Assembly precision detection system for bearing copper retainer
By performing geometric correction and deviation analysis on the top and side view images of the bearing copper retainer, the detection error problem caused by the positional offset of the rolling bearing was solved, and a higher precision assembly accuracy assessment was achieved.
Patent Information
- Application Number
- CN202511482181.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-16
- Publication Date
- 2026-02-06
AI Technical Summary
Existing technologies do not adequately consider image projection offset caused by rolling bearing position displacement in bearing retainer assembly accuracy testing, which affects testing accuracy.
The image acquisition module acquires top-view and side-view images, the image correction module performs geometric transformation correction, the deviation analysis module obtains distance dispersion and position misalignment index, and finally the assembly accuracy of the bearing copper retainer is determined by the assembly accuracy detection module.
This improves the accuracy of bearing copper cage assembly precision testing, enabling more precise assessment of assembly geometric deviations and rolling element distribution uniformity, thus ensuring the overall performance of the bearing.
Smart Images
Figure CN121481933A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application relates to the technical field of assembly precision detection, in particular to an assembly precision detection system for a bearing copper retainer. BACKGROUND
[0002] As one of key components in the field of modern mechanical manufacturing, a rolling bearing is composed of an inner ring, an outer ring, rolling elements and a bearing retainer; among them, the bearing retainer of brass type is widely used due to its good heat conductivity and wear resistance. The function of the retainer is to uniformly separate the rolling elements to prevent the rolling elements from colliding with each other during operation, thereby reducing friction, noise and improving service life. During the assembly of the bearing, the assembly precision between the retainer and the rolling elements, the inner ring and the outer ring directly affects the operation performance of the bearing. Therefore, it is necessary to detect the assembly precision of the bearing copper retainer, so as to realize real-time monitoring and accurate control of the assembly quality of the bearing, and ensure the overall performance of the bearing.
[0003] In the prior art, when the assembly effect of the bearing retainer in the rolling bearing is detected, the top view of the retainer is usually analyzed and detected by using visual detection technology. However, during the batch assembly detection of the bearing retainer, the position of the rolling bearing on the conveying belt may be offset, which may cause the rolling bearing not to be located at the center position of the captured image, and thus the geometric projection of the rolling bearing in the captured top view image may be eccentrically distorted. The prior art does not fully consider the influence of the position offset of the bearing on the assembly precision detection, which may cause errors in the assembly precision detection of the subsequent bearing retainer, and the visual detection effect is not good. SUMMARY
[0004] In order to solve the above technical problems, the purpose of the application is to provide an assembly precision detection system for a bearing copper retainer, and the technical scheme is as follows: The application provides an assembly precision detection system for a bearing copper retainer, which comprises: An image acquisition module: acquiring top view images and side view images of each rolling bearing; An image correction module: acquiring a circular edge contour in each top view image, and dividing it into an outer ring contour, an inner ring contour, a retainer contour and a rolling element contour; and performing geometric transformation correction on each top view image according to the center point of each top view image and the center point of the outer ring contour thereof; The deviation analysis module: according to the dispersion degree of the distance between the outer ring profile and the retainer profile and the dispersion degree of the distance between the retainer profile and the inner ring profile in the corrected overhead image of each rolling bearing, the distance dispersion of each rolling bearing is obtained, and the position dislocation index of each rolling bearing is obtained by combining the dispersion degree of the height of all pixel points in the upper edge profile of the rolling bearing in the side view image. The assembly precision detection module: according to the dispersion degree of the nearest distance between each rolling body profile and the center point of the outer ring profile and the dispersion degree of the distance between all adjacent rolling body profiles in the corrected overhead image of each rolling bearing, the rolling body fitting index of each rolling bearing is obtained, and the assembly abnormal index of each rolling bearing is obtained by combining the position dislocation index of each rolling bearing, and then whether the assembly precision of the bearing copper retainer in each rolling bearing is qualified is judged.
[0005] Preferably, the specific process of dividing it into an outer ring profile, an inner ring profile, a retainer profile and a rolling body profile is: converting each overhead image into a grayscale image, and obtaining the circular edge profile in each overhead grayscale image; the circular edge profile with a circumference greater than a preset segmentation threshold is recorded as a target profile, and the circular edge profile with a circumference less than or equal to the preset segmentation threshold is recorded as a rolling body profile; the two target profiles with the largest circumference are recorded as the outer ring profile, and the two target profiles with the smallest circumference are recorded as the inner ring profile, and all target profiles except the inner ring profile and the outer ring profile are recorded as the retainer profile.
[0006] Preferably, the specific process of geometric transformation correction of each overhead image is: taking the center of the outer ring profile with the largest circumference in each overhead image as the reference, and performing geometric transformation correction on each overhead image by image affine transformation method, so that the center pixel point of the corresponding overhead image coincides with the center position.
[0007] Preferably, the method for obtaining the distance dispersion of each rolling bearing is: taking the center of any outer ring profile in the corrected overhead image of each rolling bearing as the origin, and uniformly selecting a preset number of directions in the 360° direction centered on the origin, and recording the average of the variance of the outer ring distance and the variance of the inner ring distance in all directions as the distance dispersion of each rolling bearing; wherein the outer ring distance and the inner ring distance in each direction respectively refer to the Euclidean distance between the pixel points of the same outer ring profile and the pixel points of the same retainer profile, and the Euclidean distance between the pixel points of the same retainer profile and the pixel points of the same inner ring profile.
[0008] Preferably, the position dislocation index of each rolling bearing refers to the sum of the distance dispersion and the height dispersion value of each rolling bearing.
[0009] Preferably, the method for obtaining the height dispersion value of each rolling bearing comprises: obtaining all edge points in the corrected side view image of each rolling bearing; constructing a two-dimensional rectangular coordinate system with the lower left corner pixel of the corrected side view image as the origin, and recording the edge point with the largest vertical coordinate among all edge points corresponding to the same horizontal coordinate as the upper edge point of the same horizontal coordinate; and recording the variance between the vertical coordinates of all upper edge points in the corrected side view image of each rolling bearing as the height dispersion value of the rolling bearing.
[0010] Preferably, the method for obtaining the rolling body fitting index of each rolling bearing comprises: calculating the first ratio of each rolling bearing; calculating the variance of the Euclidean distance between all contour vertices and any outer ring contour center point in the corrected top view image of each rolling bearing; and recording the ratio of the first ratio and the variance as the rolling body fitting index of each rolling bearing; wherein each contour vertex in the corrected top view image refers to the pixel point closest to the outer ring contour center point on the contour of each rolling body in the corrected top view image.
[0011] Preferably, the method for obtaining the first ratio of each rolling bearing comprises: establishing a two-dimensional rectangular coordinate system with any outer ring contour center point in the corrected top view image as the coordinate origin, obtaining the angle of the line connecting each contour vertex and the origin in the two-dimensional rectangular coordinate system, and arranging all angles in ascending order to obtain the angle sequence of the corrected top view image; and recording the ratio of the minimum value and the maximum value in the first difference sequence of the angle sequence of the corrected top view image of each rolling bearing as the first ratio of each rolling bearing.
[0012] Preferably, the assembly abnormality index of each rolling bearing is positively correlated with the position misalignment index of each rolling bearing and negatively correlated with the rolling body fitting index of each rolling bearing.
[0013] Preferably, the specific process of judging whether the assembly precision of the bearing brass retainer in each rolling bearing is qualified comprises: when the assembly abnormality index of any rolling bearing is greater than a preset qualified threshold, judging that the assembly precision of the bearing brass retainer of the rolling bearing is unqualified; otherwise, judging that the assembly precision of the bearing brass retainer of the rolling bearing is qualified.
[0014] The present application has the following beneficial effects: The present application aims at the problem in the prior art that the position deviation of the bearing in the transmission process is not fully considered, resulting in image projection deviation and affecting the assembly precision detection. The present application performs image affine transformation on the top view image, so that the corrected top view image can more accurately represent the shape of the rolling bearing. The position misalignment index is constructed, so as to reflect the assembly geometric deviation of the bearing copper retainer in the plane centering and height direction, which is beneficial to the subsequent evaluation of the assembly precision of the bearing copper retainer. The rolling element fitting index is constructed, which can reflect the uniformity and radial consistency of the distribution of all rolling elements after the assembly of the bearing copper retainer. Then, the assembly abnormality index is constructed, so as to evaluate the assembly precision of the bearing copper retainer in the rolling bearing and improve the detection precision of the assembly precision of the bearing copper retainer. BRIEF DESCRIPTION OF DRAWINGS
[0015] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.
[0016] Figure 1 A block diagram of a bearing copper retainer assembly precision detection system provided by an embodiment of the present application; Figure 2 A flowchart of obtaining the assembly abnormality index of each rolling bearing provided by an embodiment of the present application. DETAILED DESCRIPTION
[0017] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific implementation, structure, features and effects of a bearing copper retainer assembly precision detection system according to the present application are described in detail as follows by combining with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.
[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.
[0019] The specific scheme of the bearing copper retainer assembly precision detection system provided by the present application is described in detail below in combination with the drawings.
[0020] Please refer to Figure 1The system includes an image acquisition module, an image correction module, a deviation analysis module, and an assembly precision detection module.
[0021] The image acquisition module acquires the top view images and the side view images of the rolling bearings.
[0022] The application takes pictures of the assembled rolling bearings in the batch assembly detection process through an industrial camera, acquires the top view images and the side view images of the rolling bearings respectively, filters and denoises the photographed top view images and side view images through an image filtering algorithm, then converts the filtered images into gray scale images, and records them as top view gray scale images and side view gray scale images respectively, and completes the preprocessing operation of the images.
[0023] The filtering algorithm is not limited to Gaussian filtering, median filtering, and bilateral filtering. The filtering algorithm and the gray scale image conversion technology are known technologies, and the specific process will not be described again.
[0024] The image correction module acquires the circular edge profile in each top view image, divides it into an outer ring profile, an inner ring profile, a retainer profile, and a rolling body profile, and performs geometric transformation correction on each top view image according to the center point of the top view image and the center point of the outer ring profile.
[0025] In the rolling bearing that has been assembled, the bearing copper retainer is located between the outer ring and the inner ring, and the inner ring, the outer ring, the rolling body, and the retainer in the rolling bearing are all standard circles. If the rolling bearing is in the center of the photographed image and the assembly precision of the retainer is good, the gap between the retainer and the outer ring in each direction around the circumference is relatively consistent, and the distance to the inner ring in each direction also tends to be consistent. Therefore, the relative distance distribution uniformity between the retainer and the inner ring and the outer ring can be analyzed by a visual detection method, so as to position the retainer and perform preliminary assembly detection analysis.
[0026] However, it needs to be considered that the rolling bearing may no longer be located in the center of the photographed image due to the influence of factors such as initial placement position and conveyor belt vibration, so at this time the shapes of the retainer and the inner and outer rings in the top view image may change from standard circles to ellipses, thereby causing the gap between the retainer and the inner and outer rings in different directions to no longer remain consistent. Therefore, the image needs to be corrected first to avoid affecting the subsequent visual detection precision.
[0027] Since the retainers, inner and outer rings and rolling elements inside the bearing are all circular or elliptical in shape, the top view gray-scale image of the rolling bearing is taken as the input of the Hough circle detection algorithm, and the output of the Hough circle detection algorithm is the circular edge profile of all the circular or semi-circular objects in the image. Since the circular profile of the rolling element is smaller than the profiles of the inner and outer rings and the retainers, it can be distinguished by the profile perimeter.
[0028] The perimeters of all the circular profiles are calculated and taken as the input of the Otsu thresholding method, and the output is the segmentation threshold, which is denoted as the preset segmentation threshold. The circular edge profile with a perimeter greater than the preset segmentation threshold is denoted as the target profile, and the circular edge profile with a perimeter less than or equal to the preset segmentation threshold is denoted as the rolling element profile.
[0029] Since the assembled structure of the rolling bearing presents a nested relationship of the outer ring enveloping the retainer, and the retainer enveloping the inner ring, the profile perimeters will also present a relationship of decreasing in turn, that is, the outer ring profile perimeter is the largest, the retainer profile perimeter is the second, and the inner ring profile perimeter is the smallest.
[0030] Therefore, the two target profiles with the largest perimeter are denoted as the outer ring profile, the two target profiles with the smallest perimeter are denoted as the inner ring profile, and the target profiles other than the inner ring profile and the outer ring profile are denoted as the retainer profile.
[0031] Taking the center of the outer ring profile in the top view gray-scale image as the reference, the top view gray-scale image is geometrically transformed and corrected by the image affine transformation method, so that the center pixel point of the corresponding top view image coincides with the center position, and the transformed image is denoted as the corrected top view image. The image affine transformation algorithm is not limited to the warpAffine function and flip function based on OpenCV.
[0032] At this point, the edge profile of the bearing retainer can be located by the visual detection method, and the image can be corrected, thereby providing support for the subsequent assembly detection of the retainer.
[0033] The deviation analysis module: according to the dispersion degree of the distance between the outer ring profile and the retainer profile in the corrected top view image of each rolling bearing in different directions, and the dispersion degree of the distance between the retainer profile and the inner ring profile in different directions, the distance dispersion degree of each rolling bearing is obtained, and combined with the dispersion degree of the height of all the pixel points in the upper edge profile of the rolling bearing in the side view image, the position dislocation index of each rolling bearing is obtained.
[0034] Since the widths of the retainers, inner and outer rings in the assembled rolling bearing are fixed, and the gap distances between the retainers and the inner and outer rings are relatively stable, the assembly effect of the retainers can be preliminarily analyzed by calculating the distance between the retainer edge and the inner and outer ring edges.
[0035] Take the outer ring profile with the largest circumference, the retainer profile with the largest circumference, and the inner ring profile with the largest circumference as examples for analysis.
[0036] In the corrected top view image of the single rolling bearing, the center of the outer ring profile is taken as the origin; in any direction starting from the origin, the Euclidean distances between the pixel points of the outer ring profile and the pixel points of the retainer profile and the Euclidean distances between the pixel points of the retainer profile and the pixel points of the inner ring profile are calculated respectively, and are recorded as the outer ring distance and the inner ring distance in the direction respectively.
[0037] In the same way, n directions are selected uniformly in the 360° direction centered on the origin, and the outer ring distances and the inner ring distances in the n directions are calculated respectively, and n is taken as 360 in this embodiment; and the mean of the dispersion degrees of the outer ring distances and the dispersion degrees of the inner ring distances in all directions is recorded as the distance dispersion degree of the rolling bearing. The distance dispersion mean can reflect the uniformity of the gap distribution between the bearing copper retainer and the inner and outer rings in the assembled rolling bearing, and the smaller the value, the better the uniformity, which means that the bearing copper retainer is more likely to be centered and not to be significantly offset, and the assembly precision is better. The calculation of the dispersion degree is not limited to variance, coefficient of variation, and mean square deviation, and variance is used for calculation in this embodiment.
[0038] Further, if the retainer has unevenness or inclination in height after assembly, and only the top view is used for analysis, the assembly precision of the retainer may be misjudged as good, and a large error may occur, so further analysis is needed.
[0039] Since the well-assembled bearing is a standard rectangle in the side view image, even if the retainer and the inner ring or the outer ring have inconsistent heights, the left and right edges in the side view image will also show straight lines. Therefore, the side view grayscale image is taken as the input of the Hough line detection algorithm for edge line extraction, and the output of the Hough line detection algorithm is all the edge lines in the side view image.
[0040] Then, all the edge points in the corrected side view image of the rolling bearing are obtained by the edge detection algorithm. The edge detection algorithm is not limited to Canny and Prewitt algorithms.
[0041] A two-dimensional rectangular coordinate system is constructed with the pixel point at the lower left corner of the corrected side view image as the origin, and the edge point with the largest vertical coordinate among all the edge points corresponding to the same horizontal coordinate is recorded as the upper edge point of the same horizontal coordinate, representing the point in the upper edge profile of the rolling bearing in the side view image. It should be noted that when there is no edge point in a column of pixel points corresponding to any horizontal coordinate, the upper edge point does not exist for the horizontal coordinate.
[0042] The dispersion degree between the longitudinal coordinates of all the upper edge points in the corrected side view image is denoted as the height dispersion value of the rolling bearing. The height dispersion value can reflect whether the height of the upper edge contour of the rolling bearing in the side view gray image is consistent, and the smaller the value is, the more consistent the height is, thereby reflecting that the assembly effect of the bearing brass retainer is better.
[0043] As a preferred embodiment, according to the distance dispersion degree and the height dispersion value of each rolling bearing, a position misalignment index of each rolling bearing is obtained, which is used to represent the possibility of assembly misalignment of the bearing brass retainer in each rolling bearing.
[0044] In the embodiment, the position misalignment index of the rolling bearing is denoted as , and the specific expression is: ; in the formula, is the position misalignment index of the rolling bearing, B is the distance dispersion degree of the rolling bearing, and C is the height dispersion value of the rolling bearing.
[0045] The position misalignment index can reflect the geometric deviation degree of the assembly of the bearing retainer in the top view image and the side view image; the smaller the value is, the better the effects of the planar centering and the height flattening of the bearing retainer in the rolling bearing are, the smaller the possibility of position deviation is, and the greater the possibility of good assembly precision is.
[0046] Assembly precision detection module: according to the dispersion degree of the nearest distance between each rolling body contour and the contour center point of the outer ring in the corrected top view image of each rolling bearing, and the dispersion degree of the distance between all adjacent rolling body contours, a rolling body fitting index of each rolling bearing is obtained, and combined with the position misalignment index of each rolling bearing, an assembly abnormality index of each rolling bearing is obtained, and then whether the assembly precision of the bearing brass retainer in each rolling bearing is qualified is judged.
[0047] Further, one of the core roles of the bearing retainer is to uniformly separate the rolling bodies to prevent the rolling bodies from colliding with each other and to ensure uniform load distribution. In the already assembled bearing, the retainer and the rolling bodies have undergone a stamping process, thereby placing the rolling bodies on the retainer.
[0048] If the rolling bodies are not correctly positioned or the positioning effect is not good after assembly, even if the retainer is centered as a whole, the spacing between the rolling bodies may not be uniform, thereby affecting the running performance and service life of the rolling bearing. Therefore, whether the assembly effect of the retainer is good can also be further evaluated according to the position distribution between each rolling body.
[0049] In the corrected top view image, the pixel point closest to the center point of the outer ring profile on each rolling body profile is recorded as the profile vertex of each rolling body profile. The dispersion degree of the Euclidean distance between all profile vertices of the rolling bearing in the corrected top view image and any center point of the outer ring profile is calculated. The dispersion degree can reflect the stability of the rolling body in the radial positioning and the tightness of the rolling body and the retainer; the smaller the value, the more likely that the distance of all profile vertices of the rolling body to the origin is close to consistent, and there is no obvious deviation and jamming phenomenon, and the retainer assembly precision is good.
[0050] Further, it needs to be considered whether there is a rolling body missing situation. If the missing phenomenon occurs, the first dispersion value may still be small, and then an error occurs.
[0051] With the center point of the outer ring profile in the corrected top view image as the coordinate origin, a two-dimensional rectangular coordinate system is established, the angle of the connecting line between each profile vertex and the origin in the two-dimensional rectangular coordinate system is obtained, and all angles are arranged in ascending order to obtain the angle sequence of the corrected top view image.
[0052] If the assembly precision of the rolling body on the retainer is good, the rolling bodies are uniformly spaced, and the angle difference between each adjacent rolling body is consistent.
[0053] The ratio of the minimum value to the maximum value in the first difference sequence of the angle sequence of the corrected top view image is recorded as the first ratio of the rolling bearing. The first difference sequence can reflect the distance interval between adjacent rolling bodies in the rolling bearing, and the first ratio can reflect the uniformity of the distance interval between the rolling bodies; when the first ratio is the maximum value 1, it means that the distance interval between each adjacent rolling body is consistent, and then the assembly effect of the retainer is better.
[0054] As a preferred embodiment, according to the dispersion degree of the closest distance between each rolling body profile and the center point of the outer ring profile in the corrected top view image of each rolling bearing, and the dispersion degree of the distance between all adjacent rolling body profiles, the rolling body fitting index of each rolling bearing is obtained, which is used to represent the uniformity of the position distribution of all rolling bodies of each rolling bearing.
[0055] In this embodiment, the rolling body fitting index of the rolling bearing is recorded as F, and the specific expression is: ; In the formula, F is the rolling body fitting index of the rolling bearing, D is the first ratio of the rolling bearing; G is the variance of the Euclidean distance between all profile vertices of the rolling bearing in the corrected top view image and any center point of the outer ring profile. is a preset constant, and its value range is [0.005, 0.01] to avoid the denominator being 0, and the value has little effect on the calculation and can be ignored, and 0.008 is taken in the embodiment.
[0056] The distribution fitting index can reflect whether the position distribution between the rolling elements after the retainer is assembled is uniform; the greater the value, the more uniform the distribution of each rolling element on the circumference, the higher the fitting degree with the ideal assembly state, and thus the greater the assembly precision of the bearing brass retainer.
[0057] Further, according to the position misalignment index and the rolling element fitting index of each rolling bearing, an assembly abnormality index of each rolling bearing is obtained, which is used to represent the possibility that the assembly precision of the bearing brass retainer in each rolling bearing is qualified. The assembly abnormality index of each rolling bearing has a positive correlation with the position misalignment index of each rolling bearing and a negative correlation with the rolling element fitting index of each rolling bearing. The flowchart of obtaining the assembly abnormality index of each rolling bearing is shown in FIG. 8. It should be noted that the dependent variable increases (decreases) with the increase (decrease) of the independent variable in the positive correlation, and the dependent variable decreases (increases) with the increase (decrease) of the independent variable in the negative correlation, and the specific calculation relationship is not specially limited in the present application. Figure 2
[0058] Preferably, in the embodiment, the assembly abnormality index of the rolling bearing is denoted as , and its specific expression is: ; in the formula, is the assembly abnormality index of the rolling bearing, is the position misalignment index of the rolling bearing, is the rolling element fitting index of the rolling bearing, is a preset constant. In the formula, the is added to the numerator to avoid the problem that the whole result is 0 when the numerator is 0, and thus the denominator is meaningless.
[0059] The assembly abnormality index of the rolling bearing can reflect whether the assembly effect of the bearing brass retainer in the rolling bearing is abnormal; the greater the value, the worse the assembly effect of the bearing brass retainer in the rolling bearing, and the more it needs to be reassembled.
[0060] According to the calculation method of the assembly abnormality index, the assembly abnormality indexes of all the rolling bearings assembled in the current batch assembly process are obtained, and the assembly abnormality indexes of all the rolling bearings are taken as the input of the abnormal value detection algorithm, and the value is obtained, wherein is the mean value of all the assembly abnormality indexes, is the standard deviation of all the assembly abnormality indexes; and the obtained The value is taken as a preset qualified threshold. When the assembly abnormality index of any rolling bearing is greater than the preset qualified threshold, it is judged that the assembly precision of the bearing brass retainer of the rolling bearing is unqualified, and re-assembly is needed to ensure the operation performance of the rolling bearing; otherwise, it is judged that the assembly precision of the bearing brass retainer of the rolling bearing is qualified.
[0061] It should be noted that the above-described order of the embodiments of the present application is merely for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.
[0062] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment mainly describes the difference from other embodiments.
[0063] The above only describes the preferred embodiments of the present application, and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.
Claims
1. A system for detecting the assembly accuracy of a bearing copper retainer, characterized in that, The system includes: Image acquisition module: Acquires top and side view images of each rolling bearing; Image correction module: acquires the circular edge contours in each top view image and divides them into outer contour, inner contour, retainer contour and rolling element contour; performs geometric transformation correction on each top view image based on the center point of each top view image and the center point of its outer contour; Deviation Analysis Module: Based on the dispersion of the distance between the outer ring contour and the cage contour in different orientations in the corrected top view image of each rolling bearing, and the dispersion of the distance between the cage contour and the inner ring contour in different orientations, the distance dispersion of each rolling bearing is obtained. Combined with the dispersion of the height of all pixels in the upper edge contour of the rolling bearing in the side view image, the positional misalignment index of each rolling bearing is obtained. Assembly accuracy detection module: Based on the dispersion of the shortest distance between the center point of each rolling element profile and the outer ring profile in the corrected top view image of each rolling bearing, and the dispersion of the distance between all adjacent rolling element profiles, the rolling element fit index of each rolling bearing is obtained. Combined with the position misalignment index of each rolling bearing, the assembly abnormality index of each rolling bearing is obtained, and then it is determined whether the assembly accuracy of the bearing copper retainer in each rolling bearing is qualified.
2. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The specific process of dividing it into outer ring contour, inner ring contour, retainer contour and rolling body contour is as follows: convert each top view image into a grayscale image, and obtain the circular edge contour in each top view grayscale image; record the circular edge contour with a perimeter greater than a preset segmentation threshold as the target contour, and record the circular edge contour with a perimeter less than or equal to the preset segmentation threshold as the rolling body contour; record the two target contours with the largest perimeter as the outer ring contour, the two target contours with the smallest perimeter as the inner ring contour, and record all target contours other than the inner ring contour and the outer ring contour as the retainer contour.
3. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The specific process of performing geometric transformation correction on each top view image is as follows: taking the center of the outer contour with the largest perimeter in each top view image as the reference, the geometric transformation correction is performed on each top view image through an image affine transformation method, so that the center pixel of the corresponding top view image coincides with the center position.
4. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The method for obtaining the distance dispersion of each rolling bearing is as follows: taking the center of any outer ring contour in the corrected top view image of each rolling bearing as the origin, a preset number of directions are uniformly selected in the 360° direction centered on the origin, and the mean of the variance of the outer ring distance and the variance of the inner ring distance in all directions is recorded as the distance dispersion of each rolling bearing; wherein, the outer ring distance and inner ring distance in each direction refer to the Euclidean distance between the same outer ring contour pixel point and the same retainer contour pixel point, and the Euclidean distance between the same retainer contour pixel point and the same inner ring contour pixel point in each direction, respectively.
5. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The positional misalignment index of each rolling bearing refers to the sum of the distance dispersion and height dispersion values of each rolling bearing.
6. The assembly accuracy detection system for a bearing copper retainer as described in claim 5, characterized in that, The method for obtaining the height discrete value of each rolling bearing is as follows: obtain all edge points in the corrected side view image of each rolling bearing; construct a two-dimensional rectangular coordinate system with the pixel point at the bottom left corner of the corrected side view image as the origin; record the edge point with the largest vertical coordinate among all edge points corresponding to the same horizontal coordinate as the upper edge point of the same horizontal coordinate; record the variance between the vertical coordinates of all upper edge points in the corrected side view image of each rolling bearing as the height discrete value of the rolling bearing.
7. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The method for obtaining the rolling element fit index of each rolling bearing is as follows: calculate the first ratio of each rolling bearing; calculate the variance of the Euclidean distance between all contour vertices and the center point of any outer ring contour in the corrected top view image of each rolling bearing; record the ratio of the first ratio to the variance as the rolling element fit index of each rolling bearing; wherein, each contour vertex in the corrected top view image refers to the pixel point on each rolling element contour in the corrected top view image that is closest to the center point of the outer ring contour.
8. The assembly accuracy detection system for a bearing copper retainer as described in claim 7, characterized in that, The method for obtaining the first ratio of each rolling bearing is as follows: taking the center point of any outer ring contour in the corrected top view image as the origin, a two-dimensional rectangular coordinate system is established, and the angle of the line connecting each contour vertex and the origin in the two-dimensional rectangular coordinate system is obtained. All angles are arranged in ascending order to obtain the angle sequence of the corrected top view image. The ratio of the minimum value to the maximum value in the first difference sequence of the angle sequence of each rolling bearing in the corrected top view image is recorded as the first ratio of each rolling bearing.
9. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The assembly anomaly index of each rolling bearing is positively correlated with the position misalignment index of each rolling bearing, and negatively correlated with the rolling element fit index of each rolling bearing.
10. The assembly accuracy detection system for a bearing copper retainer as described in claim 1, characterized in that, The specific process for determining whether the assembly accuracy of the bearing copper cage in each rolling bearing is qualified is as follows: when the assembly abnormality index of any rolling bearing is greater than the preset qualified threshold, the assembly accuracy of the bearing copper cage of that rolling bearing is determined to be unqualified; otherwise, the assembly accuracy of the bearing copper cage of that rolling bearing is determined to be qualified.