A measuring device and a measuring method for a segmented bearing cage

Abstract

The present invention provides a measuring device for a segmented bearing retainer, comprising a loading assembly, a conveying assembly, an image acquisition assembly, a blanking assembly, and an image processing assembly; the loading assembly places the segmented bearing retainer on the conveying assembly; the conveying assembly transports the segmented bearing retainer to the image acquisition assembly; the image acquisition assembly comprises a camera, which captures an image of the segmented bearing retainer; the image processing assembly is connected to the camera and is used to process the image captured by the camera to obtain the dimensions of the segmented bearing retainer; and the blanking assembly blanks the segmented bearing retainer. The above-mentioned measuring device is used to measure the segmented bearing retainer, including capturing an image of the retainer to be tested, processing the image by the image processing assembly, and comparing and analyzing the processing results of the current retainer to be tested with the design requirements of the retainer to be tested.

Description

Technical Field

[0001] The present invention relates to the technical field of shield machine main bearings, and in particular to a measuring device and a measuring method for a segmented bearing retainer. Background Art

[0002] Because the segmented cage is a critical component of the main bearing, 100% dimensional inspection is essential to ensure proper operation. Segmented bearing cages typically exhibit the following characteristics: they possess both linear and arc features; the arc radius is large, the segment length is short, and the result is a noticeably short arc; there are numerous pockets; and there are numerous cage sets. However, due to the relatively small market in this area, unique dimensional measurement technologies for similar structures are unavailable. Manual measurement or three-dimensional coordinate measurement are commonly used. However, switching between measurement points requires significant time, making efficient batch testing impossible. Furthermore, with short arc features, small deviations in the measurement points are amplified into large deviations in the radius using the least-squares method for fitting a circle, making it difficult to assess whether the short arc meets design requirements. Summary of the Invention

[0003] The present invention provides a measuring device for a segmented bearing retainer, comprising a loading component, a conveying component, an image acquisition component, a blanking component and an image processing component; the loading component places the segmented bearing retainer on the conveying component; the conveying component transports the segmented bearing retainer to the image acquisition component; the image acquisition component comprises a camera, which takes an image of the segmented bearing retainer; the image processing component is connected to the camera and is used to process the image taken by the camera to obtain the size of the segmented bearing retainer; the blanking component blanks the segmented bearing retainer.

[0004] In addition to the above structure, the measuring device of the bearing segmented retainer also includes an alarm component and a rejection component; the alarm component is connected to the image processing component and is used to display the processing results of the image processing component; the rejection component is connected to the image processing component and is used to reject the bearing segmented retainer from the conveying component when the image processing component feeds back that the current bearing segmented retainer it has detected is an unqualified product.

[0005] The present invention also provides a method for measuring a segmented bearing cage, comprising the following steps:

[0006] Step 1: Install the measuring device for the segmented bearing cage as described above, and place the cages to be tested one by one on the measuring device for the segmented bearing cage, so that they are moved by the transmission assembly;

[0007] Step 2: Image acquisition of the holder to be inspected: The image acquisition component captures an image of the holder to be inspected, and transmits the captured image to the image processing component;

[0008] Step 3: The image processing component processes the image;

[0009] Step 4: Compare and analyze the processing result of the current cage to be tested with the design requirements of the cage to be tested, and feed back and process the analysis result.

[0010] Optionally, in step 2, the image captured by the image acquisition component on the holder to be inspected is a color image.

[0011] Optionally, in step three, the specific process of the image processing component processing the image is: using a hole filling algorithm to divide the contour lines in the color image into straight line features and arc features, and then searching and calculating the straight line features and arc features separately to obtain the corresponding structural dimensions in the retaining frame to be detected.

[0012] Optionally, the specific process of using a hole filling algorithm to divide the contour lines in the color image into straight line features and arc features is as follows:

[0013] (1) converting the color image into a binary image, and setting the binarization value of the pocket portion of the cage to be inspected to 0 and the binarization value of the frame portion to 1, thereby obtaining a binary image I of the cage to be inspected;

[0014] (2) extracting the edge of the entire binary image I, and setting the binary value of the edge position to 1, and the binary value of the non-edge position to 0, to obtain the edge image F;

[0015] (3) Using a 3*3 matrix of all 1s as a structural element, dilation processing is performed on the edge image F to obtain an edge result after the dilation processing of the edge image F;

[0016] (4) Find the intersection of the edge result and the inverse of the binary image I to obtain the result F1.

[0017] (5) Compare F1 and F. If they are not equal, assign F1 to F and repeat steps (3) and (4) until F1 and F are equal, obtaining a hole filling result I1, which is the inverse of F1. Express the hole filling result I1 as an arc segment of the cage, and the difference I2 between I1 and I as a straight line segment of the cage.

[0018] (VI) Perform edge extraction on I1 and I2 based on the Sobel operator to obtain the arc edge point set P of the cage to be detected. a and the line edge point set P l .

[0019] Optionally, the specific process of searching and calculating the pocket features in the cage to be inspected is as follows:

[0020] (1) According to the straight line edge point set P l The distribution of each point in the line edge point set P l Perform Euclidean clustering to form a sub-point set P for each pocket l1 ,P l2 ,…,P ln , where n is the total number of sub-point sets of each pocket; and the coordinate value boundaries of each sub-point set after segmentation are obtained to obtain the area estimation value of each sub-point set region;

[0021] (2) Calculate the minimum bounding rectangle of each sub-point set respectively, and calculate the pocket width and length corresponding to the sub-point set based on its corresponding minimum bounding rectangle:

[0022] (2.1) According to the connecting lines of the diagonal points of the minimum circumscribed rectangle, find the subpoint set P l1 The corresponding coordinates of the pocket center O l1 (X ol1 ,Y ol1 ), where: X ol1 is the coordinate of the center of the pocket in the X direction, Y ol1 is the coordinate of the pocket center in the Y direction;

[0023] (2.2), according to the center coordinates O of the pocket l1 and the center of the pocket to the subpoint set P l1 The minimum distance coordinates of the minimum enclosing rectangle P(X P ,Y P ), calculate the deflection angle θ of the pocket relative to the coordinate system; and search for the pocket center O l1 To the subpoint set P l1 The minimum distance point P in the X direction of the minimum circumscribed rectangle X1 The minimum distance point P in the Y direction Y1 :

[0024]

[0025] Where: X P From the center of the pocket to the subpoint set P l1 The X coordinate of the minimum distance point P of the minimum enclosing rectangle, Y P From the center of the pocket to the subpoint set P l1 The Y coordinate of the minimum distance point P of the minimum circumscribed rectangle, the center coordinate of the pocket O l1 to P X1 The product of the distance between them and cos(θ) is the half width of the pocket hole. The center coordinate of the pocket hole is O l1 to P Y1The product of the distance between and cos(θ) is the half length of the pocket;

[0026] (2.3) Repeat steps (2.1) and (2.2) to obtain the pocket width and length values ​​corresponding to other sub-point sets.

[0027] Optionally, the process of comparing and analyzing the characteristic dimensions of the pockets in the cage to be inspected with the designed dimensions of the pockets in the cage is as follows:

[0028] I. Determining whether the obtained pocket width and length values ​​meet the design requirements of the cage to be inspected: The obtained pocket width and length values ​​are compared and analyzed with the pocket design values ​​corresponding to the cage to be inspected to confirm whether the pocket dimensions of the cage to be inspected are qualified. If qualified, the cage to be inspected is transported by the conveyor assembly to the position of the unloading assembly and unloaded. If unqualified, the computer sends corresponding instructions to the alarm assembly and the rejection assembly, respectively, to reject the cage to be inspected.

[0029] II. Determine whether the distribution of pockets meets the design requirements of the cage to be tested: Assume that the center coordinates of all sub-points in the pocket feature are O l1 ,O l2 ,…,O ln And O l1 and O ln It just falls on the theoretical distribution circle. According to O l1 and O ln The coordinates of the theoretical distribution center O' are obtained by taking the coordinates of the center of the pocket hole and the design value of the center diameter of the pocket hole. l1 ,O l2 ,…,O ln The distance between the circle center O'' coordinate obtained by the least squares method and the valid O' coordinate is less than the center diameter, confirm the valid O' coordinate; and calculate O l1 ,O l2 ,…,O ln The difference between the distance to O' and the center diameter is compared with the radius design tolerance to confirm whether the pocket distribution is qualified.

[0030] Optionally, the specific process of searching and comparing the arc features in the cage to be inspected is as follows:

[0031] A. Use the following method to search for arc features in the cage to be inspected:

[0032] A1. Based on the RANSAC algorithm, the arc edge point set P in the cage to be detected is a Perform segmentation to obtain the sub-point set P after the first arc feature search is completed a1 , Pa2 ,…,P at , and the subpoint set P a1 , P a2 ,…,P at From the arc edge point set P a Delete in;

[0033] A2. Arc edge point set P a Repeatedly search for arc features in the arc until the arc edge point set P a The arc feature can no longer be found in the search, and the arc edge point set P is obtained. a The arc sub-point sets found in the search are used to complete the search for arc features in the cage to be inspected;

[0034] B. Based on the arc sub-point set P ai The deviation between the arc feature and the corresponding theoretical value in the cage to be tested is calculated to determine whether the arc feature meets the design requirements of the cage to be tested.

[0035] Optionally, in step A1, the arc edge point set P in the cage to be detected is converted based on the RANSAC algorithm. a The specific process of segmentation is as follows:

[0036] A1-1. In an arc subpoint set P a1 Choose any three non-collinear points from the set and add them to the set to form set s1;

[0037] A1-2. Perform arc fitting on all points in set s1 to obtain arc r1;

[0038] A1-3. Calculate the P values ​​other than the points included in the set s1. a1 The distance between the point in and the arc r1 is calculated to obtain several distance values; all points with distance values ​​less than the threshold ε are added to the set to form a set s2; and the optimal fitting residual v1 is set to infinity;

[0039] A1-4. Perform arc fitting on all points in set s2 to obtain arc r2;

[0040] A1-5. Determine whether the fitting residual v2 of arc r2 is less than the current optimal fitting residual v1: If the fitting residual v2 of the arc is less than the current optimal fitting residual v1, save the point index, fitted parameters, and fitting residuals in the set s2, and use this fitting residual v2 as the optimal fitting residual for the next iteration; if the fitting residual v2 of the arc is greater than the current optimal fitting residual v1, the current optimal fitting residual v1 is still used as the optimal fitting residual for the next iteration;

[0041] A1-6. Calculate the P values ​​other than the points included in set s2. a1 The distance between the point in and the arc r2 is obtained to obtain several distance values, and all arc sub-point sets with distance values ​​less than the threshold ε are added into the set to form a set s3;

[0042] A1-7, repeat steps B4 to B6 until the number of iterations reaches the preset number, and output the arc sub-point set P with the minimum fitting residual ai ;

[0043] A1-8, due to the output arc sub-point set P ai With the minimum fitting residual, the arc subpoint set P is considered ai belong to the same arc.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention provides a measuring device for a segmented bearing cage. The device uses image analysis to quickly and accurately determine cage dimensions, thereby enabling efficient batch testing. A calculation method for short arc dimension deviations is also proposed.

[0046] In addition to the above-described objects, features and advantages, the present invention has other objects, features and advantages. The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are intended to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings:

[0048] Figure 1 is a schematic diagram of a measuring device for a segmented bearing cage according to an embodiment of the present invention;

[0049] Figure 2 yes Figure 1 Axonometric diagram of the image acquisition component;

[0050] Figure 3 yes Figure 2 A-direction local schematic diagram;

[0051] Figure 4 1 is a flow chart of a method for measuring a segmented bearing cage according to an embodiment of the present invention;

[0052] Figure 5 (a)- Figure 5 (f) is a method for measuring a segmented bearing cage according to an embodiment of the present invention, wherein the center of the pocket is searched. l1 To the subpoint set P l1The minimum distance point P in the X direction of the minimum circumscribed rectangle X1 The minimum distance point P in the Y direction Y1 Detailed process diagram;

[0053] Figure 6 It is a coordinate diagram for finding the center O' of the theoretical distribution circle of pockets in a measurement method of a segmented bearing cage in an embodiment of the present invention.

[0054] in:

[0055] 1. First robotic arm, 2. Conveyor belt, 3. Holding frame, 4. Image acquisition component, 4-1. Base, 4-2. Retractable cantilever frame, 4-3. Camera, 4-4. Lighting equipment, 5. Computer, 6. Second robotic arm, 7. Third robotic arm, 8. Warning light, 9. Pocket hole, 10. Minimum circumscribed rectangle of the pocket hole. DETAILED DESCRIPTION

[0056] In order to make the above-mentioned purposes, features and advantages of the present invention more clear and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. It should be noted that the drawings of the present invention are all simplified and non-precisely scaled, and are only used to conveniently and clearly assist in explaining the implementation of the present invention; the "numbers" mentioned in the present invention are not limited to the specific quantities in the examples in the accompanying drawings; the directions or positional relationships indicated by "front", "middle", "back", "left", "right", "up", "down", "top", "bottom", "middle", etc. mentioned in the present invention are based on the directions or positional relationships shown in the drawings of the present invention, and do not indicate or imply that the devices or components referred to must have a specific direction, nor can they be understood as limitations on the present invention.

[0057] The present embodiment:

[0058] See also Figures 1 to 3 As shown, a measuring device for a segmented bearing cage includes a first robotic arm 1, a conveyor belt 2, an image acquisition component 4, a computer 5, and a third robotic arm 7;

[0059] The first robotic arm 1, image acquisition component 4, computer 5 and third robotic arm 7 are all arranged on one side of the conveyor belt 2; the holder 3 to be inspected is loaded onto the conveyor belt 2 by the first robotic arm 1, and is transported to the position of the image acquisition component 4 via the conveyor belt 2; the image acquisition component 4 takes an image of the holder 3 to be inspected, and transmits the captured image to the computer 5 for processing to obtain the structural dimensions of the holder 3 to be inspected; at the same time, the holder 3 to be inspected continues to be transported to the first robotic arm 7 via the conveyor belt 2 to be unloaded by the third robotic arm 7.

[0060] Optionally, to ensure that the image acquisition component 4 can clearly capture the color of the holder 3 to be inspected, the color of the conveyor belt 2 and the color of the holder 3 to be inspected are set to different colors. Preferably, the color of the conveyor belt 2 can be blue to distinguish it from the metallic silver-white and yellow colors of the holder 3 to be inspected.

[0061] Optionally, the structures of the first robotic arm 1, the second robotic arm 6 and the third robotic arm 7 are consistent, and the first robotic arm 1, the second robotic arm 6 and the third robotic arm 7 all include a robotic arm body that can be displaced in all directions and a suction cup structure arranged at the driving end of the robotic arm body, and the suction cup structure is driven by the robotic arm body to absorb or place the retaining frame 3 to be detected.

[0062] Optionally, the image acquisition component 4 includes a base 4-1, a retractable cantilever frame 4-2, a camera 4-3 and a lighting device 4-4, the retractable cantilever frame 4-2 is fixedly mounted on the base 4-1, the driving end of the retractable cantilever frame 4-2 is fixedly mounted with a ring-shaped lighting device 4-4, and the center of the lighting device 4-4 is arranged with a camera 4-3 to reduce or prevent the shadow of the camera 4-3 during the shooting of the holder 3 to be inspected through the auxiliary lighting of the lighting device 4-4 during the measurement process.

[0063] Optionally, the retractable arm 4-2 is configured as a height-adjustable structure to ensure that the holder 3 to be detected can appear completely in the field of view of the camera 4-3 during the shooting process of the camera 4-3.

[0064] Optionally, since the holder 3 to be inspected has pockets and arcs whose dimensions are to be measured, to ensure that the plane of the pockets and arcs is parallel to the lens of the image acquisition assembly 4, the conveyor belt 2 is provided with several raised support structures, and the holder 3 to be inspected is placed on the raised support structures. Preferably, the raised support structures are sequentially distributed along the displacement direction of the conveyor belt 2.

[0065] In addition to the above structure, the measuring device of the bearing segmented retainer also includes a warning light 8 and a second robotic arm 6 connected to a computer 5. The computer 5 compares and analyzes the structural dimensions of the retainer 3 to be detected with the design dimensions of the retainer 3 to be detected. When the result of its analysis exceeds the design requirement range of the retainer 3 to be detected, the computer 5 sends corresponding instructions to the warning light 8 and the second robotic arm 6. The warning light 8 is displayed and the second robotic arm 6 grabs the retainer 3 to be detected to remove the retainer 3 to be detected.

[0066] See also Figure 4 As shown, the specific process of measuring the cage using the above-mentioned cage rapid measurement device is as follows:

[0067] Step 1: Install the quick measurement device for the cage, and adjust the distance between the extendable cantilever frame and the conveyor belt in the image acquisition component to ensure that the camera can clearly capture the cage to be inspected;

[0068] Step 2: The holder to be inspected is grabbed by the first robotic arm and transported to the conveyor belt, and then transported to the position of the image acquisition component via the conveyor belt;

[0069] Step 3: The image acquisition component takes a color image of the holder to be inspected and transmits the taken color image to a computer;

[0070] Step 4: The computer processes the color image of the current cage to be inspected;

[0071] Step 5: Compare and analyze the processing result of the current cage to be tested with the design requirements of the cage to be tested, and feed back and process the analysis result.

[0072] Optionally, the method for processing the color image is as follows: using a hole filling algorithm to divide the contour lines in the color image into straight line features and arc features, and searching and calculating the pocket features and arc features in the retaining frame to be inspected respectively to obtain the corresponding structural dimensions of the retaining frame to be inspected.

[0073] Optionally, the specific process of using a hole filling algorithm to separate the contour lines in the color image into straight line features and arc features is as follows:

[0074] (1) Since the features of the cage to be inspected that require structural dimension calculation only include the pockets (i.e., the holes) and the arcs (since the cage to be inspected is a segmented structure, the straight line connecting the two arcs does not need to be inspected), the color image of the cage to be inspected is converted into a binary image, and the binarization values ​​of the portions of the binarized image of the cage to be inspected that are consistent with the background color are set to 0 (i.e., the binarization values ​​of the pockets of the cage to be inspected and the background are both set to 0), and the binarization values ​​of the portions of the binarized image of the cage to be inspected that are inconsistent with the background color are set to 1 (i.e., the binarization value of the frame portion of the cage to be inspected is set to 1), thereby obtaining a binary image I;

[0075] (2) extracting the edge of the entire binary image I, and setting the binary value of the edge position to 1, and the binary value of the non-edge position to 0, to obtain the edge image F;

[0076] (3) Using a 3*3 matrix of all 1s as a structural element, dilation processing is performed on the edge image F to obtain an edge result after the dilation processing of the edge image F;

[0077] (4) Find the intersection of the edge result and the inverse of the binary image I to obtain the result F1.

[0078] (5) Compare F1 and F. If they are not equal, assign F1 to F and repeat steps (3) and (4) until F1 and F are equal, obtaining the hole filling result I1 (the hole filling result I1 is the inverse of F1). The hole filling result I1 is expressed as the arc segment of the cage, and the difference I2 between I1 and I is expressed as the straight line segment of the cage.

[0079] (VI) Perform edge extraction on I1 and I2 based on the Sobel operator (in the process of performing the Sobel operator on I1 and I2, the threshold is set to 0.15), and obtain the arc edge point set P of the cage to be detected a and the line edge point set P l ;

[0080] (7) Further based on the Zernike moment sub-pixel edge detection algorithm (threshold selection is based on Otsu adaptive threshold method), the edge contour size is refined from pixel level to sub-pixel level to improve the resolution of the detection algorithm.

[0081] Optionally, the specific process of searching and calculating the pocket features in the cage to be inspected is as follows:

[0082] (1) According to the straight line edge point set P l The distribution of each point in the line edge point set P l Perform Euclidean clustering segmentation (the segmentation threshold depends on the product of the minimum spacing between each pocket and the conversion factor between the pixel point and the length unit, where the conversion factor is set as the ratio between the pixel size and the actual size) to form a sub-point set P for each pocket. l1 ,P l2 ,…,P ln , where n is the total number of sub-point sets in each pocket; and the coordinate value boundaries of each sub-point set after segmentation are obtained to obtain the estimated area value of each sub-point set region; if the estimated area value of the region corresponding to each sub-point set is less than 1 / 10 of the average sum of the area of ​​all sub-point sets, the region corresponding to the sub-point set is determined to be a noise point and is removed;

[0083] (2) Calculate the minimum bounding rectangle of each sub-point set respectively (calculating the corresponding minimum bounding rectangle of each sub-point set is a prior art in this field), and calculate the pocket width and length corresponding to the sub-point set based on the corresponding minimum bounding rectangle:

[0084] (2.1) According to the connecting line of the diagonal points of the minimum circumscribed rectangle (the center of the connecting line of the diagonal points is the center of the pocket), find the subpoint set P l1The corresponding coordinates of the pocket center O l1 (X ol1 ,Y ol1 ), where: X ol1 is the coordinate of the center of the pocket in the X direction, Y ol1 is the coordinate of the pocket center in the Y direction;

[0085] (2.2), according to the center coordinates O of the pocket l1 and the center of the pocket to the subpoint set P l1 The minimum distance coordinates of the minimum enclosing rectangle P(X P ,Y P ), calculate the deflection angle θ of the pocket relative to the coordinate system; and search for the pocket center O l1 To the subpoint set P l1 The minimum distance point P in the X direction of the minimum circumscribed rectangle X1 The minimum distance point P in the Y direction Y1 (For detailed search process, see Figure 5 shown):

[0086]

[0087] Where: X P From the center of the pocket to the subpoint set P l1 The minimum distance in the X direction of the minimum enclosing rectangle, P From the center of the pocket to the subpoint set P l1 The minimum distance in the Y direction of the minimum circumscribed rectangle, the center coordinate of the pocket O l1 to P X1 The product of the distance between and cos(θ) is the half-width of the pocket (the half-width of the pocket = 1 / 2 of the pocket width), and the center coordinate of the pocket is O l1 to P Y1 The product of the distance between them and cos(θ) is the half length of the pocket (half length of the pocket = 1 / 2 of the pocket length);

[0088] (2.3) Repeat steps (2.1) and (2.2) to obtain the pocket width and length values ​​corresponding to other sub-point sets.

[0089] This pocket feature calculation method avoids point set segmentation and segmented fitting in a single pocket, greatly reducing the amount of calculation.

[0090] Optionally, the process of comparing and analyzing the characteristic dimensions of the pockets in the cage to be inspected with the designed dimensions of the pockets in the cage is as follows:

[0091] I. Determine whether the obtained pocket width and length values ​​meet the design requirements of the cage to be inspected: Compare and analyze the obtained pocket width and length values ​​with the pocket design values ​​corresponding to the cage to be inspected to confirm whether the pocket size of the cage to be inspected is qualified; if qualified, the cage to be inspected continues to run from the conveyor belt to the position of the third robotic arm for unloading; if unqualified, the computer sends corresponding instructions to the warning light and the second robotic arm, and the second robotic arm sucks the cage to be inspected out of the conveyor belt to achieve the removal of the cage to be inspected.

[0092] II. Determine whether the distribution of pockets meets the design requirements of the cage to be tested: Assume that the center coordinates of all sub-points in the pocket feature are O l1 ,O l2 ,…,O ln And O l1 and O ln It just falls on the theoretical distribution circle. According to O l1 and O ln The coordinates of the theoretical distribution circle center O' are obtained by taking the design value of the center diameter of the pocket hole distribution as an example (there are 2 solutions); and the coordinates of the center of the pocket hole O are obtained by taking the design value of the center diameter of the pocket hole distribution as an example (there are 2 solutions). l1 ,O l2 ,…,O ln The distance between the fitting circle center O″ coordinate obtained by the least square method and the effective O′ coordinate is less than the center diameter. Confirm the effective O′ coordinate and calculate O l1 ,O l2 ,…,O ln The difference between the distance to O' and the center diameter is compared with the radius design tolerance to confirm whether the pocket distribution is qualified (the coordinates of the theoretical distribution center O' are shown in the table). Figure 6 This judgment method does not rely on the least squares method, so it will not amplify the deviation of the measurement point and can better describe the actual measurement results.

[0093] Furthermore, let the center coordinate of the pocket hole be O l1 The coordinate values ​​in the X and Y directions are (X1, Y1), the center coordinate of the pocket is O ln The coordinate values ​​in the X and Y directions are (X n ,Y n ), the design value of the center diameter of the pocket distribution is R, then the coordinate values ​​(X′, Y′) of the theoretical distribution center O′ of the pocket in the X and Y directions are solved as follows:

[0094] ①Calculate O l1 and O ln The distance D between:

[0095]

[0096] Among them: X1 is the subpoint set P l1 The X-direction coordinate of the center of the pocket, Y1 is the sub-point set P l1 The Y coordinate of the center of the pocket hole, X n is the subpoint set P ln The X-axis coordinates of the pocket center, Y n is the subpoint set P ln The Y coordinate of the pocket center;

[0097] Calculate the theoretical distribution from the center O' to the chord Distance d:

[0098]

[0099] When X1≠X n And Y1≠Y n When , the 2 solutions of O′(X′,Y′) can be expressed as:

[0100] or

[0101]

[0102] Where: α corresponds to the chord The inclination angle is , X′ is the coordinate of the theoretical distribution center O′ in the X direction, and Y′ is the coordinate of the theoretical distribution center O′ in the Y direction;

[0103] When X1=X n When , the 2 solutions of O′(X′,Y′) can be expressed as:

[0104] or

[0105] When Y1=Y n When , the 2 solutions of O′(X′,Y′) can be expressed as:

[0106] or

[0107] Combined criteria Invalid solutions can be deleted.

[0108] Optionally, the specific process of searching and comparing the arc features in the cage to be inspected is as follows:

[0109] A. Use the following method to search for arc features in the cage to be inspected:

[0110] A1. Based on the RANSAC algorithm, the arc edge point set P in the cage to be detected is aTo split, since the arc edge point set P a There is more than one arc in the set P. Therefore, when the arc edge point set P is first a After the arc feature search is completed, all the arc edge point sets found are saved as sub-point set P a1 , P a2 ,…,P at , and the subpoint set P a1 , P a2 ,…,P at From the arc edge point set P a Delete in;

[0111] A2. Arc edge point set P a Repeatedly search for arc features in the arc until the arc edge point set P a No arc features can be found in the search, and the arc sub-point set P is obtained. a1 ,P a2 ,…,P am , where: t≤m, m is the total number of arc sub-point sets; complete the search for arc features in the cage to be detected.

[0112] Optionally, in step A1, the arc edge point set P in the cage to be detected is converted based on the RANSAC algorithm. a The specific process of segmentation is as follows:

[0113] A1-1. Point set P on the arc edge a Pick any three points (pick any three points in the arc edge point set and check whether the three points are collinear. If the three points are collinear, replace the third point with the remaining points until the three points are not collinear, and add the three points to the set) to determine whether the points are collinear. If the three points are collinear, add the three points to the arc edge point set P. a Reselect until the three selected points are not collinear with each other, and add the three points finally selected to the set to form set s1;

[0114] A1-2. Perform arc fitting on the points in set s1 to obtain arc r1;

[0115] A1-3. Calculate the P values ​​other than the points included in the set s1. a The distance between the point in and the arc r1 is obtained to obtain several distance values; all points with distance values ​​less than the threshold ε are added to the set to form a set s2; and the best fit residual v1 is set to infinity (the best fit residual v1 refers to the difference between the minimum fit value of the arc r1 and the actual measured value);

[0116] A1-4. Perform arc fitting on all points in set s2 to obtain arc r2;

[0117] A1-5. Determine whether the fitting residual v2 of arc r2 is less than the current optimal fitting residual v1: If the fitting residual v2 of the arc is less than the current optimal fitting residual v1, save the point index, fitted parameters, and fitting residuals in the set s2, and use this fitting residual v2 as the optimal fitting residual for the next iteration; if the fitting residual v2 of the arc is greater than the current optimal fitting residual v1, the current optimal fitting residual v1 is still used as the optimal fitting residual for the next iteration;

[0118] A1-6. Calculate the P values ​​other than the points included in set s2. a The distance between the point in and the arc r2 is obtained to obtain several distance values, and all arc sub-point sets with distance values ​​less than the threshold ε are added into the set to form a set s3;

[0119] A1-7. Repeat steps B4 to B6 until the number of iterations reaches the preset number, and output the arc sub-point set with the minimum fitting residual;

[0120] A1-8. Since the output arc sub-point set has the minimum fitting residual, it can be considered to belong to the same arc. Let this sub-point set be P ai .

[0121] B. Use the following process to determine whether the arc feature in the cage to be tested meets the design requirements of the cage to be tested:

[0122] According to the subpoint set P ai Sort the slopes of the points in the graph to point (0, 0) and find the set of points P that contains the points ai The two endpoints of the arc, assuming that these two endpoints fall on the theoretical circle, combine the theoretical circle radius to find the theoretical center O 1 , according to P ai The least squares method is used to fit the center of the circle O 2 and the theoretical center O 1 and the fitting circle center O 2 The distance between them is less than the theoretical circle radius, so the theoretical center O can be clearly determined. 1 (Find the theoretical center O 1 and the theoretical center O 1 and the fitting circle center O 2 The distance between the two holes is calculated by referring to the theoretical distribution center O' and O l1 and O ln The calculation process of the distance between

[0123] Calculate P ai From the point in the circle to the theoretical center O 1The difference between the distance and the theoretical circle radius is calculated, and the difference is compared with the radius design tolerance to confirm whether the arc feature is qualified: if it is determined that the arc feature of the holder to be inspected does not meet its design requirements, the computer will send a warning display signal to the warning light and a suction signal to the second robot arm to remove the holder to be inspected from the conveyor belt; if it is determined that the arc feature of the holder to be inspected meets its design requirements, the conveyor belt continues to transport the holder to be inspected to the suction position of the third robot arm, so that the third robot arm can unload the holder to be inspected and complete the inspection of the holder to be inspected.

[0124] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for measuring a segmented bearing cage, characterized in that: The following steps are involved: Step 1: Install the measuring device for the segmented bearing cage, and place the cages to be tested one by one on the measuring device for the segmented bearing cage, so that they are displaced by the driving of the transmission assembly; Step 2: Image acquisition of the holder to be inspected: The image acquisition component captures an image of the holder to be inspected, and transmits the captured image to the image processing component; The image captured by the image acquisition component on the holder to be inspected is a color image; Step 3: The image processing component processes the image; The specific process of image processing by the image processing component is as follows: the contour lines in the color image are divided into straight line features and circular arc features using a hole filling algorithm, and then the straight line features and circular arc features are searched and calculated separately to obtain the corresponding structural dimensions of the cage to be inspected; The specific process of searching and calculating the pocket features in the cage to be inspected is as follows: (1) According to the line edge point set P l The distribution of each point in the line edge point set P l Perform Euclidean clustering to form a sub-point set P for each pocket l1 ,P l2 ,…,P ln , where n is the total number of sub-point sets of each pocket; and the coordinate value boundaries of each sub-point set after segmentation are obtained to obtain the area estimation value of each sub-point set region; (2) Calculate the minimum bounding rectangle of each sub-point set respectively, and calculate the pocket width and length corresponding to the sub-point set based on its corresponding minimum bounding rectangle: (2.1) According to the connecting lines of the diagonal points of the minimum circumscribed rectangle, find the subpoint set P l1 The corresponding coordinates of the pocket center O l1 (X ol1 ,Y ol1 ), where: X ol1 is the coordinate of the center of the pocket in the X direction, Y ol1 is the coordinate of the pocket center in the Y direction; (2.2), according to the center coordinates O of the pocket l1 and the center of the pocket to the subpoint set P l1 The minimum distance coordinates of the minimum enclosing rectangle P(X P ,Y P ), calculate the deflection angle θ of the pocket relative to the coordinate system; and search for the pocket center O l1 To the subpoint set P l1 The minimum distance point P in the X direction of the minimum circumscribed rectangle X1 The minimum distance point P in the Y direction Y1 : Where: X P From the center of the pocket to the subpoint set P l1 The X coordinate of the minimum distance point P of the minimum enclosing rectangle, Y P From the center of the pocket to the subpoint set P l1 The Y coordinate of the minimum distance point P of the minimum circumscribed rectangle, the center coordinate of the pocket O l1 to P X1 The product of the distance between them and cos(θ) is the half width of the pocket hole. The center coordinate of the pocket hole is O l1 to P Y1 The product of the distance between and cos(θ) is the half length of the pocket; (2.3) Repeat steps (2.1) and (2.2) to obtain the pocket width and length values ​​corresponding to other sub-point sets; Step 4: Compare and analyze the processing result of the current cage to be tested with the design requirements of the cage to be tested, and provide feedback and process the analysis results; The measuring device for the segmented bearing retainer comprises a loading component, a transmission component, an image acquisition component (4), a unloading component, an image processing component, an alarm component and a rejection component; The loading assembly places the bearing segmented cage on the conveying assembly; The conveying assembly transports the bearing segment retainer to the image acquisition assembly; The image acquisition component (4) comprises a camera (4-3), and the camera (4-3) captures an image of the bearing segmented retainer; The image processing component is connected to the camera (4-3) and is used to process the image captured by the camera (4-3) to obtain the size of the bearing segmented cage; The blanking assembly blanks the segmented bearing cage; The alarm component is connected to the image processing component and is used to display the processing results of the image processing component; The rejection component is connected to the image processing component and is used to reject the bearing segmented retainer out of the conveying component when the image processing component feeds back that the current bearing segmented retainer it has detected is an unqualified product.

2. The method for measuring a segmented bearing cage according to claim 1, characterized in that: The specific process of using the hole filling algorithm to divide the contour lines in the color image into straight line features and arc features is as follows: (1) converting the color image into a binary image, and setting the binarization value of the pocket portion of the cage to be inspected to 0 and the binarization value of the frame portion to 1, thereby obtaining a binary image I of the cage to be inspected; (2) extracting the edge of the entire binary image I, and setting the binary value of the edge position to 1, and the binary value of the non-edge position to 0, to obtain the edge image F; (3) Using a 3*3 matrix of all 1s as a structural element, dilation processing is performed on the edge image F to obtain an edge result after the dilation processing of the edge image F; (4) Find the intersection of the edge result and the inverse of the binary image I to obtain the result F1. (5) Compare F1 and F. If they are not equal, assign F1 to F and repeat steps (3) and (4) until F1 and F are equal, obtaining a hole filling result I1, which is the inverse of F1. Express the hole filling result I1 as an arc segment of the cage, and the difference I2 between I1 and I as a straight line segment of the cage. (VI) Perform edge extraction on I1 and I2 based on the Sobel operator to obtain the arc edge point set P of the cage to be detected. a and the line edge point set P l .

3. The method for measuring a segmented bearing cage according to claim 1, characterized in that: The specific process of comparing and analyzing the characteristic dimensions of the pockets in the cage to be inspected with the design dimensions of the pockets in the cage is as follows: I. Determining whether the obtained pocket width and length values ​​meet the design requirements of the cage to be inspected: The obtained pocket width and length values ​​are compared and analyzed with the pocket design values ​​corresponding to the cage to be inspected to confirm whether the pocket dimensions of the cage to be inspected are qualified. If qualified, the cage to be inspected is transported by the conveyor assembly to the position of the unloading assembly and unloaded. If unqualified, the computer sends corresponding instructions to the alarm assembly and the rejection assembly, respectively, to reject the cage to be inspected. II. Determine whether the distribution of pockets meets the design requirements of the cage to be tested: Assume that the center coordinates of all sub-points in the pocket feature are O l1 ,O l2 ,…,O ln And O l1 and O ln It just falls on the theoretical distribution circle. According to O l1 and O ln The coordinates of the theoretical distribution center O' are obtained by taking the coordinates of the center of the pocket hole and the design value of the center diameter of the pocket hole. l1 ,O l2 ,…,O ln The distance between the circle center O'' coordinate obtained by the least squares method and the valid O' coordinate is less than the center diameter, confirm the valid O' coordinate; and calculate O l1 ,O l2 ,…,O ln The difference between the distance to O' and the center diameter is compared with the radius design tolerance to confirm whether the pocket distribution is qualified.

4. The method for measuring a segmented bearing cage according to claim 2, characterized in that: The specific process of searching and comparing the arc features in the cage to be inspected is as follows: A. Use the following method to search for arc features in the cage to be inspected: A1. Based on the RANSAC algorithm, the arc edge point set P in the cage to be detected is a Perform segmentation to obtain the sub-point set P after the first arc feature search is completed a1 , P a2 ,…,P at , and the subpoint set P a1 , P a2 ,…,P at From the arc edge point set P a Delete in; A2. Arc edge point set P a Repeatedly search for arc features in the arc until the arc edge point set P a The arc feature can no longer be found in the search, and the arc edge point set P is obtained. a The arc sub-point sets found in the search are used to complete the search for arc features in the cage to be inspected; B. Based on the arc sub-point set P ai The deviation between the arc feature and the corresponding theoretical value in the cage to be tested is calculated to determine whether the arc feature meets the design requirements of the cage to be tested.

5. The method for measuring a segmented bearing cage according to claim 4, characterized in that: In step A1, based on the RANSAC algorithm, the arc edge point set P in the cage to be detected is a The specific process of segmentation is as follows: A1-1. In an arc subpoint set P a1 Choose any three non-collinear points from the set and add them to the set to form set s1; A1-2. Perform arc fitting on all points in set s1 to obtain arc r1; A1-3. Calculate the P values ​​other than the points included in the set s1. a1 The distance between the point in and the arc r1 is calculated to obtain several distance values; all points with distance values ​​less than the threshold ε are added to the set to form a set s2; and the optimal fitting residual v1 is set to infinity; A1-4. Perform arc fitting on all points in set s2 to obtain arc r2; A1-5. Determine whether the fitting residual v2 of arc r2 is less than the current optimal fitting residual v1: If the fitting residual v2 of the arc is less than the current optimal fitting residual v1, save the point index, fitted parameters, and fitting residuals in the set s2, and use this fitting residual v2 as the optimal fitting residual for the next iteration; if the fitting residual v2 of the arc is greater than the current optimal fitting residual v1, the current optimal fitting residual v1 is still used as the optimal fitting residual for the next iteration; A1-6. Calculate the P values ​​other than the points included in the set s2. a1 The distance between the point in and the arc r2 is obtained to obtain several distance values, and all arc sub-point sets with distance values ​​less than the threshold ε are added into the set to form a set s3; A1-7, repeat steps B4 to B6 until the number of iterations reaches the preset number, and output the arc sub-point set P with the minimum fitting residual ai ; A1-8, due to the output arc sub-point set P ai With the minimum fitting residual, the arc subpoint set P is considered ai belong to the same arc.

Images