) above, it is obtained:

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; BC</span>}}{{\mathrm{<span\; class="notranslate">\; BO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; C</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; o</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

$\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

According to formula (3), where,

h ₁ =AC=10mm, h ₂ =A'C'=20mm,

It can also be known that y ₁ =K*△y ₁ =83*0.0291=2.4153, y ₂ =K*△y ₂ =83*0.03497=2.9024;

Substitute the known amount to get

Calculated H=120.1, Y=-6.8379

Depend on Get tanθ=0.2787

So far, the calibration of the structured light vision system parameters with the camera is completed, and the values of the parameters K, H, Y, and tanθ are respectively obtained: K=83, H=120.1, tanθ=0.2787, Y=-6.8379, where the minus sign does not mean Size, which only represents direction information.

The fourth step, the acquisition of weld height information

In the case of weld seam tracking, what is to be measured is the weld height information at the intersection of the weld seam and the structured light belt, and the weldment 4 whose weld height information is to be measured is moved until the weld point where the weld seam height is to be measured Intersect with the structured light belt, record the intersection point as M, make the extension line of OM intersect the 0mm plane at point N, record the distance between point N and the acute angle vertex D of the triangular wedge in step 2-I as y ₃ =ND, it can be seen that y ₃ It is the horizontal distance between the intersection point M of the structured light strip and the welding point of the welding seam to measure the height of the welding seam and the acute angle vertex D of the triangular wedge block in step 2-I. Make a vertical line from point M to the 0mm plane, and the vertical point is P. Note that the distance between point M and point P is h ₃ =MP,

According to △MNP∽△ONO’similarity:

$\frac{\mathrm{<span\; class="notranslate">\; MP</span>}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; NP</span>}}{{\mathrm{<span\; class="notranslate">\; NO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$

That is:

$\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Right now: $h_3=\frac{{\mathrm{the\; y}}_3*h}{{\mathrm{the\; y}}_3+Y+\mathrm{the\; tan}\mathrm{\&theta;}*h}=f\left({\mathrm{the\; y}}_3\right)---\left(7\right)$

Move the weld along the direction of the vertical structured light strip until the structured light strip intersects with the welding spot where the height of the weld is to be measured, and record the distance between the intersection point of the weld seam and the structured light strip and the optical central axis OO' of the camera lens through a two-dimensional image The distance between △y3'=-0.05168mm, according to △y3=△y3'-Y/K=-0.05168-(-6.8379)/83=0.0307mm, the offset △y3 can be obtained, and then y3= K*△y3=83*0.0307=2.5481. The above negative signs do not represent the size, but only the direction information.

According to formula (7) $h_3=\frac{{\mathrm{the\; y}}_3*h}{{\mathrm{the\; y}}_3+Y+\mathrm{the\; tan}\mathrm{\&theta;}*h},$

Available weld height

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2.5481</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 120.1</span>}}{\mathrm{<span\; class="notranslate">\; 2.5481</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6.8379</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.2787</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 120.1</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10.50</span>}\mathrm{<span\; class="notranslate">\; mm</span>}$

Use the electronic wedge vernier to measure the actual height of the weld seam to obtain a weld seam height of 10.45mm.

By multiplying the two-dimensional image coordinates and the ratio K, the two-dimensional coordinates (181, 130.5) of the weld spot to be tested can be determined, and the reconstruction of the weld three-dimensional camera coordinates (181, 130.5, 10.50) is realized.

Example 2

In addition to controlling the motor to move the camera along the vertical structured light direction, the triangular wedge moves forward to make the structured light strip rise to a height of h ₁ =AC=5mm above the horizontal plane and stop at a distance of y _x1 =10.079mm and the structured light is at The moving distance on the image plane is △y ₁ =0.01424mm; the triangular wedge moves forward to make the structured light belt rise to a height of h ₂ =A′C′=15mm above the horizontal plane and stop at a distance of y _x2 =20.157mm and the distance that the structured light moves on the image plane Δy ₂ =0.0325mm; other than that, the others are the same as in Embodiment 1,

The camera calibration results are K=82.8, H=121.3, tanθ=0.2826, Y=-6.8612, where the negative sign does not represent the size, but only the direction information.

Move the weld along the vertical direction of the structured light until the structured light intersects the welding point where the height of the weld is to be measured, and record the distance between the intersection of the weld and the structured light band and the optical central axis OO' of the camera lens through a two-dimensional image Distance △y3'=-0.05176mm, according to △y3=△y3'-Y/K=-0.05176-(-6.8612)/82.8=0.0311mm, the offset can be obtained, and then y3=K*△y3 can be obtained =82.8*0.0311=2.5751. The above negative signs do not represent the size, but only the direction information.

According to formula (7) Available weld height

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2.5751</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 121.3</span>}}{\mathrm{<span\; class="notranslate">\; 2.5751</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6.8612</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.2826</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 121.3</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10.31</span>}\mathrm{<span\; class="notranslate">\; mm</span>}$

Using the electronic wedge vernier to measure the actual height of the weld seam, the weld seam height is 10.31mm.

The two-dimensional coordinates (181.5, 130.2) of the weld point to be tested can be determined by the product of the two-dimensional image coordinates and the ratio K, and the reconstruction of the three-dimensional camera coordinates (181.5, 130.2, 10.31) of the weld can be realized.

Example 3

In addition to controlling the motor to move the camera along the parallel direction of the structured light strip by 15mm, the triangular wedge moves forward to make the structured light strip rise to a height of h ₁ =AC=15mm above the horizontal plane and stop at a distance of y _x1 =28.737mm and The moving distance of the structured light on the image plane is △y ₁ =0.02594; the triangular wedge moves forward to make the structured light belt rise to a height of h ₂ =A′C′=25mm above the horizontal plane and stop at a distance of y _x2 = 21.443mm and the moving distance of structured light on the image plane △y ₂ =0.05701mm; other than that, the same as embodiment 1,

The camera calibration results are K=83.1, H=119.8, tanθ=0.2741, Y=-6.8856, where the negative sign does not represent the size, but only the direction information.

Move the weld along the vertical direction of the structured light until the structured light intersects the welding point where the height of the weld is to be measured, and record the distance between the intersection of the weld and the structured light band and the optical central axis OO' of the camera lens through a two-dimensional image Distance △y3'=-0.05316mm, according to △y3=△y3'-Y/K=-0.05316-(-6.8856)/83.1=0.0297mm, the offset can be obtained, and then y3=K*△y3 can be obtained =83.1*0.0297=2.4681. The above negative signs do not represent the size, but only the direction information.

According to formula (7) Available weld height

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2.4681</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 119.8</span>}}{\mathrm{<span\; class="notranslate">\; 2.4681</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6.8856</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.2741</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 119.8</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10.40</span>}\mathrm{<span\; class="notranslate">\; mm</span>}$

Use the electronic wedge vernier to measure the actual height of the weld seam to obtain a weld seam height of 10.45mm.

Through the product of the two-dimensional image coordinates and the ratio K, the two-dimensional coordinates (181.2, 130.2) of the weld point to be tested can be determined to realize the reconstruction of the three-dimensional camera coordinates of the weld (181.2, 130.3, 10.40).

Example 4

Adopting a triangular wedge block with an acute angle of 20° except that the triangular wedge target used; in addition, other is the same as embodiment 1, and the different results obtained in the calibration process are as follows:

In the first step of horizontal deviation identification step, the camera is moved 10mm along the parallel direction of the structured light belt by controlling the motor, and then the pixel value of the welding seam moving in the up and down direction in the two-dimensional image is calculated, and then each The actual distance represented by the pixel can get the proportional coefficient K between the pixel value and the actual value, K=83.2;

In the second step of identifying the deviation in the vertical direction, in step 2-II, the triangular wedge moves forward y _x1 to make the structured light strip rise to a height of AC=10mm above the horizontal plane and stop, and the movement mechanism The encoder records the forward moving distance of the triangular wedge block y _x1 =30.306mm, processes and records the moving distance of the structured light strip on the image plane in the two-dimensional image △y ₁ =0.0297mm; in step 2-Ⅲ, repeat 2- In the operation process of step II, just continue to move the triangular wedge forward for a distance of y _x2 , so that the structured light strip rises to a height of A'C'=20mm above the horizontal plane and stops, and the encoder in the motion mechanism records the triangular wedge The block moves forward for a distance of y _x2 =30.403mm, and the distance △y ₂ =0.0369mm of the structured light strip moving on the image plane is processed and recorded in the two-dimensional image;

In the third step, in the process of calibrating the parameters of the structured light vision system with the camera,

By formula (3)

$\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.$

in,

h ₁ =AC=10mm, h ₂ =A'C'=20mm, α=20°

It can also be known that y ₁ =K*△y ₁ =83.2*0.0297=2.4680, y ₂ =K*△y ₂ =83.2*0.0369=3.0729;

Substitute the known amount to get

Calculated H=121.1, Y=-6.8643

Depend on Get tanθ=0.2831

So far, the calibration of the structured light vision system parameters with the camera is completed, and the values of the parameters K, H, Y, and tanθ are respectively obtained: K=83.2, H=121.1, tanθ=0.2831, Y=-6.8643, where the minus sign does not mean Size, only represents direction information.

In the fourth step of obtaining the weld height information, the weld is moved along the vertical direction of the structured light until the structured light intersects the spot where the weld height is to be measured, and the intersection point of the weld and the structured light band is recorded by a two-dimensional image The distance △y3'=-0.05221mm between the point and the optical central axis OO' of the camera lens can be obtained according to △y3=△y3'-Y/K=-0.05221-(-6.8643)/83.2=0.0303mm Offset, and then y3=K*△y3=83.2*0.0303=2.5210 can be obtained. The above negative signs do not represent the size, but only the direction information.

According to formula (7) Available weld height

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2.5210</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 121.1</span>}}{\mathrm{<span\; class="notranslate">\; 2.5210</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6.8643</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.2831</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 121.1</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10.20</span>}\mathrm{<span\; class="notranslate">\; mm</span>}$

Using the electronic wedge vernier to measure the actual height of the weld seam, the weld seam height is 10.43mm.

The two-dimensional coordinates (179.5, 131.2) of the weld spot to be tested can be determined by the product of the two-dimensional image coordinates and the ratio K, and the reconstruction of the weld three-dimensional camera coordinates (179.5, 131.2, 10.20) is realized.

Example 5

Adopting a triangular wedge block with an acute angle of 60° except that the triangular wedge target used; in addition, other is the same as embodiment 1, and the different results obtained in the calibration process are as follows:

In the first step of horizontal deviation identification step, the camera is moved 10mm along the parallel direction of the structured light belt by controlling the motor, and then the pixel value of the welding seam moving in the up and down direction in the two-dimensional image is calculated, and then each The actual distance represented by the pixel can get the proportional coefficient K between the pixel value and the actual value, K=83.7;

In the second step of identifying the deviation in the vertical direction, in step 2-II, the triangular wedge moves forward y _x1 to make the structured light strip rise to a height of AC=10mm above the horizontal plane and stop, and the movement mechanism The encoder records the forward movement distance of the triangular wedge block y _x1 =8.5675mm, processes and records the distance △y ₁ =0.0289mm of the structured light strip moving on the image plane in the two-dimensional image; in step 2-Ⅲ, repeat 2- In the operation process of step II, just continue to move the triangular wedge forward for a distance of y _x2 , so that the structured light strip rises to a height of A'C'=20mm above the horizontal plane and stops, and the encoder in the motion mechanism records the triangular wedge The block moves forward for a distance of y _x2 =8.5694mm, and the distance △y ₂ =0.0348mm of the structured light strip moving on the image plane is processed and recorded in the two-dimensional image;

In the third step, in the process of calibrating the parameters of the structured light vision system with the camera,

By formula (3)

$\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.$

in,

h ₁ =AC=10mm, h ₂ =A'C'=20mm, α=60°

It can also be known that y ₁ =K*△y ₁ =83.7*0.0289=2.4204, y ₂ =K*△y ₂ =83.7*0.0348=2.9108;

Substitute the known amount to get

Calculated H=119.2, Y=-6.8731

Depend on Get tanθ=0.2794

So far, the calibration of the structured light vision system parameters with the camera is completed, and the values of the parameters K, H, Y, and tanθ are obtained respectively: K=83.7, H=119.2, tanθ=0.2794, Y=-6.8731, where the minus sign does not mean Size, only represents direction information.

In the fourth step of obtaining the weld height information, the weld is moved along the vertical direction of the structured light until the structured light intersects the spot where the weld height is to be measured, and the intersection point of the weld and the structured light band is recorded by a two-dimensional image The distance between △y3'=-0.05315mm and the optical central axis OO' of the camera lens can be obtained according to △y3=△y3'-Y/K=-0.05315-(-6.8731)/83.7=0.0289mm Offset, and then y3=K*△y3=83.7*0.0289=2.4189 can be obtained. The above negative signs do not represent the size, but only the direction information.

According to formula (7) Available weld height

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2.4189</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 119.2</span>}}{\mathrm{<span\; class="notranslate">\; 2.4189</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 6.8731</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.2794</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 119.2</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 9.99</span>}\mathrm{<span\; class="notranslate">\; mm</span>}$

Use the electronic wedge vernier to measure the actual height of the weld seam to obtain a weld seam height of 10.23mm.

The two-dimensional coordinates (179.1, 130.2) of the weld point to be tested can be determined by the product of the two-dimensional image coordinates and the ratio K, and the reconstruction of the three-dimensional camera coordinates of the weld (179.1, 130.2, 9.99) is realized.

In the above embodiments, the laser used is a line laser.

Welds with different positions and heights of the weldment were selected for multiple tests. In the test, the electronic wedge vernier was used to actually measure the height of the weld to obtain the information of the weld height. The actual three-dimensional coordinate information of the weld and the three-dimensional coordinate information obtained by calibration were respectively fitted, and the fitting result curve is shown in Fig. 3.

The experimental results shown in Figure 3 show that the absolute error in the X direction is within 0.5mm, the absolute error in the height Z measurement is within 1mm, and the mean square error is 0.707. The experiment shows that the algorithm has high accuracy and stability and meets the engineering accuracy requirements. .

The equipment involved in the above examples are all commercially available.

Claims (3)

1. The structured light vision sensor parameter calibration method for obtaining weld height information is characterized in that: it is a structured light vision system parameter calibration method based on a wedge-shaped target, and the steps are: The wedge-shaped target adopts a triangular wedge, and the main equipment used includes a camera, a laser, an encoder in the motion mechanism, a triangular wedge with an acute angle of 20° to 60°, weldment, electronic wedge vernier and servo The motor, the camera mainly involves the camera lens; The first step is to identify the horizontal deviation By controlling the servo motor, the camera is moved along the parallel direction of the structured light belt for a distance of 5mm to 25mm and measured and recorded, and then the pixel value of the welding seam moving in the up and down direction in the two-dimensional image is calculated, and then the actual value represented by each pixel is calculated. The distance can be used to obtain the proportional coefficient K between the pixel value and the actual distance value; The second step is to identify the vertical deviation Step 2-I, place the triangular wedge under the camera to ensure that the laser belt hits the edge of the triangular wedge; Step 2-II, move the triangular wedge forward for a distance of y _x1 , the structured light strip produced by the laser will move upward along the hypotenuse of the triangular wedge until the structured light strip rises to The height above the horizontal plane AC=10mm stops, the moving distance of the structured light strip on the image plane is △y ₁ , the encoder in the motion mechanism will record the forward moving distance y _x1 of the triangular wedge during the moving process, in two dimensions Process and record the distance Δy ₁ of the structured light band moving on the image plane in the image; Step 2-Ⅲ, repeat the operation process of step 2-Ⅱ, but continue to move the triangular wedge block forward for a distance of y _x2 until the structured light strip rises to the height A'C'=20mm above the horizontal plane and stops. At this time, the moving distance of the structured light strip on the image plane is △y ₂ , and the encoder in the motion mechanism will record the distance y _x2 of the forward movement of the triangular wedge during the moving process, and process and record the structured light strip in the two-dimensional image The moving distance △y ₂ on the image plane; The third step is to use the camera to calibrate the parameters of the structured light vision system Mark the position of the horizontal plane 0 mm where the acute angle apex of the triangular wedge in the position of the triangular wedge in the operation of step 2-I is D, and place it in the position of the triangular wedge in the operation of step 2-II The position of the acute-angle apex of the triangular wedge in the horizontal plane 0 mm is marked as D1, and the position of the acute-angle apex of the triangular wedge in the position of the triangular wedge in the operation process of step 2-Ⅲ is marked as D2, Then the distance between D and D1 is y _x1 , the distance between D1 and D2 is y _x2 , Let the intersection point of the line laser emitted by the laser and the hypotenuse of the triangular wedge in the position of the triangular wedge in step 2-Ⅱ be A, the center of the camera lens be marked as O, and the intersection of the extension line of the camera lens center O and A The horizontal plane with a height of 0mm is at point B, draw a vertical line from A to the horizontal plane with a height of 0mm, and intersect the horizontal plane at point C, then the three points form △ABC; draw a vertical line from the center point O of the camera lens to the horizontal plane with a height of 0mm, and intersect the horizontal plane at Point O', then three points O, B, O' form a triangle △OBO', record the distance between AC as h ₁ , According to △ABC∽△OBO’similarity: $\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; BC</span>}}{{\mathrm{<span\; class="notranslate">\; BO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$ Let the intersection point of the line laser emitted by the laser and the hypotenuse of the triangular wedge at the position of the triangular wedge in step 2-Ⅲ be A', and the intersection of the extension line of the camera lens center O and A' with a height of 0mm is at At point B', make a vertical line from A' to the horizontal plane with a height of 0mm, and intersect the horizontal plane at point C', then the three points form △A'B'C'; make a vertical line from the center point O of the camera lens 1 to the horizontal plane with a height of 0mm Intersect the horizontal plane with point O', then three points O, B', O' form a triangle △OB'O', record the distance between A'C' as h ₂ , According to the similarity of △A'B'C'∽△OB'O': $\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; C</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; B</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{\mathrm{<span\; class="notranslate">\; o</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$ Denote the angle between the central axis of the laser and the vertical direction as θ, denote BD=y ₁ , B'B=y ₂ , Then BC=y ₁ -(y _x1 -h ₁ /tanθ), B'C'=y ₁ +y ₂ -(y _x1 +y _x2 -h ₁ /tanα), record the distance OO from the lens center O to the height 0mm ' is H, the distance from the acute angle vertex D of the triangular wedge block to the camera optical central axis OO' in step 2-I is Y, that is, OO'=H, DO'=Y, and BO'=Y+y ₁ , B can be obtained 'O'＝Y+y ₁ +y ₂ , put in (1) and (2) to get: $\left\{\begin{array}{c}\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\\ \frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$ Among them, h ₁ =10mm, h ₂ =20mm, α=20°～60°, taking the known quantity into it, we can get: The values of parameters H and Y can be obtained from formula (4). In addition, the geometric relationship in the figure can be obtained: y _x1 -AC/tanα=AC*tanθ Right now: $\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; AC</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}{\mathrm{<span\; class="notranslate">\; AC</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$ It is known that h ₁ =AC=10mm, α=20°～60°substituting into (5), From the above formula, tanθ can be calculated, So far, the camera is used to calibrate the parameters of the structured light vision system, and the values of the parameters K, H, Y, and tanθ are respectively obtained; The fourth step, the acquisition of weld height information In the case of seam tracking, what is to be measured is the seam height information at the intersection of the seam and the structured light belt, and the weldment to be measured for seam height information is moved until the weld point where the seam height is to be measured is in contact with the The structured light bands intersect, mark the intersection point as M, make the extension line of OM intersect the 0mm plane at point N, record the distance between point N and the acute angle vertex D of the triangular wedge in step 2-I as y ₃ =ND, it can be seen that y ₃ is The horizontal distance between the intersection point M of the structured light strip and the welding point of the weld to be measured for the height of the weld and the apex D of the acute angle of the triangular wedge in step 2-I is the perpendicular line from point M to the 0mm plane, and the perpendicular point is P, record The distance between point M and point P is h ₃ =MP, According to △MNP∽△ONO’similarity: $\frac{\mathrm{<span\; class="notranslate">\; MP</span>}}{{\mathrm{<span\; class="notranslate">\; OO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; NP</span>}}{{\mathrm{<span\; class="notranslate">\; NO</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}$ That is: $\frac{{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; tan</span>}\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$ Right now: $h_3=\frac{{\mathrm{the\; y}}_3*h}{{\mathrm{the\; y}}_3+Y+\mathrm{the\; tan}\mathrm{\&theta;}*h}=f\left({\mathrm{the\; y}}_3\right)---\left(7\right)$ It can be seen from the above formula that h ₃ is a function of y ₃ , Move the weld along the direction of the vertical structured light strip until the structured light strip intersects with the welding point where the height of the weld is to be measured, and record the distance between the intersection point of the weld seam and the structured light strip and the optical central axis OO' of the camera lens through a two-dimensional image The distance between them is △y3', then △y3=△y3'-Y/K, that is, the size of the offset △y3 is obtained, then the size of y3 is y3=K*△y3, and the level height of the initial welding point is set 0mm, when there is a vertical deviation of the camera relative to the weld, it will cause the laser stripes in the image to move in the left and right directions, and the weld does not move at this time. When such a change occurs in the image, through the image Process and record the displacement offset △y3 of the intersection point of the welding seam and the structured light belt, and then know that y3=K*△y3, combined with the formula (7), the deviation between the welding torch and the actual welding seam in the vertical direction can be identified. So far, The height information of the welding seam at the intersection of the welding seam and the structured light belt is obtained, and the reconstruction of the coordinates of the three-dimensional camera is realized. 2. The method for calibrating parameters of a structured light vision sensor for obtaining weld height information according to claim 1, wherein the laser is a line laser. 3. The method for calibrating parameters of a structured light vision sensor for obtaining weld height information according to claim 1, wherein one of the acute angles of the triangular wedge is 30°.