Shaft part radial total run-out error measurement method based on image processing

Abstract

The invention discloses a shaft part radial total run-out error measurement method based on image processing. The method comprises the following steps: S1, acquiring an image of a measured part; s2, establishing an initial point cloud image based on a three-dimensional coordinate system O1 according to the image of the measured part; s3, in the initial point cloud image, determining the axis of the image by using a fitting algorithm; s4, establishing a new three-dimensional coordinate system Oe by taking the axis as a coordinate axis; s5, establishing a conversion matrix T between the three-dimensional coordinate system O1 and the three-dimensional coordinate system Oe, and converting the point cloud image data based on the three-dimensional coordinate system Oe into the point cloud image data in the three-dimensional coordinate system O1 through the conversion matrix T; and S6, calculating the distance from the midpoint of the point cloud to the Z axis, and obtaining an error value by using an error judgment method. According to the error measurement method, the problems of unreasonable reference establishment and inaccurate data acquisition in the prior art are solved.

Description

Technical Field

[0001] The invention belongs to the technical field of instrument measurement, and in particular relates to a method for measuring radial total runout errors of shaft parts based on image processing. Background Art

[0002] In the field of precision measurement, the accuracy of mechanical parts is very important for the operation of the machine. The participation rate of shaft parts in mechanical transmission is 100%, which has the functions of transmission, positioning, load bearing and support. The form and position tolerances of shaft parts include straightness, coaxiality, roundness, cylindricity, circular runout and total runout error, among which the total runout error is one of the comprehensive errors of shaft parts. The total runout is a comprehensive error item set based on a specific detection method, which can fully reflect the shape and position errors of the measured elements. Therefore, the runout tolerance can comprehensively control the position, direction and shape errors of the measured elements. The comprehensiveness of the total runout can include roundness, cylindricity and circular runout in a certain sense. Therefore, there is a saying in the industry: "If in doubt, use runout". The total runout refers to the runout along the entire measured surface when the part rotates continuously around the reference axis. The total runout tolerance is the maximum runout allowed when the actual element being measured rotates continuously around the reference axis and indicates the relative movement of the ideal profile along its entire surface profile. The ideal rotating surface of radial total runout is a cylindrical surface with the datum element as the axis. The tolerance zone of radial total runout is a cylindrical surface with the same axis, and the error value is the difference between the two cylinder radii.

[0003] So far, the method for calculating the radial total runout error is approximate processing, which is to install the measured part on the left and right tips, use the dial indicator and the digital display pointer to move on the continuously rotating part, read or record the measured data, and obtain the difference between the maximum and minimum values ​​as the error value. The installation basis of the measured part is inconsistent with the evaluation basis of the radial total runout error, which leads to poor accuracy and low credibility of the measurement results obtained by the traditional evaluation method. Summary of the invention

[0004] The purpose of the present invention is to provide a method for measuring the radial total runout error of shaft parts based on image processing, which solves the problems of unreasonable benchmark establishment and inaccurate data collection in the prior art.

[0005] The technical solution adopted by the present invention is a method for measuring the radial total runout error of shaft parts based on image processing, comprising the following steps: S1, collecting images of the parts to be tested; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

[0006] The present invention is also characterized in that: The specific process of S1 is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3. Collect the cross-sectional circular image information of the section A, the section B and the sampling area W at the shaft section of the measured part between the sections A and B respectively.

[0007] The process of collecting the cross-sectional circular image information of the sampling area W in S1.3 is as follows: when collecting the image at the W position, interval sampling is performed, and the interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of several cross-sectional circles.

[0008] The specific process of S2 is: S2.1. Project the contour information of the measured part through the telecentric lens imaging method; collect Q+1 images of the contour information at the prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extract pixel information at the cross-section circle; the pixel position information of the pixel information is j, j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.4. Determine pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.5. Combined with motion control, add Z-axis height information to the cross-sectional profile polar coordinate circle according to the corresponding distance, and finally display the collected data in the three-dimensional coordinate system O1; where L = ab, L2 = eb, and fe is the width of the sampling area W.

[0009] The specific process of S3 is: S3.1. Determine the fitting circle centers Oa, Ob, and Ob of A, B and section W respectively by the least squares circle method. e With O f ; S3.2, Connect O a , O b , the straight line O a O b The extension line G is used as the reference axis for error analysis.

[0010] The specific process of S4 is: e As the reference point, O e O f The extension line of is the Z axis, passing through O e With the XO1Z plane parallel and perpendicular to O e Z1 O e X1 is the new X axis; e , O f It is the intersection of the first cross-sectional circle and the last cross-sectional circle of the sampling area W with the reference axis G.

[0011] The specific process of S5 is as follows: S5.1. Calculate the angle between the reference axis and the Z axis in the O1 coordinate system. ; S5.2. Calculate O1 and O e The transformation matrix T between the two coordinate systems; S5.3, according to the angle , will be in O e The point cloud image data under the coordinates is converted into point cloud image data of the three-dimensional coordinate system O1 through the transformation matrix T.

[0012] The specific process of S5.2 is as follows: S5.2.1. Assuming a direction vector is a normalized vector, i.e. , according to the direction vector, construct the antisymmetric moment according to formula (1) ; (1); S5.2.2. Calculate the rotation matrix using formula (2) ; (2); In the formula, is the identity matrix, is the angle between the direction vector and the Z axis, by calculating the direction vector With the standard basis vectors The dot product gives ,Right now ,as well as ; S5.2.3. Calculate according to formula (3) ; (3); S5.2.4. According to formula (4), the rotation matrix is ​​obtained ; (4); S5.2.5, according to formula (5), the transformation matrix T is obtained; (5).

[0013] The specific process of S6 is as follows: S6.1, projecting the converted point cloud image data of the three-dimensional coordinate system O1 into the two-dimensional coordinate system XO1Y after removing the Z-axis height information; S6.2, with the reference point O1 as the center of the circle, find the farthest point P1 and the closest point P2 from the center O1 on the point cloud, draw circles with R1 and R2 as radii respectively, and obtain two concentric circles F1 and F2; S6.3. Calculate the final radial total runout error value t according to equations (6) and (7); (6); t = R2 - R1 (7); In the formula, R 1= O1P1, R2=O1P2; (Xi, Yj) is a point on the point cloud.

[0014] The beneficial effects of the present invention are: The image processing-based radial total runout error measurement method for shaft parts provided by the present invention combines the two-dimensional image information collected by a two-dimensional camera with the one-dimensional position information provided by a motion control system to synthesize the three-dimensional space point cloud information of the shaft parts. Compared with data acquisition methods such as three-coordinate and laser ranging, this method is simpler, has a shorter acquisition time and has higher accuracy; based on the three-dimensional space point cloud established by the motion control system, the error analysis reference axis is determined by positional relationship and least squares circle method. The determination of the axis is the most important step in calculating the error. This point is ignored in most measurement methods, or the inaccuracy of establishing the reference will have a great impact on the calculation error. The present invention determines the reference based on the definition of radial total runout error, uses the coordinate transformation method to calculate the distance from the midpoint of the converted point cloud to the Z axis, and uses the error judgment method to directly obtain the error value, reducing the difficulty of error analysis. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 It is a flow chart of a method for measuring radial total runout error of shaft parts based on image processing of the present invention; Figure 2It is a schematic diagram of an image acquisition method of a method for measuring radial total runout error of shaft parts based on image processing of the present invention; Figure 3 It is the cross-section edge profile information diagram of the measured component of the present invention; Figure 4 is a schematic diagram of a polar coordinate circle of a cross-sectional profile of a component under test according to the present invention; Figure 5 It is an amplified diagram of the polar coordinate circle OSPA error of the cross-sectional profile of the measured component of the present invention; Figure 6 is a schematic diagram of an original point cloud image of a component under test of the present invention; Figure 7 is a schematic diagram of the central axis of the original point cloud image of the measured component of the present invention; Figure 8 is the new coordinate system O in the original point cloud image of the measured component of the present invention. e Schematic diagram; Fig. 9 is a schematic diagram of a point cloud image of a measured component in a three-dimensional coordinate system O1 of the present invention; Fig.10 It is a distance and error range diagram of the point cloud of the measured component to the Z axis of the present invention; Fig.11 is a schematic diagram of the appearance of a step shaft to be measured in Example 6 of the present invention; Fig.12 is a stepped shaft profile image diagram in Example 6 of the present invention; Fig.13 is a polar coordinate diagram of the center position of the contour fitting circle of the step shaft reference circle A in Example 6 of the present invention; Fig.14 is a polar coordinate diagram of the center position of the contour fitting circle of the step shaft reference circle B in Example 6 of the present invention; Fig.15 3 is a diagram of the circular runout measurement error range in Example 6 of the present invention. DETAILED DESCRIPTION

[0015] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0016] Example 1 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Figure 1 As shown, the following steps are included: S1, collecting images of the parts to be tested; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

[0017] Example 2 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Figure 1 As shown, the following steps are included: S1, collecting images of the parts to be tested; like Figure 2 As shown, the specific process is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3, respectively collect cross-sectional circular image information of section A, section B and sampling area W at the shaft section of the measured part between section A and section B; The process of collecting the cross-sectional circular image information of the sampling area W is as follows: when collecting the image at the W position, interval sampling is performed, and interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of several cross-sectional circles; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

[0018] Example 3 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Figure 1 As shown, the following steps are included: S1, collecting images of the parts to be tested; like Figure 2 As shown, the specific process is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3, respectively collect cross-sectional circular image information of section A, section B and sampling area W at the shaft section of the measured part between section A and section B; The process of collecting the cross-sectional circular image information of the sampling area W is as follows: when collecting the image at the W position, interval sampling is performed, and interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of several cross-sectional circles; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; like Figure 3 As shown, the specific process is: S2.1. Project the contour information of the measured part through the telecentric lens imaging method; collect Q+1 images of the contour information at the prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extract pixel information at the cross-section circle; the pixel position information of the pixel information is j, j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.4, such as Figure 4 , 5 As shown, determine the pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.5, combined with motion control, add Z-axis height information to the cross-section profile polar coordinate circle according to the corresponding distance, such as Figure 6 As shown, the collected data is finally displayed in the three-dimensional coordinate system O1; wherein L = ab, L2 = eb, and fe is the width of the sampling area W; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

[0019] Example 4 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Figure 1 As shown, the following steps are included: S1, collecting images of the parts to be tested; like Figure 2 As shown, the specific process is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3, respectively collect cross-sectional circular image information of section A, section B and sampling area W at the shaft section of the measured part between section A and section B; The process of collecting the cross-sectional circular image information of the sampling area W is as follows: when collecting the image at the W position, interval sampling is performed, and interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of several cross-sectional circles; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; like Figure 3 As shown, the specific process is: S2.1. Project the contour information of the measured part through the telecentric lens imaging method; collect Q+1 images of the contour information at the prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extract pixel information at the cross-section circle; the pixel position information of the pixel information is j, j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.4, such as Figure 4 , 5 As shown, determine the pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.5, combined with motion control, add Z-axis height information to the cross-section profile polar coordinate circle according to the corresponding distance, such as Figure 6 As shown, the collected data is finally displayed in the three-dimensional coordinate system O1; wherein L = ab, L2 = eb, and fe is the width of the sampling area W; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; The specific process is: S3.1, if Figure 7 As shown in the figure, the fitting centers Oa, Ob, and Ob of A, B and the cross section W are determined by the least squares circle method. e With O f ; S3.2, Connect O a , O b , the straight line O a O b The extension line G of is used as the reference axis for error analysis; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; The specific process is: Figure 8 As shown, e As the reference point, O e O f The extension line of is the Z axis, passing through O e With the XO1Z plane parallel and perpendicular to O e Z1 O e X1 is the new X axis; e , O f It is the intersection of the first and last cross-sectional circles of the sampling area W with the reference axis G; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

[0020] Example 5 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Figure 1 As shown, the following steps are included: S1, collecting images of the parts to be tested; like Figure 2 As shown, the specific process is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3, respectively collect cross-sectional circular image information of section A, section B and sampling area W at the shaft section of the measured part between section A and section B; The process of collecting the cross-sectional circular image information of the sampling area W is as follows: when collecting the image at the W position, interval sampling is performed, and interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of several cross-sectional circles; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; like Figure 3 As shown, the specific process is: S2.1. Project the contour information of the measured part through the telecentric lens imaging method; collect Q+1 images of the contour information at the prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extract pixel information at the cross-section circle; the pixel position information of the pixel information is j, j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.4, such as Figure 4 , 5 As shown, determine the pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.5, combined with motion control, add Z-axis height information to the cross-section profile polar coordinate circle according to the corresponding distance, such as Figure 6 As shown, the collected data is finally displayed in the three-dimensional coordinate system O1; wherein L = ab, L2 = eb, and fe is the width of the sampling area W; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; The specific process is: S3.1, if Figure 7 As shown in the figure, the fitting centers Oa, Ob, and Ob of A, B and the cross section W are determined by the least squares circle method. e With O f ; S3.2, Connect O a , O b , the straight line O a O b The extension line G of is used as the reference axis for error analysis; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; The specific process is: Figure 8 As shown, e As the reference point, O e O f The extension line of is the Z axis, passing through O e With the XO1Z plane parallel and perpendicular to O e Z1 Oe X1 is the new X axis; e , O f It is the intersection of the first and last cross-sectional circles of the sampling area W with the reference axis G; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; The specific process is: S5.1. Calculate the angle between the reference axis and the Z axis in the O1 coordinate system. ; S5.2. Calculate O1 and O e The transformation matrix T between the two coordinate systems; The specific process is: S5.2.1. Assuming a direction vector is a normalized vector, i.e. , according to the direction vector, construct the antisymmetric moment according to formula (1) ; (1); S5.2.2. Calculate the rotation matrix using formula (2) ; (2); In the formula, is the identity matrix, is the angle between the direction vector and the Z axis, by calculating the direction vector With the standard basis vectors The dot product gives ,Right now ,as well as ; S5.2.3. Calculate according to formula (3) ; (3); S5.2.4. According to formula (4), the rotation matrix is ​​obtained ; (4); S5.2.5, according to formula (5), the transformation matrix T is obtained; (5); S5.3, according to the angle ,like Fig. 9 As shown, it will be e The point cloud image data under the coordinates is converted into point cloud image data of the three-dimensional coordinate system O1 through the transformation matrix T; S6, calculating the distance from the midpoint of the point cloud to the Z axis, and using the error determination method to obtain the error value; The specific process is: S6.1, if Fig.10 As shown, the converted point cloud image data of the three-dimensional coordinate system O1 is projected into the two-dimensional coordinate system XO1Y after removing the Z-axis height information; S6.2, with the reference point O1 as the center of the circle, find the farthest point P1 and the closest point P2 from the center O1 on the point cloud, draw circles with R1 and R2 as radii respectively, and obtain two concentric circles F1 and F2; S6.3. Calculate the final radial total runout error value t according to equations (6) and (7); (6); t = R2 - R1 (7); In the formula, R 1= O1P1, R2=O1P2; (Xi, Yj) is a point on the point cloud.

[0021] Example 6 The radial total runout error measurement method of shaft parts based on image processing proposed in this embodiment is as follows: Fig.12 As shown in the figure, the stepped shaft for turbine shaft has a total runout of 0.01 mm. The measuring platform is used for image acquisition, and the collected images are processed by Halcon 22.11, including the following steps: S1, collecting images of the parts to be tested; like Fig.13 As shown, the specific process is: S1.1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.3, respectively collect the cross-sectional circular image information of the sampling area W at the shaft section of the measured part between the sections A and B. Since the stepped shaft is mounted on the turntable of the equipment through the upper and lower vertices, the collected image includes the vertices; The process of collecting the cross-sectional circular image information of the sampling area W is as follows: when collecting the image at the W position, interval sampling is performed, and interval sampling is performed along the direction from the reference B to the reference A at the position L2 from the reference B section. Collect information of 10 cross-section circles; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; The specific process is: S2.1. Project the contour information of the measured part through the telecentric lens imaging method; collect Q+1 images of the contour information at the prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extract pixel information at the cross-section circle; the pixel position information of the pixel information is j, j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.4. Determine pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.5, combined with motion control, add Z-axis height information to the cross-section profile polar coordinate circle according to the corresponding distance, and finally display the collected data in the three-dimensional coordinate system O1; where L = ab, L2 = eb, and fe is the width of the sampling area W; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; The specific process is: S3.1, if Fig.14 As shown in the figure, the fitting centers Oa, Ob, and Ob of A, B and the cross section W are determined by the least squares circle method. e With O f ;O a (0.033, 0.0196) and O b (0.032, 0.095); Based on the height of the reference circle, the Z-axis coordinates of the two centers are 18.894 and 29.043, and the plane expression containing the axis is obtained using the axis fitting algorithm: ; S3.2, Connect O a , O b , the straight line O a O b The extension line G of is used as the reference axis for error analysis; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; The specific process is: e As the reference point, O e O f The extension line of is the Z axis, passing through O e With the XO1Z plane parallel and perpendicular to O e Z1 O e X1 is the new X axis; e , O f It is the intersection of the first and last cross-sectional circles of the sampling area W with the reference axis G; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; The specific process is: S5.1. Calculate the angle between the reference axis and the Z axis in the O1 coordinate system. ; S5.2. Calculate O1 and O e The transformation matrix T between the two coordinate systems; The specific process is: S5.2.1. Assuming a direction vector is a normalized vector, i.e. , according to the direction vector, construct the antisymmetric moment according to formula (1) ; (1); S5.2.2. Calculate the rotation matrix using formula (2) ; (2); In the formula, is the identity matrix, is the angle between the direction vector and the Z axis, by calculating the direction vector With the standard basis vectors The dot product gives ,Right now ,as well as ; S5.2.3. Calculate according to formula (3) ; (3); S5.2.4. According to formula (4), the rotation matrix is ​​obtained ; (4); S5.2.5, according to formula (5), the transformation matrix T is obtained; (5); S5.3, according to the angle , will be in O e The point cloud image data under the coordinates is converted into point cloud image data of the three-dimensional coordinate system O1 through the transformation matrix T; Taking the three-dimensional coordinates of a set of sampling circles as an example, the coordinates are transformed and the homogeneous transformation matrix is ​​used to operate , get the rotated point cloud data; S6, calculating the distance from the midpoint of the point cloud to the Z axis, and using the error determination method to obtain the error value; The specific process is: S6.1, projecting the converted point cloud image data of the three-dimensional coordinate system O1 into the two-dimensional coordinate system XO1Y after removing the Z-axis height information; S6.2, with the reference point O1 as the center of the circle, find the farthest point P1 and the closest point P2 from the center O1 on the point cloud, draw circles with R1 and R2 as radii respectively, and obtain two concentric circles F1 and F2; S6.3. Calculate the final radial total runout error value t according to equations (6) and (7); (6); t = R2 - R1 (7); In the formula, R 1= O1P1, R2=O1P2; (Xi, Yj) is a point on the point cloud; like Fig.15 As shown, the circumferential data of the 10 sampled circles are uniformly represented in the axis coordinate system, the total runout value is calculated, and based on the total runout tolerance of the workpiece design, it is determined that the circular runout of the workpiece exceeds the tolerance, and the tolerance value is 0.0092mm.

Claims

1. A method for measuring the radial total runout error of shaft parts based on image processing, characterized in that: The following steps are involved: S1, collecting images of the parts to be tested; S2. According to the image of the measured part, an initial point cloud image based on the three-dimensional coordinate system O1 is established; S3, in the initial point cloud image, using a fitting algorithm to determine the axis of the image; S4. Establish a new three-dimensional coordinate system O with the axis as the coordinate axis e ; S5, establish three-dimensional coordinate system O1 and O e The transformation matrix T between them will be based on the three-dimensional coordinate system O e The point cloud image data in is converted into point cloud image data in the three-dimensional coordinate system O1 through the transformation matrix T; S6. Calculate the distance from the midpoint of the point cloud to the Z axis, and use the error determination method to obtain the error value.

2. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 1 is characterized in that: The specific process of S1 is as follows: S1.

1. Determine the shaft section of the part to be measured for measuring the total radial runout error; S1.

2. Define the common axis of the two ends A and B of the shaft segment of the measured part as the reference; S1.

3. Collect the cross-sectional circular image information of the section A, the section B and the sampling area W at the shaft section of the measured part between the sections A and B respectively.

3. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 2 is characterized in that: The process of collecting the cross-sectional circular image information of the sampling area W in S1.3 is as follows: when collecting the image at the W position, interval sampling is performed, and the interval sampling is performed along the direction from the reference B to the reference A at the L2 position of the reference B section. Collect information of several cross-sectional circles.

4. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 3 is characterized in that: The specific process of S2 is: S2.

1. Project the contour information of the measured part by telecentric lens imaging method; collect Q+1 images of the contour information at a prescribed interval; S2.2, collect the projection information of the entire circumference of the part by rotating the camera along the axis of the part to be measured; S2.3, extracting pixel information at the cross-section circle; The pixel position information of the pixel information is j, where j=1.2, ..., Q+1; the number of pixel rows of the pixel information is i, and the number of pixels in each row of pixel information is set to R i ; S2.

4. Determine pixel center O p , according to the pixel center O p Create a polar coordinate circle for the cross-section contour; set the center of the polar coordinate circle O p The distance between distance O is X1; S2.

5. Combined with motion control, add Z-axis height information to the cross-sectional profile polar coordinate circle according to the corresponding distance, and finally display the collected data in the three-dimensional coordinate system O1; where L = ab, L2 = eb, and fe is the width of the sampling area W.

5. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 4 is characterized in that: The specific process of S3 is as follows: S3.

1. Determine the fitting circle centers Oa, Ob, and Ob of A, B and section W respectively by the least squares circle method. e With O f ; S3.2, Connect O a , O b , the straight line O a O b The extension line G is used as the reference axis for error analysis.

6. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 5 is characterized in that: The specific process of S4 is: e As the reference point, O e O f The extension line of is the Z axis, passing through O e With the XO1Z plane parallel and perpendicular to O e Z1 O e X1 is the new X axis; e , O f It is the intersection of the first cross-sectional circle and the last cross-sectional circle of the sampling area W with the reference axis G.

7. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 6 is characterized in that: The specific process of S5 is as follows: S5.

1. Calculate the angle between the reference axis and the Z axis in the O1 coordinate system. ; S5.

2. Calculate O1 and O e The transformation matrix T between the two coordinate systems; S5.3, according to the angle , will be in O e The point cloud image data under the coordinates is converted into point cloud image data of the three-dimensional coordinate system O1 through the transformation matrix T.

8. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 7 is characterized in that: The specific process of S5.2 is as follows: S5.2.

1. Assuming a direction vector is a normalized vector, i.e. , according to the direction vector, construct the antisymmetric moment according to formula (1) ; （1）； S5.2.

2. Calculate the rotation matrix using formula (2) ; （2）； In the formula, is the identity matrix, is the angle between the direction vector and the Z axis, by calculating the direction vector With the standard basis vectors The dot product gives ,Right now ,as well as ; S5.2.

3. Calculate according to formula (3) ; （3）； S5.2.

4. According to formula (4), the rotation matrix is ​​obtained ; （4）； S5.2.5, according to formula (5), the transformation matrix T is obtained; （5）。 9. The method for measuring the radial total runout error of shaft parts based on image processing according to claim 8 is characterized in that: The specific process of S6 is as follows: S6.1, projecting the converted point cloud image data of the three-dimensional coordinate system O1 into the two-dimensional coordinate system XO1Y after removing the Z-axis height information; S6.2, with the reference point O1 as the center of the circle, find the farthest point P1 and the closest point P2 from the center O1 on the point cloud, draw circles with R1 and R2 as radii respectively, and obtain two concentric circles F1 and F2; S6.

3. Calculate the final radial total runout error value t according to equations (6) and (7); （6）； t = R2 - R1 (7); In the formula, R 1= O1 P1, R2=O1 P2; (Xi, Yj) is a point on the point cloud.

Images