Method for evaluating manufacturing defects of bearing

Abstract

Disclosed is a method for evaluating manufacturing defects of a bearing, comprising: obtaining image data of the bearing, identifying a first number of defects on the bearing based on the image data of the bearing, performing a defect integration operation on the first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number, the defect integration operation includes determining whether to integrate two or more defects in the first number of defects based on a distance between the first number of defects and determining a target defect on the bearing based on the second number of defects. According to the method for evaluating the manufacturing defects of the bearing, some defects existing in the bearing can be integrated, and a more accurate defect recognition result can be obtained.

Description

Technical Field

[0001] This disclosure relates to machining equipment, and more particularly to a method for assessing manufacturing defects in bearings. Background Technology

[0002] Bearings are an important component in mechanical equipment. They support rotating shafts, maintaining the shaft's centered position when other components move relative to each other on the shaft. This supports the rotating parts, reduces friction during movement, and ensures rotational accuracy. Generally, a bearing includes components such as an inner ring, outer ring, rolling elements, and a cage.

[0003] During the manufacturing process, bearings may contain a certain number of defects. Currently, to check for defects in bearings, quality control engineers can identify them manually. However, depending on lighting conditions and the qualifications and fatigue level of the maintenance engineer, manual defect assessment of bearings involves significant uncertainty. Furthermore, manual defect identification by quality control engineers is extremely costly in terms of manpower, time, and money, and the results may not be sufficiently accurate.

[0004] Therefore, a method and apparatus for assessing manufacturing defects in bearings are desired. Summary of the Invention

[0005] Embodiments of this disclosure provide a method for evaluating manufacturing defects in a bearing, comprising: obtaining image data of the bearing; identifying a first number of defects on the bearing based on the image data of the bearing; performing a defect integration operation on the first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number; wherein the defect integration operation includes determining whether to integrate two or more defects from the first number of defects based on the distance between the first number of defects; and determining a target defect on the bearing based on the second number of defects.

[0006] According to an embodiment of the present disclosure, the defect integration operation includes: determining to integrate the two or more adjacent defects in the first number of defects based on the fact that the distance between two or more adjacent defects in the first number of defects is less than an integration distance threshold.

[0007] According to an embodiment of the method of this disclosure, the defect integration operation includes: determining an integration auxiliary attribute for each defect in a first number of defects; increasing an integration distance threshold for two or more defects based on the high similarity of the integration auxiliary attributes of two or more defects in the first number of defects; and decreasing the integration distance threshold for two or more defects based on the low similarity of the integration auxiliary attributes of two or more defects in the first number of defects.

[0008] According to an embodiment of the present disclosure, the defect integration operation includes: determining an integration auxiliary attribute for each defect in a first number of defects; and integrating defects belonging to the same cluster by clustering multiple defects using a clustering model based on the distance between adjacent defects in the first number of defects and the integration auxiliary attribute for each defect in the first number of defects.

[0009] According to the method of embodiments of this disclosure, the integrated auxiliary attributes include at least one of: the curvature of the edge of the defect, the grayscale of the pixel of the defect, and the uniformity of the pixel of the defect.

[0010] According to an embodiment of the present disclosure, determining a target defect on the bearing includes: determining the defect type of the second number of defects using a trained model based on a second number of defects.

[0011] The method according to an embodiment of the present disclosure further includes: determining a target defect on the bearing based on the size of each defect in the second number of defects.

[0012] According to an embodiment of the method of this disclosure, determining the target defect on the bearing further includes: for each defect type, determining a defect size threshold corresponding to the defect type, and in response to one or more defects in the second number of defects having a size greater than the defect size threshold corresponding to the defect type, determining the one or more defects as candidate target defects associated with the defect type.

[0013] According to the method of an embodiment of the present disclosure, determining the target defect on the bearing further includes: for each of the second number of defects, determining a defect auxiliary attribute of the defect; if the defect auxiliary attribute of the defect matches the defect type to a high degree, reducing the defect size threshold for the defect; and if the defect auxiliary attribute of the defect matches the defect type to a low degree, increasing the defect size threshold for the defect.

[0014] According to the method of an embodiment of the present disclosure, determining the target defect on the bearing further includes: determining the degree of deviation between each candidate target defect and its corresponding defect size threshold, and selecting one or more defects from the candidate target defects as the target defect based on the degree of deviation.

[0015] According to the method of embodiments of the present disclosure, for a batch of bearings, the number of bearings with each defect type is counted to generate statistical data results of the target defects.

[0016] According to the method of embodiments of the present disclosure, for a batch of bearings, the number of bearings with each defect type and the number of target defects for each defect type are counted to generate statistical data results of target defects.

[0017] According to the method of embodiments of the present disclosure, statistical data of target defects in a batch of bearings are compared with statistical data of target defects in a previous batch of bearings.

[0018] According to the method of an embodiment of the present disclosure, the defect auxiliary attributes include at least one of: the curvature of the edge of the defect, the grayscale of the pixel of the defect, the uniformity of the pixel of the defect, and the number of defects.

[0019] Embodiments of this disclosure provide a bearing manufacturing defect assessment apparatus, comprising: a camera configured to acquire image data of a bearing; a processor coupled to the camera and executing program code stored in a memory to perform the following operations: identifying a first number of defects on the bearing based on the image data of the bearing; performing a defect integration operation on the first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number; wherein the defect integration operation includes determining whether to integrate two or more defects from the first number of defects based on the distance between the first number of defects; and determining a target defect on the bearing based on the second number of defects.

[0020] Embodiments of this disclosure provide one or more non-transitory storage media storing instructions that, when executed by a processor, cause the processor to perform the bearing manufacturing defect assessment method as described above.

[0021] The bearing manufacturing defect assessment method, apparatus, and storage medium according to embodiments of this disclosure can identify the type of defect and determine the target defect in a low-cost, fast, and accurate manner based on an image of the bearing. By integrating several defects present in the bearing, more accurate defect identification results can be obtained. Attached Figure Description

[0022] The above and other aspects, features, and advantages of specific embodiments of the present disclosure will become clearer from the following description taken in conjunction with the accompanying drawings, in which:

[0023] Figure 1 This is an example schematic diagram of an image of a bearing containing defects.

[0024] Figure 2 This is an example flowchart of a method for evaluating manufacturing defects in bearings according to embodiments of the present disclosure.

[0025] Figure 3 This is an example schematic diagram of an image of a bearing according to an embodiment of the present disclosure.

[0026] Figure 4 This is an example flowchart of another method for evaluating manufacturing defects in bearings according to embodiments of the present disclosure.

[0027] Figure 5A This is an example flowchart of a method for evaluating manufacturing defects in a batch of bearings according to embodiments of the present disclosure.

[0028] Figure 5B This is an example flowchart of another evaluation method for manufacturing defects in a batch of bearings according to embodiments of this disclosure.

[0029] Figure 6 This is an example schematic diagram of a bearing manufacturing defect assessment apparatus according to an embodiment of the present disclosure.

[0030] Figure 7 This is an example schematic diagram of a non-transitory computer-readable storage medium according to at least one embodiment of the present disclosure. Detailed Implementation

[0031] Before proceeding with the detailed description below, it may be advantageous to define certain words and phrases used throughout this disclosure. The terms “comprising” and “including” and their derivatives mean, but are not limited to, “including”. The phrase “at least one”, when used with a list of items, means that different combinations of one or more of the listed items may be used, and that only one item in the list may be required. For example, “at least one of A, B, and C” includes any one of the following combinations: A, B, C, A and B, A and C, B and C, A and B and C.

[0032] Definitions of other specific words and phrases are provided throughout this disclosure. Those skilled in the art will understand that, in many, if not most, cases, such definitions apply to the prior and future use of the words and phrases thus defined.

[0033] The various embodiments of the principles of this disclosure described below with reference to the accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this disclosure in any way. Those skilled in the art will understand that the principles of this disclosure can be implemented in any suitably arranged system or device. In some cases, the actions described in this disclosure can be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific order or sequential sequence to achieve the desired result. In certain embodiments, multitasking and parallel processing may be advantageous.

[0034] The text and accompanying drawings are provided by way of example only to aid in understanding this disclosure. They should not be construed as limiting the scope of the claims appended to this disclosure in any way. Throughout the drawings, the same reference numerals generally indicate the same elements. Although certain embodiments and examples have been provided, it will be apparent to those skilled in the art, based on the content of this disclosure, that changes may be made to the illustrated embodiments and examples without departing from the scope of this disclosure.

[0035] Figure 1 This is an example schematic diagram of an image of a bearing containing defects.

[0036] like Figure 1 As shown, bearing image 100 may include multiple defects, such as defect 101, defect 102, defect 103, defect 104, defect 105, and defect 106. When evaluating manufacturing defects in the bearing, the defect size can be compared to a defect size threshold. For example, the size of each of defects 101-106 can be compared to a defect size threshold. Figure 1 As shown, Figure 1 All defects, including defects 101-106, have dimensions smaller than the defect size threshold. Therefore, it is possible to identify... Figure 1 The bearing image 100 in the image does not include manufacturing defects. However, such a manufacturing defect assessment is inaccurate because defects 101 and 102 could actually be a single defect, and the defect size of that single defect could be larger than the defect threshold size. Therefore, in Figure 1 In the example, the actual defect was incorrectly missed.

[0037] Figure 2 This is an example flowchart of a method for evaluating manufacturing defects in bearings according to embodiments of the present disclosure.

[0038] like Figure 2 As shown, in S201, image data of the bearing can be obtained.

[0039] In S202, the first number of defects on the bearing can be identified based on the image data of the bearing.

[0040] In S203, a defect integration operation can be performed on a first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number, wherein the defect integration operation includes determining whether to integrate two or more defects from the first number of defects based on the distance between the first number of defects.

[0041] In S204, the target defect on the bearing can be determined based on the second number of defects.

[0042] Figure 3 This is an example schematic diagram of an image of a bearing according to an embodiment of the present disclosure.

[0043] like Figure 3 As shown, a bearing image 300 can be obtained. For example, a bearing or bearing component (e.g., inner ring, outer ring, rolling elements, cage, etc.) prepared for manufacturing defect assessment can be cleaned. After removing surface impurities or debris from the bearing or bearing component, it is placed under specific lighting to clearly photograph the bearing or bearing component including several defects, thereby obtaining a magnified and focused bearing image 300. The specific lighting can be light with an intensity higher than a certain threshold. Alternatively or additionally, the specific lighting can be one or more of visible light and ultraviolet light. The photographed bearing image 300 can contain all defects of that part of the bearing. When multiple parts / side surfaces require manufacturing defect assessment, multiple images can be taken.

[0044] A first number of defects on the bearing can be identified based on the bearing image 300. For example, six defects can be identified on the bearing image 300, including defects 301, 302, 303, 304, 305, and 306.

[0045] A defect integration operation can be performed on a first number of defects (defects 301-305) on the bearing image 300 to obtain a second number of defects. The defect integration operation includes determining whether to integrate two or more defects from the first number of defects based on the distance between them. The distance between defects can refer to the distance between the closest points of two defects.

[0046] For example, the distance between two or more adjacent defects in a first number of defects on bearing image 300 can be determined. For example, the distance between any two defects from 301 to 306 can be determined. It can be determined that two or more adjacent defects in the first number of defects will be integrated based on the distance between them being less than an integration distance threshold. For example, defects 301 and 302 can be integrated into one defect, namely defect 307, based on the distance 'a' between them being less than the integration distance threshold. It can be determined that two or more adjacent defects in the first number of defects will not be integrated based on the distance between them being greater than or equal to the integration distance threshold. For example, defects 304 and 305 can be kept separate because the distance 'b' between them is greater than the integration distance threshold.

[0047] For example, depending on the integration distance threshold and the distance between defects 301 and 306, more or fewer defects can be integrated. For instance, when the distances between all defects 301 and 306 are greater than or equal to the integration distance threshold, no defects may be integrated. In this case, the second quantity can be equal to the first quantity. When the distance between at least two defects among defects 301 and 306 is less than the integration distance threshold, at least two defects with distances less than the integration distance threshold can be integrated. In this case, the second quantity can be less than the first quantity.

[0048] In one embodiment, an integrated auxiliary attribute can be determined for each defect in a first number of defects. The integrated auxiliary attribute may include, but is not limited to, the curvature of the defect's edge, the grayscale of the defect's pixels, the uniformity of the defect's pixels, etc.

[0049] For example, based on the high similarity of integration auxiliary attributes between two or more defects in a first number of defects, the requirements for integrating two or more defects can be reduced; for example, the integration distance threshold between the two or more defects can be increased. When the integration distance threshold is increased, the distance between the two or more defects is more likely to meet the requirements for defect integration; that is, the distance between the two or more defects is more likely to be less than the integration distance threshold. For example, in an alternative embodiment, if it is determined that the similarity of integration auxiliary attributes between defects 304 and 305 is high, the requirements for integrating defects 304 and 305 can be reduced; that is, the integration distance threshold between defects 304 and 305 can be increased. If the increased integration distance threshold between defects 304 and 305 is greater than distance b, then defects 304 and 305 can be integrated into a single defect in this case.

[0050] For example, based on the low similarity of integration auxiliary attributes of two or more defects among a first number of defects, the requirements for integrating two or more defects can be increased; for example, the integration distance threshold of two or more defects can be reduced. When the integration distance threshold is reduced, the distance between two or more defects is less likely to meet the requirements for defect integration; that is, the distance between two or more defects is more likely to exceed the integration distance threshold. For example, in an alternative embodiment, if it is determined that the similarity of integration auxiliary attributes between defects 301 and 302 is very low, the requirements for integrating defects 304 and 305 can be increased; that is, the integration distance threshold between defects 301 and 302 can be reduced. If the reduced integration distance threshold between defects 301 and 302 is less than distance 'a', then defects 301 and 302 may not be integrated into a single defect in this case, but rather remain separate.

[0051] In another embodiment, multiple defects can be clustered using a clustering model based on the distance between adjacent defects in a first number of defects and the integration auxiliary attribute of each defect in the first number of defects, thus integrating defects belonging to the same cluster. For example, the clustering model may include, but is not limited to, K-means clustering, hierarchical clustering, density-based spatial clustering of applications with noise (DBSCAN), Gaussian mixture model (GMM), density peaks clustering, and spectral clustering. When clustering using a clustering model, the distance between two or more defects can be given a higher weight than the integration auxiliary attribute of two or more defects.

[0052] Target defects on the bearing can be identified based on a second number of defects after performing defect integration operations. For example, the dimensions of a second number of defects, including defects 303, 304, 305, 306, and 307, can be compared with a defect size threshold. Based on the fact that the defect size of defect 307 is greater than the defect size threshold, defect 307 is identified as the target defect.

[0053] and Figure 1 In contrast, in embodiments according to this disclosure Figure 3During the process, due to the implementation of defect integration operations, defect 307, as the target defect, can be correctly identified. By conducting a more accurate manufacturing defect assessment of the bearing, more realistic defect results can be provided to the user, allowing them to adjust the manufacturing process or workflow.

[0054] Figure 4 This is an example flowchart of another method for evaluating manufacturing defects in bearings according to embodiments of the present disclosure.

[0055] Figure 4 Steps S401-S403 and Figure 2 Steps S201-S203 are similar and will not be described again to avoid redundancy.

[0056] In S404, the defect types of a second number of defects can be determined using a trained model. The trained model can include trained machine learning models and trained deep learning models. For example, the model can be trained based on features (such as shape features) of different types of defects to determine the defect types of the second number of defects. The machine learning model can include one or more of decision trees, random forests, extreme gradient boosting (XGBoost), support vector machines (SVM), etc. The deep learning model can include one or more of convolutional neural networks (CNN), object detection models (YOLO, Fast R-CNN), etc.

[0057] In S405, the target defect on the bearing can be determined based on the size of the defects of various defect types in the second number of defects.

[0058] In one embodiment, defect size thresholds can be determined for various defect types. A defect can be identified as a target defect in response to a second number of defects having a defect size threshold corresponding to its defect type.

[0059] In another embodiment, a defect size threshold can be determined for each defect type. A defect can be identified as a candidate target defect if it exceeds the defect size threshold corresponding to its defect type in a second number of defects. For each defect in the second number of defects, auxiliary defect attributes can be determined. These auxiliary attributes may include, but are not limited to, the curvature of the defect's edges, the grayscale of the defect's pixels, the uniformity of the defect's pixels, and the number of defects.

[0060] Reference Figure 3For example, based on a high degree of match between the defect auxiliary attributes and the defect type, the requirements for identifying a defect as a candidate target defect can be lowered; for example, the defect size threshold for that defect can be reduced. In an alternative embodiment, if the defect size of defect 305 is smaller than the defect size threshold corresponding to the defect type of defect 305, and if the defect auxiliary attributes of defect 305 match the type of defect 305 highly, the defect size threshold for defect 305 can be reduced. The reduced defect size threshold for defect 305 may be smaller than the defect size of defect 305. In this way, defect 305 can be identified as a candidate target defect. For example, based on a low degree of match between the defect auxiliary attributes and the defect type, the requirements for identifying a defect as a candidate target defect can be increased; for example, the defect size threshold for that defect can be increased. In an alternative embodiment, if the defect size of defect 306 is smaller than the defect size threshold corresponding to the defect type of defect 306, and if the defect auxiliary attributes of defect 306 match the type of defect 306 poorly, the defect size threshold for defect 306 can be increased. The increased defect size threshold for defect 306 may be greater than the defect size of defect 306. In this way, defect 306 can be identified as not a candidate target defect.

[0061] In one embodiment, candidate target defects can be directly identified as target defects. In another embodiment, the degree of deviation between each candidate target defect and its corresponding defect size threshold can be determined, and one or more defects among the candidate target defects can be selected as target defects based on the degree of deviation. In one embodiment, the degree of deviation can be the ratio of the difference between the defect size of the candidate target defect and the defect size threshold to the defect size threshold. In another embodiment, the degree of deviation can be the ratio of the defect size of the candidate target defect to the defect size threshold. Those skilled in the art should understand that other implementations of the degree of deviation are also possible. According to embodiments of this disclosure, candidate target defects can be sorted in descending order of degree of deviation, and one or more defects that rank higher (e.g., have a larger degree of deviation) among the candidate target defects can be identified as target defects. For example, the defect with the largest degree of deviation can be identified as the target defect. See also Figure 3In one embodiment, defects 307 and 305 can be identified as candidate target defects. When the ratio of the difference between the defect size and a defect size threshold of a candidate target defect to the defect size threshold is used as the degree of deviation, the degree of deviation between defect 307 and its corresponding defect size threshold can be 40%, and the degree of deviation between defect 305 and its corresponding defect size threshold can be 2%. After sorting defects 307 and 305 according to their degree of deviation from largest to smallest, when selecting the defect with the highest ranking, defect 307 can be identified as the target defect. In this way, the number of target defects can be reduced, and the most prominent defect among the candidate target defects can be identified as the target defect. Since other candidate target defects may be related to the target defect, adjusting the bearing manufacturing process or workflow for the target defect can indirectly solve or mitigate other candidate target defects.

[0062] Figure 5A This is an example flowchart of a method for evaluating manufacturing defects in a batch of bearings according to embodiments of the present disclosure.

[0063] like Figure 5A As shown, in S501A, the manufacturing defect assessment results of one bearing in a batch of bearings can be obtained. For example, this can be achieved through methods such as... Figure 2 as well as Figure 4 The method for assessing manufacturing defects in bearings shown in the diagram evaluates the manufacturing defects in the bearing.

[0064] In S502A, the target defect type of the bearing can be recorded. For example, in Figure 3 In the example, defects 307 and 305 might be identified as target defects. Defect type A for defect 307 and defect type B for defect 305 can be recorded.

[0065] In S503A, the statistical data on the number of bearings appearing for each defect type can be updated. For example, since defects 307 and 305 are identified as target defects, the number of bearings appearing for defect type A can be updated from 0 to 1, the number of bearings appearing for defect type B can be updated from 0 to 1, and the number of bearings appearing for other defect types (e.g., defect type C, defect type D, etc.) can be counted as 0.

[0066] In S504A, it can be determined whether the current bearing is the last bearing in the batch. If the current bearing is not the last bearing in the batch, the process proceeds to S501A. If the current bearing is the last bearing in the batch, the process proceeds to S505A. Here, the last bearing in the batch refers to the last bearing in the batch that needs to be counted, and does not mean that it is necessary to count every bearing in the batch.

[0067] In S505A, statistical data results of the target defect can be generated. For example, the generated statistical data results of the target defect can be reported to the user so that the user can adjust the bearing manufacturing process or production flow. In an exemplary and non-limiting embodiment, the statistical data results may be as shown in Table 1 below.

[0068] Table 1

[0069]

[0070] Alternatively or additionally, in S506A, the statistical data of the current target defect can be compared with the statistical data of the previous target defect to obtain a comparison result. For example, the statistical data of the previous target defect might be for a batch of bearings produced before the adjustment of the bearing manufacturing process or procedure, while the statistical data of the current target defect might be for a batch of bearings produced after the adjustment of the bearing manufacturing process or procedure. By comparing the two statistical results, the user can be assisted in determining whether the adjustments taken by the user are effective.

[0071] Figure 5B This is an example flowchart of another evaluation method for manufacturing defects in a batch of bearings according to embodiments of the present disclosure.

[0072] like Figure 5B As shown, in S501B, the manufacturing defect assessment results of one bearing in a batch of bearings can be obtained. For example, this can be achieved through methods such as... Figure 2 as well as Figure 4 The method for assessing manufacturing defects in bearings shown in the diagram evaluates the manufacturing defects in the bearing.

[0073] In S502B, the defect type of the target defect in the bearing and the quantity of target defects of the corresponding defect type can be recorded. For example, in Figure 3 In the example, defects 307, 303, and 304 might be identified as target defects. The defect type A of defect 307, the number of target defects of defect type A (1), and the defect type C of defects 303 and 304, and the number of target defects of defect type C (2) can be recorded.

[0074] In S503B, statistical data can be updated for the number of bearings appearing for each defect type and the number of target defects for each defect type. For example, since defects 307 and 305 are identified as target defects, the number of bearings appearing for defect type A can be updated from 0 to 1, the number of bearings appearing for defect type C can be updated from 0 to 1, and the number of bearings appearing for other defect types (e.g., defect type B, defect type D, etc.) can be counted as 0; the number of target defects for defect type A can be updated from 0 to 1, and the number of target defects for defect type C can be updated from 0 to 2.

[0075] In S504B, it can be determined whether the current bearing is the last bearing in the batch. If the current bearing is not the last bearing in the batch, the process proceeds to S501B. If the current bearing is the last bearing in the batch, the process proceeds to S505B. Here, the last bearing in the batch refers to the last bearing in the batch that needs to be counted, and does not mean that it is necessary to count every bearing in the batch.

[0076] In S505B, statistical data results of the target defect can be generated. For example, the generated statistical data results of the target defect can be reported to the user so that the user can adjust the bearing manufacturing process or production flow. In an exemplary and non-limiting embodiment, the statistical data results may be as shown in Table 2 below.

[0077] Table 2

[0078]

[0079] Alternatively or additionally, in S506B, the statistical data of the current target defect can be compared with the statistical data of the previous target defect to obtain a comparison result. For example, the statistical data of the previous target defect might be for a batch of bearings produced before the adjustment of the bearing manufacturing process or procedure, while the statistical data of the current target defect might be for a batch of bearings produced after the adjustment of the bearing manufacturing process or procedure. By comparing the two statistical results, the user can be assisted in determining whether the adjustments taken by the user are effective.

[0080] Those skilled in the art will understand that more granular data statistics can be performed on the target defects to provide users with more information for adjusting manufacturing processes or techniques, and all of these configurations are within the scope of this disclosure.

[0081] Figure 6 This is an example schematic diagram of a bearing manufacturing defect assessment apparatus according to an embodiment of the present disclosure.

[0082] like Figure 6As shown, the bearing manufacturing defect assessment device 600 may include a camera 610, a processor 620, and a memory 630.

[0083] Camera 610 can be configured to acquire image data of the bearing.

[0084] The processor 620 can be coupled to the camera 610 and the memory 630, and executes program code stored in the memory 630 to perform the following operations: identify a first number of defects on the bearing based on the image data of the bearing; perform a defect integration operation on the first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number; wherein the defect integration operation includes determining whether to integrate two or more defects from the first number of defects based on the distance between the first number of defects; and determining a target defect on the bearing based on the second number of defects.

[0085] The memory 630 and the processor 620 can be interconnected via a bus system and / or other forms of connection mechanism (not shown). For example, the bus can be a Peripheral Component Interconnect Standard (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The communication bus can be divided into an address bus, a data bus, a control bus, etc.

[0086] For example, processor 620 may be a central processing unit (CPU), digital signal processor (DSP), graphics processing unit (GPU), or other form of processing unit with data processing capabilities and / or program execution capabilities, such as a field-programmable gate array (FPGA). Processor 620 may be a general-purpose processor or a special-purpose processor, capable of controlling other components in bearing manufacturing defect assessment device 600 to perform desired functions.

[0087] Exemplarily, memory 630 may include any combination of one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, erasable programmable read-only memory (EPROM), portable compact disc read-only memory (CD-ROM), USB memory, flash memory, etc. Processor 620 may execute program code stored in memory 630 to implement various functions of bearing manufacturing defect evaluation device 600. Various application programs and various data, as well as various data used and / or generated by the application programs, may also be stored in the computer-readable storage medium.

[0088] For example, the bearing manufacturing defect assessment device 600 may also include input devices such as a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, and gyroscope; output devices such as a liquid crystal display, speaker, and vibrator; storage devices such as magnetic tape and hard disk (HDD or SSD); and communication devices such as network interface cards such as LAN cards and modems. The communication devices allow the bearing manufacturing defect assessment device 600 to communicate wirelessly or wiredly with other devices to exchange data and perform communication processing via a network such as the Internet. A drive is connected to the I / O interface as needed. Removable storage media, such as disks, optical disks, magneto-optical disks, and semiconductor memories, are installed on the drive as needed so that computer programs read from them can be installed into the storage device as required.

[0089] For example, the bearing manufacturing defect assessment device 600 may further include a peripheral interface (not shown in the figure). This peripheral interface can be various types of interfaces, such as a USB interface, a Lightning interface, etc. The communication device can communicate wirelessly with networks and other devices, such as the Internet, intranets, and / or wireless networks such as cellular telephone networks, wireless local area networks (LANs), and / or metropolitan area networks (MANs). Wireless communication can use any of a variety of communication standards, protocols, and technologies, including but not limited to Global System for Mobile Communications (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (W-CDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Bluetooth, Wi-Fi (e.g., based on IEEE 802.11a, IEEE 802.11b, IEEE 802.11g, and / or IEEE 802.11n standards), Voice over Internet Protocol (VoIP), Wi-MAX, protocols for email, instant messaging, and / or Short Message Service (SMS), or any other suitable communication protocol.

[0090] The bearing manufacturing defect assessment device 600 can be, for example, a system-on-a-chip (SOC) or a device including the SOC. For instance, it can be any device such as a mobile phone, personal digital assistant (PDA), tablet computer, laptop computer, desktop computer, part of bearing processing and manufacturing equipment, server, etc., or any combination of data processing devices and hardware. The embodiments of this disclosure do not limit this. The specific functions and technical effects of the bearing manufacturing defect assessment device 600 can be found in the foregoing description of various methods and additional aspects of bearing manufacturing defect assessment according to at least one embodiment of this disclosure, and will not be repeated here.

[0091] Figure 7This is an example schematic diagram of a non-transitory computer-readable storage medium according to at least one embodiment of the present disclosure.

[0092] like Figure 7 As shown, a non-transitory readable storage medium 700 stores computer instructions 710, which, when executed by a processor, perform one or more steps of the various methods and their additional aspects as described above.

[0093] For example, the non-temporarily readable storage medium 700 may be any combination of one or more computer-readable storage media, such as a computer-readable storage medium containing program code for performing the various methods described above.

[0094] For example, when the program code is read by a computer, the computer can execute the program code stored in the computer storage medium to perform one or more steps of the various methods and additional aspects described above, such as those according to at least one embodiment of the present disclosure.

[0095] For example, the non-transitory readable storage medium may include a memory card of a smartphone, a storage component of a tablet computer, a hard disk of a personal computer, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), portable compact disc read-only memory (CD-ROM), flash memory, and other non-transitory readable storage media or any combination thereof.

[0096] The bearing manufacturing defect assessment method, apparatus, and storage medium according to embodiments of this disclosure can identify the type of defect and determine the target defect in a low-cost, fast, and accurate manner based on an image of the bearing. By integrating several defects present in the bearing, more accurate defect identification results can be obtained.

[0097] The text and accompanying drawings are provided by way of example only to aid in understanding this disclosure. They should not be construed as limiting the scope of this disclosure in any way. Although certain embodiments and examples have been provided, it will be clear to those skilled in the art, based on the content disclosed herein, that changes can be made to the illustrated embodiments and examples without departing from the scope of this disclosure.

[0098] Although this disclosure has been described with reference to exemplary embodiments, various changes and modifications may be suggested to those skilled in the art. This disclosure is intended to cover such changes and modifications that fall within the scope of the appended claims.

[0099] Any description in this invention should not be construed as implying that any particular element, step, or function is an essential element that must be included within the scope of the claims. The scope of the patent subject matter is defined only by the claims.

Claims

1. A method for evaluating manufacturing defects in bearings, comprising: Obtain the image data of the bearing. A first number of defects on the bearing are identified based on the image data of the bearing. A defect integration operation is performed on a first number of defects on the bearing to obtain a second number of defects, the second number being less than or equal to the first number. The defect integration operation includes determining, based on the distance between the first number of defects, whether to integrate two or more defects from the first number of defects. Based on the second number of defects, the target defect on the bearing is determined.

2. The method according to claim 1, wherein, The defect integration operation includes: Based on the fact that the distance between two or more adjacent defects in the first number of defects is less than the integration distance threshold, it is determined that the two or more adjacent defects in the first number of defects will be integrated.

3. The method according to claim 2, wherein, The defect integration operation includes: Determine the integrated auxiliary attributes of each defect in the first number of defects. The integration distance threshold of two or more defects is increased based on the high similarity of the integration auxiliary attributes of two or more defects in the first number of defects, and the integration distance threshold of two or more defects is decreased based on the low similarity of the integration auxiliary attributes of two or more defects in the first number of defects.

4. The method according to claim 2, wherein, The defect integration operation includes: Determine the integrated auxiliary attributes of each defect in the first number of defects. Based on the distance between adjacent defects in the first number of defects and the integration auxiliary attributes of each defect in the first number of defects, multiple defects are clustered using a clustering model, and defects belonging to the same cluster are integrated.

5. The method according to claim 3 or 4, wherein, The integrated auxiliary attributes include at least one of the following: the curvature of the defect's edge, the grayscale of the defect's pixels, and the uniformity of the defect's pixels.

6. The method according to claim 1, wherein, Determining the target defect on the bearing includes: determining the defect type of the second number of defects using a trained model based on a second number of defects.

7. The method according to claim 6, further comprising: Based on the size of each defect in the second number of defects, the target defect on the bearing is determined.

8. The method according to claim 6, wherein, Determining the target defect on the bearing further includes: for each defect type, Determine the defect size threshold corresponding to this defect type. In response to one or more defects in the second number of defects having a size greater than a defect size threshold corresponding to the defect type, the one or more defects are identified as candidate target defects associated with the defect type.

9. The method according to claim 8, wherein, Determining the target defect on the bearing further includes: for each of the second number of defects, Determine the defect auxiliary attributes of this defect. If the defect auxiliary attribute of the defect matches the defect type well, the defect size threshold for the defect is reduced; if the defect auxiliary attribute of the defect matches the defect type poorly, the defect size threshold for the defect is increased.

10. The method according to claim 9, wherein, Determining the target defect on the bearing also includes: Determine the degree of deviation between each candidate defect and its corresponding defect size threshold, and select one or more defects from the candidate defects as the target defect based on the degree of deviation.

Images