Aviation ball bearing sorting system and sorting method
Through machine vision technology combined with neural network and Hough circle transformation algorithm, the problem of inconsistent detection of steel ball diameters during bearing assembly is solved, and the precision measurement and sorting of steel balls embedded in bearings is realized, which improves detection efficiency and accuracy, and ensures the bearings pass rate.
Patent Information
- Application Number
- CN202211150028.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-21
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The problem of inconsistent diameter of steel balls in the prior art is difficult to detect and sort out during bearing assembly, especially when the steel ball is embedded between the inner ring, cage, and outer ring of the bearing, conventional contact measurement methods cannot achieve losslessness. Precision measurement leads to the generation of unqualified bearing finished products.
Using an embedded steel ball ruler diameter precision measurement system based on machine vision, the image acquisition device and bearing positioning device are combined with R2U-Net++ neural network and improved Hough circle transformation algorithm to realize lossless precision measurement of steel ball diameter and bearing separation. Pick.
The precision measurement accuracy of the diameter of the inner steel ball in the bearing is 0.001mm, avoiding errors caused by manual participation, improving detection efficiency and sorting accuracy, and ensuring the passing rate of the bearing.
Smart Images

Figure CN115646841B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of machine vision and bearing detection, and in particular to an aviation ball bearing sorting system and a sorting method. Background Art
[0002] Aircraft bearings are one of the important parts of the aviation industry. They work under high speed, high load and high temperature conditions. Their health status directly affects the safety of aircraft and even the life safety of pilots. It is necessary to control the quality of aircraft bearings to improve the qualified rate of bearings. Steel balls are the most important parts of bearings and the most used rolling elements in the bearing industry. According to Hertz's point contact elasticity theory: when steel balls with large diameter differences are installed in a set of bearings, it will directly affect the bearing's accuracy, motion performance, and service life.
[0003] During the transportation, storage, distribution and assembly of bearing steel balls, steel balls of different diameters and batches may be mixed together. During the assembly process, mixed steel balls are assembled into the bearings, and the consistency of the steel ball size cannot be guaranteed, resulting in unqualified bearing products. Especially in the aviation field, when the bearing parts fail, other host parts will be damaged in a chain reaction, causing incalculable catastrophic consequences.
[0004] According to the current testing methods of finished bearings by manufacturers, there is no specific method to deal with the very hidden problem of mis-installation of steel balls. There are two reasons: First, before the final assembly of the bearing, the bearing manufacturer orders the required specifications of steel balls from the steel ball manufacturer and inspects the goods by random sampling. It takes a lot of manpower and material resources to fully test the size of all steel balls, which is not realistic. Even if the full inspection is completed before assembly, it cannot guarantee whether there will be mixed balls during the circulation process; second, after the final assembly of the bearing, the steel balls are in the confined space of the inner ring, cage, and outer ring of the bearing, and the existing steel ball diameter detection methods cannot complete the detection task.
[0005] Based on this, those skilled in the art are committed to providing a steel ball detection device and method for aviation ball bearings. Summary of the invention
[0006] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is that the steel ball is embedded between the inner ring, retaining frame and outer ring of the aviation rolling bearing. The spatial size of the arc contour of the steel ball is measured in millimeters. Conventional contact measurement methods are difficult to complete non-destructive measurement of the steel ball size, and it is impossible to determine whether the bearing assembly is installed with the wrong ball. A system and method for precise measurement of the diameter of an embedded steel ball ruler based on machine vision are proposed, which realizes non-destructive precision measurement of the steel ball diameter and sorting of bearings.
[0007] To achieve the above object, the present invention provides an aviation ball bearing sorting system on one hand, wherein the aviation ball bearing comprises a steel ball, a track and a transfer device, wherein the track is provided with a fixed entrance station, a detection station, a normal product outlet and a defective product outlet in sequence;
[0008] The indexing device includes at least one push rod and at least one push assembly, one end of the push rod extends to the upper end of the track to push the aviation ball bearing to move on the track, and the push assembly is used to push the aviation bearing on the track to the normal product outlet or the defective product outlet;
[0009] The inspection station is provided with an image acquisition device and a bearing positioning device. The image acquisition device includes a camera, the camera is located above the track, and the bearing positioning device is located below the camera, so that each steel ball in the aviation ball bearing located at the inspection station is positioned directly below the camera.
[0010] Furthermore, a receiving hole extending vertically through the track at the detection station is provided, and the bearing positioning device includes a rotating platform, which is located in the receiving hole, and the upper surface of the rotating platform is coplanar with the bottom surface of the track.
[0011] Furthermore, a plane mirror is provided on the upper surface of the rotating platform.
[0012] Furthermore, the bearing positioning device includes a positioning assembly, and a vertically penetrating avoidance hole is provided at the center of the rotating platform, and the positioning assembly is located in the avoidance hole. The positioning assembly includes a vertical pole, a rotating rod, a resistance rod and a telescopic cylinder. The vertical pole is vertically arranged, one end of the rotating rod is hingedly connected to the top end of the vertical pole, the other end of the rotating rod is fixedly connected to the resistance rod, the fixed end of the telescopic cylinder is installed on the vertical pole, and the telescopic end of the telescopic cylinder is fixedly connected to the middle part of the rotating rod.
[0013] Furthermore, the indexing device includes a power mechanism, and the output end of the power mechanism drives the push rod to move.
[0014] Furthermore, bristles are provided at the lower end of the push rod located at the upper end of the track.
[0015] Furthermore, the image acquisition device includes a camera bracket, a coaxial light source and a photoelectric switch, the camera is fixed on the camera bracket, the camera lens is located above the detection station, and the photoelectric switch is installed on the camera bracket.
[0016] A second aspect of the present invention provides an aviation ball bearing sorting method, which is implemented based on the aviation ball bearing sorting system and includes the following steps:
[0017] S1, collect photos of the left and right sides of the contours of one of the steel balls in the rolling bearing;
[0018] S2, after filtering and denoising the image, use the R2U-Net++ neural network to perform image segmentation;
[0019] S3, extracting edges of the segmented image and fitting circles using an improved Hough circle transform algorithm;
[0020] S4, calculating the diameter of the fitted circle of the same steel ball image and taking the average value as the diameter of the embedded steel ball;
[0021] S5. Repeat steps S1-S4 until the diameters of all the steel balls of the aviation ball bearing are detected. If all the steel balls of the aviation ball bearing to be detected meet the requirements, the transfer device will push the aviation ball bearing that has been tested to the normal product outlet. Otherwise, the aviation ball bearing that has been tested will be pushed to the defective product outlet.
[0022] Compared with the prior art, the present invention has the following beneficial effects:
[0023] 1. In the sorting system of the present invention, the bearings to be sorted move on a circular track to realize assembly line bearing sorting. The bearings are pushed to the corresponding outlet according to the test results. No human intervention is required throughout the process, avoiding the problem of sorting errors caused by human physiological and psychological factors, and effectively improving the detection efficiency and sorting accuracy.
[0024] 2. The sorting method of the present invention adopts the method of imaging the single-sided contour of the local steel ball and then fuses the images using an image fusion algorithm. This method reduces the imaging field of view and improves the pixel accuracy under the condition of a certain camera resolution, thereby improving the measurement accuracy to 0.001mm, which is sufficient to identify unqualified bearings with misinstalled steel balls.
[0025] 3. Since the fully assembled bearing ball is severely obscured, the part that can be imaged and measured through the coaxial light source is a 0.5mm gap. This application separates the contour of the same ball into images, then segments them separately, fits the arcs separately to calculate the diameter, and then takes the average value. This process can effectively utilize the imaging part, improve the pixel accuracy under the 8-pixel size limit of the industrial camera, and obtain high-quality original image information for subsequent image segmentation and circle fitting. The method of directly taking a picture of the entire steel ball to calculate the diameter cannot achieve the required 1μm system detection accuracy through theoretical calculation under the camera pixel limit. The method of this application improves the detection accuracy under the current mainstream industrial camera pixel size limit.
[0026] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1It is a schematic diagram of the rectification structure of an aviation ball bearing sorting system according to a specific embodiment of the present invention;
[0028] Figure 2 It is a structural schematic diagram of a push rod and a pushing assembly of a transfer device according to a specific embodiment of the present invention;
[0029] Figure 3 It is a structural schematic diagram of an imaging device and a bearing positioning device according to a specific embodiment of the present invention;
[0030] Figure 4 is a schematic diagram of U-Net++ according to a specific embodiment of the present invention, Figure 4 a is a schematic diagram of the circular convolution unit structure. Figure 4 b is a schematic diagram of the structure of the convolution unit. Figure 4 c is the structural principle diagram of U-Net++. DETAILED DESCRIPTION
[0031] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0032] It should be noted that the illustrations provided in the following embodiments are only used to illustrate the basic concept of the present invention in a schematic manner, and thus the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.
[0033] While some exemplary embodiments of the present invention have been described for the purpose of illustration, it should be understood that the present invention may be implemented in other ways not specifically shown in the drawings.
[0034] Aviation ball bearings include an inner ring, a retaining frame, an outer ring and a steel ball. The steel ball is embedded between the inner ring, retaining frame and outer ring of the aviation rolling bearing. The space size of the arc contour of the steel ball is measured in millimeters. Conventional contact measurement methods are difficult to complete non-destructive measurement of the steel ball size and cannot determine whether the bearing assembly is installed with the wrong ball. For this reason, in a specific embodiment, a sorting system for precise measurement of the diameter of an embedded steel ball ruler based on machine vision and a sorting method based on the system are proposed, which realizes non-destructive precision measurement of the ball diameter and sorting of bearings.
[0035] like Figure 1As shown, an aviation ball bearing sorting system includes a track 5 and a transfer device, wherein the track 5 is provided with a fixed entrance station 1, a detection station 2, a normal product outlet 3 and a defective product outlet 4 in sequence;
[0036] The shape of the track 5 can be set as needed, for example, it can be a circular track, a linear track, an arc track, an S-shaped track, etc., and its laying is designed according to the specific environment and location, etc. The track in this embodiment is a circular track. The width of the track is slightly larger than the major diameter of the bearing, the side height of the track is 2 / 3 of the thickness of the bearing, and the bottom surface of the track is a horizontal plane.
[0037] like Figure 2 As shown, the transfer device is used to push the aviation ball bearings on the track 5 to the detection station 2, and push the detected aviation bearings to different outlets for sorting according to the detection results; the transfer device includes at least one push rod 6 and at least one push assembly 7, one end of the push rod 6 extends to the upper end of the track 5 to push the aviation ball bearings to move on the track 5, and the number and movement trajectory of the push rods 6 can be set according to actual conditions, such as Figure 1 As shown, the transfer device of this embodiment includes a power mechanism 19 and four push rods 6. One ends of the four push rods 6 are fixed together, and the other ends extend in different directions to form a "cross-shaped" structure. The power mechanism 19 is located at the center of the circular track 5. The power mechanism 19 adopts a rotating motor 18. The output end of the rotating motor is fixedly connected to the fixed points of the four push rods 6. During the working process, the power mechanism 19 rotates and drives the four push rods 6 to transfer at a fixed point.
[0038] In the specific use process, the number of the push rods 6 can be changed. For example, increasing the number of push rods 6 can form a "M"-shaped structure, and reducing the number of push rods 6 can form a "V"-shaped structure. The push rods 6 can also be installed at two or more output ends of the power mechanism 19 as needed, dividing the track 5 into several areas, and each area completes the pushing of the bearing through the corresponding push rod 6. The position of the power mechanism 19 can also be moved as needed, and the power mechanism 19 can be set outside the circular track 5. When the track 5 is a linear track 5, the power mechanism 19 can be a telescopic cylinder 23, a telescopic hydraulic rod or a telescopic screw structure, which drives the push rod 6 to move along the linear track 5.
[0039] The pushing group can use a hydraulic cylinder or a pneumatic cylinder as power, and the number of pushing components 7 can be set as needed. In this embodiment, the pushing components 7 include four, which are respectively installed beside the push rod 6. When the push rod 6 pushes the bearing to the corresponding outlet, the pushing component 7 moves to push the bearing to the outlet; in the specific application process, the installation position of the pushing component 7 can be set as needed, as long as it can push the bearing to the corresponding outlet.
[0040] like Figure 3 As shown, the inspection station 2 is provided with an image acquisition device and a bearing positioning device, the image acquisition device includes a camera 8, the camera 8 is located above the track 5, and the bearing positioning device is located below the camera 8, so that each steel ball in the aviation ball bearing located at the inspection station 2 is positioned directly below the camera 8.
[0041] In order to capture images of each steel ball in the bearing, the bearing positioning device shown includes a rotating platform 14 and a rotating motor 18. The output end of the rotating motor 18 is connected to the rotating platform 14 to drive the rotating platform 14 to be installed. Accordingly, a receiving hole 17 that passes through from top to bottom is opened on the track 5 at the detection station 2. The rotating platform 14 is located in the receiving hole 17. The rotating platform drives the bearing to rotate at the shooting position of the camera 8, so that the contours of the embedded steel balls are exposed one by one in the field of view of the camera 8 lens 9. The method of partially shooting the bearing ball solves the contradiction between the measurement accuracy that the system must achieve and the resolution limit of the camera 8. The upper surface of the rotating platform 14 shown is coplanar with the bottom surface of the track 5 to ensure the smooth transposition of the bearing.
[0042] The upper surface of the rotating platform 14 is paved with a plane mirror 15 to ensure a large contrast between the steel ball outline and the background. The bearing surface is usually stained with impurities such as oil. Since the diameter of the steel ball itself is very small, the impurities on the steel ball surface may affect the calculation of the steel ball diameter. In this embodiment, the lower end of the push rod 6 located at the upper end of the track 5 is provided with bristles. The width of the push rod 6 with a brush is 2 to 3 mm less than the thickness of the ring, which is used to wipe off the bearing oil contaminated by the plane mirror 15 placed on the bearing rotating platform.
[0043] In order to make each steel ball of the bearing appear accurately near the field of view of the camera 8, the bearing positioning device includes a positioning component, the center of the rotating platform 14 is provided with an avoidance hole 16 which runs through from top to bottom, the positioning component is located in the avoidance hole 16, the positioning component includes a vertical rod 20, a rotating rod 21, a contact rod and a telescopic cylinder 23, the vertical rod 20 is arranged vertically, the lower end of the vertical rod 20 is fixedly connected to the output end of the rotating motor 18, the top end of the vertical rod 20 is hingedly connected to one end of the rotating rod 21, and the other end of the rotating rod 21 is hingedly connected to the output end of the rotating motor 18. The end is fixedly connected to the abutment rod, and the fixed end of the telescopic cylinder 23 is installed on the vertical pole 20. The telescopic end of the telescopic cylinder 23 is fixedly connected to the middle part of the rotating rod 21. When positioning is required, the telescopic cylinder 23 drives the rotating rod 21 to lift up and the abutment rod 22 abuts against the bottom end of the bearing to be tested 11 to make the bearing and the rotating platform 14 rotate coaxially, thereby ensuring that each ball contour can appear near the center of the field of view of the camera 8. When positioning is not required, the telescopic cylinder 23 drives the rotating rod 21 to descend, and the rolling bearing is freed from the constraints of the positioning component.
[0044] The image acquisition device includes a camera 8 support, a coaxial light source 10 and a photoelectric switch, the camera 8 is fixed on the camera 8 support, the lens 9 of the camera 8 is located above the detection station 2, and the photoelectric switch is installed on the camera 8 support to obtain a high-quality image with sufficient resolution, clear edges and accurate imaging. The photoelectric switch includes a transmitter 12 and a receiver 13;
[0045] The camera 8 uses pulse triggering to shoot the ball contour, uses a coaxial light source 10 to arrange the light field, and uses a telecentric lens 9 to reduce imaging distortion. The above-mentioned photoelectric sensor switch is used as a pulse generator. According to the uniform distribution of the bearing balls, it is arranged at the current non-shooting steel ball. With the rotation of the bearing, the periodic pulse is generated to control the shooting time of the camera 8, and each image is a single-sided contour of the ball.
[0046] Based on the above-mentioned aviation ball bearing sorting system, in a specific embodiment, a method for sorting aviation ball bearings is provided, comprising the following steps:
[0047] S1, collect photos of the left and right sides of the contours of one of the steel balls in the rolling bearing;
[0048] When the bearing is pushed onto the rotating platform 14 of the inspection station 2 through the entrance, the positioning assembly causes the bearing to rotate coaxially with the rotating platform 14. During the rotation process, when the steel ball blocks the photoelectric receiver, the camera 8 is triggered to capture the single-sided contour of the steel ball. The upper and lower trigger edges of each photoelectric sensing pulse correspond to capturing the left and right contours of the same ball.
[0049] To accurately measure the ball contour size, a coaxial light source 10 is required. The detection accuracy requirement is 1μm=0.001mm, and the ball contour range is 1mm*1m. The actual size measured is 1mm, and the required field of view is selected as 1.2*1.2mm; pixel accuracy: 2 pixels represent a boundary, 0.001mm / 2; camera 8 pixels = field of view / pixel accuracy = 1.2 / (0.001 / 2)*1.2 / (0.001 / 2)=2400*2400=576 0000, and a 5 million pixel camera 8 is selected.
[0050] S2, after filtering and denoising the image, use the R2U-Net++ neural network to perform image segmentation;
[0051] The image segmentation method established is: Recurrent Residual Convolutional Neural Network based on U-Net++ (R2U-Net++). The ordinary convolutional layer of the network is replaced with a recurrent residual convolutional layer, and the network structure of U-Net++ with nested and dense skip connections is used.
[0052] Residual units help train deep architectures, while feature accumulation with cyclic residual convolution layers ensures better feature representation for segmentation tasks, i.e. it allows us to design better U-Net++ architectures. With the same number of network parameters, using cyclic residual convolutions will have better image segmentation performance. U-Net++ is essentially a deeply supervised encoder-decoder network, where the encoder and decoder subnetworks are connected by a series of nested dense skip paths. The redesigned skip paths aim to reduce the semantic gap between the feature maps of the encoder and decoder subnetworks, making it easier for the optimizer to handle learning tasks, and U-Net++ with deep supervision has a higher average IoU gain in image segmentation.
[0053] S3, extracting edges of the segmented image and fitting circles using an improved Hough circle transform algorithm;
[0054] In the present invention, the proportion of the image arc to the entire circle is limited. In order to make full use of the arc information carried by the image and perform high-precision detection of the circle contour, based on the traditional Hough method, this embodiment adopts two steps to perform Hough circle fitting. The equation of the circle can be expressed as (ax) in Hough transform. 2 +(by) 2 =r 2 In order to reduce the operation time, in the first step, a relatively rough parameter space division is used to find the approximate Hough space parameters of the circle to be fitted. In the second step, the original image is first sub-pixel processed to obtain a more refined measurement image, and then the parameters obtained in the first step are used within a certain range to find the more refined Hough space parameters generated on the pixel image. The specific Hough circle transform algorithm process is as follows:
[0055] S31, reading the original image (I);
[0056] S32, image preprocessing, median filtering (M), grayscale (G), and binarization to obtain an image (B);
[0057] S33, roughly divide the parameters a, b, r, perform a standard Hough transform process, find the coordinates of the maximum value in the accumulator, and record them as (a1, b1, r1);
[0058] S34, performing sub-pixel division on the median filtered image (M) to obtain a sub-pixel image (S);
[0059] S35. In the appropriate neighborhood of (a1, b1, r1) obtained in step S333, finely divide the parameter a, b, r space, and perform the standard Hough transform process again to find the coordinates of the maximum value in the accumulator, recorded as (a2, b2, r2), which are the parameters of the circle to be fitted.
[0060] S4, calculating the diameter of the fitted circle of the same steel ball image and taking the average value as the diameter of the embedded steel ball;
[0061] Two images are obtained for the same ball, and two circle edges are obtained after segmentation. Then, two circle diameters d1 and d2 are obtained by Hough transform fitting. The average of the two circle diameters is obtained to obtain the final measured diameter d of the ball, that is, d = (d1 + d2) / 2.
[0062] S5. Repeat steps S1-S4 until the diameters of all the steel balls of the aviation ball bearing are detected. When the diameters of each steel ball of the aviation ball bearing are the same, it indicates that the aviation ball bearing meets the requirements. Otherwise, it is an aviation ball bearing that does not meet the requirements. When the aviation ball bearing meets the requirements, the transfer device pushes the tested aviation ball bearing to the normal product outlet 3. Otherwise, the tested aviation ball bearing is pushed to the defective product outlet 4.
[0063] The sorting system and sorting method of the above embodiment of the present invention are used to image the two arc edges of an embedded ball separately, which reduces the field of view of the camera 8, helps to improve the pixel accuracy, and obtains the required measurement accuracy within the limited camera 8 pixels; at the same time, the cyclic residual convolution layer is used as the neural network base layer, and the nested and densely connected network structure of U-Net++ is combined to accurately segment the image. Machine vision technology completes the precise measurement of the size of the embedded steel ball in the assembly bearing, with an accuracy of up to 1μm.
[0064] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.
Claims
1. A method for sorting aviation ball bearings, characterized in that: The invention is realized based on an aviation ball bearing sorting system, wherein the aviation ball bearing comprises a steel ball, and the aviation ball bearing sorting system comprises a track and a transfer device, wherein the track is provided with a fixed entrance station, a detection station, a normal product outlet and a defective product outlet in sequence; The indexing device includes at least one push rod and at least one push assembly, one end of the push rod extends to the upper end of the track to push the aviation ball bearing to move on the track, and the push assembly is used to push the aviation bearing on the track to the normal product outlet or the defective product outlet; The inspection station is provided with an image acquisition device and a bearing positioning device, the image acquisition device includes a camera, the camera is located above the track, and the bearing positioning device is located below the camera, so that each steel ball in the aviation ball bearing located at the inspection station is positioned directly below the camera; the track of the inspection station is provided with a receiving hole that passes through from top to bottom, and the bearing positioning device includes a rotating platform, the rotating platform is located in the receiving hole, and the upper surface of the rotating platform is coplanar with the bottom surface of the track; The bearing positioning device includes a positioning assembly, a vertically penetrating avoidance hole is provided at the center of the rotating platform, the positioning assembly is located in the avoidance hole, the positioning assembly includes a vertical pole, a rotating rod, a contact rod and a telescopic cylinder, the vertical pole is arranged vertically, one end of the rotating rod is hingedly connected to the top end of the vertical pole, the other end of the rotating rod is fixedly connected to the contact rod, the fixed end of the telescopic cylinder is installed on the vertical pole, and the telescopic end of the telescopic cylinder is fixedly connected to the middle part of the rotating rod; the lower end of the push rod located at the upper end of the track is provided with bristles; The aviation ball bearing sorting method comprises the following steps: S1, collect photos of the left and right sides of the contours of one of the steel balls in the rolling bearing; S2, after filtering and denoising the image, use the R2U-Net++ neural network to perform image segmentation; S3, extracting edges of the segmented image and fitting circles using an improved Hough circle transform algorithm; S4, calculating the diameter of the fitted circle of the same steel ball image and taking the average value as the diameter of the embedded steel ball; S5. Repeat steps S1-S4 until the diameters of all the steel balls of the aviation ball bearing are detected. If all the steel balls of the aviation ball bearing to be detected meet the requirements, the transfer device will push the aviation ball bearing that has been tested to the normal product outlet. Otherwise, the aviation ball bearing that has been tested will be pushed to the defective product outlet.
2. A method for sorting aviation ball bearings according to claim 1, characterized in that: A plane mirror is laid on the upper surface of the rotating platform.
3. The method for sorting aviation ball bearings according to claim 1, characterized in that: The indexing device comprises a power mechanism, and the output end of the power mechanism drives the push rod to move.
4. The method for sorting aviation ball bearings according to claim 1, characterized in that: The image acquisition device comprises a camera bracket, a coaxial light source and a photoelectric switch. The camera is fixed on the camera bracket, the lens of the camera is located above the detection station, and the photoelectric switch is installed on the camera bracket.
