Inspection system for yarn bobbins and method for inspecting yarn bobbins

Through artificial intelligence and deep learning systems combined with image acquisition devices, a database is established to automatically identify and classify various defects of yarn bobbins, the problem of low detection efficiency in the existing technology is solved, efficient defect identification and classification is achieved, and the improvement of the improvement of Product quality management.

CN118946803APending Publication Date: 2024-11-12SANCO TEXTILE ENTERPRISE IND & TRADING CO BASPINAL BRANCH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202280094167.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2022-01-28
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The prior art relies on human eye examination or simple image processing in yarn bobbin inspection, making it difficult to effectively detect a variety of defects, especially complex, non-elastic defects and wear defects, and the detection efficiency is low and difficult to automate.

Method used

Using artificial intelligence and deep learning systems combined with image acquisition devices, a database is established by obtaining image data of yarn bobbins, and the application is used to automatically identify and classify different types of defects, including non-elastic defects, damaged bobbin defects, and wear defects. , color defects and diameter defects, and mark the defect location on the monitor.

Benefits of technology

It realizes automatic identification and classification of various defects of yarn bobbins, improves detection efficiency, reduces manual intervention, enhances the visualization and data management of defects, and ensures the accuracy of product quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118946803A_ABST
    Figure CN118946803A_ABST
Patent Text Reader

Abstract

An inspection system for yarn bobbins is provided. A system capable of improving automation of yarn bobbin inspection should be provided. The system comprises: an image acquisition device (5) for acquiring an image of a yarn bobbin (12); a database (3) in which a first data set regarding at least a first flaw type is stored, the first data set being generated by using samples of yarn bobbins (12) having the first flaw type; and an application program (4) for determining whether the yarn bobbin (12) checked by the image acquisition device (5) has the first defect type. In addition, a method for inspecting yarn bobbins and assigning different flaw types to the yarn bobbins is provided.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] The invention relates to an inspection system for yarn bobbins, with which the yarn bobbins can be inspected to search for defects. In addition, a corresponding method for inspecting such yarn bobbins is provided.

[0002] Until today, in industrial processes for yarn production, humans have inspected the yarn bobbins with their eyes to search for specific flaws or defects in the yarn bobbins. If such flaws or defects are present, the corresponding bobbins are marked or removed from the bobbins to be sent to the customer. However, some techniques are disclosed in scientific papers on how inspection can be done by physical measurements, in particular by image inspection systems. For example, in the scientific paper Automatic Bobbin Inspection System (J. Baumgartinger, R. Lenzinger Berichte, 82 (2003) 70-75), and in the scientific paper An Innovative Solution for Abrage Fault Detection on Yarn Bobbin and Fabric Surface (HICelik, E. Gultekin, LCDulger, HISunbul, H. Kani, ICENS, 5 th This prior art was described at the International Conference on Engineering and Natural Science, June 12-16, 2019, Prague).

[0003] In the latter paper, a generic prototype machine vision inspection system using an illumination unit and a camera is described. The illumination unit irradiates light with a specific temperature onto the surface of the yarn bobbin in order to make so-called wear defects (defects from wear) more visible. These defects are, for example, on the surface of the yarn bobbin and are caused by wear. It is shown that by varying the brightness and adjusting the image processing of the acquired images, the corresponding wear defects (wear imperfections) can be made more visible.

[0004] Scientific paper Automatic Bobbin Inspection System (J.Baumgartinger & R. 2003) describes the different commonly known defects of yarn bobbins in terms of geometry, macrostructural defects and microstructural defects. In particular, it is described how the detection of broken filaments can be performed. A CCD camera records and analyzes the laser beam reflected on the surface of the yarn bobbin as it rotates, and when the reflected beam meets preset conditions, it is determined that the broken filament is located on the corresponding surface. These surfaces are the end faces of the bobbin in the length direction of the bobbin. The system requires great effort to set up the corresponding boundary conditions, etc., in order to detect broken filaments.

[0005] In view of said prior art documents, the present inventors applied a new approach and found that a system working with artificial intelligence and / or deep learning can be used to determine whether a specific defect is present at the yarn bobbin.

[0006] In view of this, an inspection system is provided having the features defined in claim 1. In a first step, a sample of yarn bobbins having a specific defect (or defect) type is used to acquire images of the respective yarn bobbins via an image acquisition device. Data is obtained from the acquired images and stored in a database.

[0007] Thus, a plurality of yarn bobbins with corresponding defect types are used to establish a first data set in a database. Then, later in the industrial process, when further evaluating the yarn bobbins for the presence of a specific defect type, this defect can be automatically determined or derived. It can then be decided whether this defect type is present.

[0008] That is, it is decided by the application program whether the corresponding data acquired by the image acquisition device corresponds or substantially corresponds to the data in the first data set. If this is the case, it is decided that the first defect is present. In order to have a valid first data set in the database, it is beneficial to generate the data set using samples of yarn bobbins with the specific defect type, which is named "first defect type", wherein the first defect type is respectively arranged at different areas or has different aspects. Since the various aspects of the first defect type are extensive in the corresponding different samples, the corresponding first data set can be better used to identify whether the first defect is contained in the yarn bobbins subjected to inspection during production.

[0009] The inspection system may have an image acquisition device for acquiring an image of the yarn tube. The image data are compared in the application with a first data set stored in the database, and the application determines whether the yarn tube inspected by the image acquisition device has a first defect type.

[0010] According to a further development, a second data set is stored in the database, the second data set corresponding to the second defect type. The second data set can also be established by using samples of yarn bobbins with the second defect type. The aspects described above for the first data set generation can also be applied to the second data set generation.

[0011] Advantageously, the sample yarn bobbins used to create the respective first and second data sets only have the respective first defects and / or second defects.

[0012] The first data set and the second data set do not need to be provided in different parts of the database, but can both be stored together in one part of the database. Corresponding data related to the first data set and the second data set may have corresponding identifiers assigning them to the first defect type or the second defect type. Samples having both defect types may also be used to establish the data set. The data may have identifiers corresponding to the first defect type and the second defect type.

[0013] Therefore, with the system it can also be decided whether the defect in the yarn bobbin to be checked is of the second defect type.

[0014] The corresponding yarn bobbins with these first defect types and / or second defect types can be removed from the batch of bobbins to be shipped to the customer. The corresponding bobbins with defects can be sorted to form batches of yarn bobbins with the first defect type, yarn bobbins with the second defect type, and yarn bobbins with both defect types.

[0015] The respective batches can be placed on different pallets etc. The respective defect type can be an inelastic defect, a damaged tube defect, a wear defect, a color defect or a diameter defect.

[0016] Inelastic defects should be understood as follows. Some yarns (e.g., yarns used to produce denim) may have a certain elasticity. In order to ensure this elasticity, at least one elastic filament may be provided in the yarn to adjust the corresponding properties. Additional filaments without elasticity may be provided, and these filaments may be twisted together during ring spinning to form a core. The core may be encapsulated by a covering layer, which includes, for example, staple fibers such as natural fibers. However, during the spinning (particularly ring spinning) process, it may occur that the elastic filaments contained in the core break, so there is an area in the core where no elastic filaments are provided. This area is called an inelastic zone. Therefore, an inelastic defect is a defect in which there is a zone along the yarn in which there is an area in which no elastic is provided in the yarn containing an elastic. The elastic is not limited to a material corresponding to the trademark "Elastane", but may be provided by any elastic filament such as Lycra. The inventors have found for the first time that such inelastic defects can be detected by a specific image acquisition device and by using artificial intelligence.

[0017] The damaged bobbin defect is a defect related to any damage to the bobbin itself. In contrast to the term yarn bobbin, the term bobbin is used in this application only for the intermediate shaft around which the yarn is wound. Thus, a yarn bobbin includes the yarn wound around the bobbin. In the following, whenever the term bobbin is used, it is meant to refer to the shaft around which the yarn is wound. In the following, when the term yarn bobbin is used, it is meant to include the complete arrangement of the yarn wound around the bobbin.

[0018] The corresponding bobbin may be broken or have any other damage. This can be seen, for example, when the image acquisition system is focused on the front face of the yarn bobbin in the longitudinal direction.

[0019] Wear defects are defects caused by the wear of yarn and bobbin during handling.

[0020] A color defect is a defect in the color of the yarn. In a specific production process, the yarn may be colored. The corresponding result should correspond to a standardized color code. This can be the case, for example, by depicting the color code on the inside of the bobbin via a sticker. A corresponding system can acquire the corresponding color code and can decide whether the detected color of the yarn matches this color code. Thus, the corresponding color that meets the customer's order can be ensured for the corresponding yarn bobbin.

[0021] A diameter defect is a defect in the diameter of a yarn bobbin. For each batch of bobbins, there is usually a predetermined diameter, such as the diameter ordered by a customer. Therefore, in order to manage the respective yarn bobbins to be packaged and shipped to the customer, an inspection system is used to determine the diameter defect.

[0022] The same inspection system may be utilized to determine any of the defects individually or in any combination.

[0023] This means that, in addition to inelasticity defects, damaged tube defects and / or wear defects and / or color defects and / or diameter defects can also be detected.

[0024] For the corresponding defects, different data sets are generated using sample yarn bobbins and saved in a database.

[0025] Based on the database, it is decided which stored data corresponds to the data of the acquired image, so that a corresponding allocation of the different defects is carried out.

[0026] The application is configured to determine a type of defect of a yarn bobbin inspected using the image acquisition device.

[0027] The inventors have also noticed that the application can advantageously have a defect assignment module and additionally a defect marking module. With the defect assignment module, the defect type in the yarn bobbin inspected with the image acquisition device is assigned to the corresponding yarn bobbin. With the defect marking module, data are generated to provide a marking that can be superimposed on the image of the yarn bobbin acquired by the image acquisition device.

[0028] With such a configuration, a person operating the inspection system (operator) can better identify the respective defect and its location on the respective bobbin.

[0029] Advantageously, the marking is defined by a cluster of pixels, and the size and form of the cluster of pixels corresponds to the area of ​​the defect.

[0030] Different defect types can be assigned to different colors for marking. For example, the color of the pixel showing the inelastic defect can be red, the color of the pixel showing the damaged bobbin defect can be green, the color of the pixel showing the wear defect can be blue, and the color of the pixel showing the diameter defect can be yellow. The corresponding mark can be superimposed on the corresponding image of the corresponding yarn bobbin acquired by the image acquisition device, and the superimposed image can be visualized on the display. Then, the corresponding person operating the system can view the position assigned to the corresponding defect on the display. In order to further facilitate the operator to determine the position of the defect, it is beneficial that when on the display, the image of the yarn bobbin (without superimposed marks) is depicted next to the image of the yarn bobbin (with superimposed marks) and both images have the same size (magnification). Compared with in the image without marks, the position of the defect is more obvious.

[0031] Although the corresponding transport of the yarn bobbins through the system is not defined, it is seen as beneficial that the corresponding system comprises a bobbin holder mounted on a guide rail. The bobbin holder can be moved along the guide rail through the inspection area (which is the area where the image acquisition is performed).

[0032] From the batch of bobbins to be inspected, the individual yarn bobbins can be placed manually by a user (operator) or automatically by a robot into the corresponding bobbin holder. Via the bobbin holder, the yarn bobbins are moved through the inspection area. Thereafter, the yarn bobbins can be taken out of the yarn holder manually by a person or automatically by a robot and can be automatically distributed to a tray storing bobbins with corresponding defects and / or a tray storing bobbins without defects.

[0033] The allocation can be done automatically or by a corresponding light system with red or green light. Green light corresponds to a yarn bobbin without defects, so this yarn bobbin can be allocated by the operator to the package of yarn bobbins that should be sent to the customer. Red light corresponds to yarn with defects. In addition, different light types can be used to allocate yarn bobbins to different defect types.

[0034] According to a further development, the yarn bobbin holder is adapted to rotate the yarn bobbin along an axis which is perpendicular to the bobbin center axis and which may also be perpendicular to the direction in which the guide rail extends. Advantageously, each yarn bobbin moves along the guide rail in its longitudinal direction. However, in the inspection area, it is advantageous to rotate each yarn bobbin by 90° so that the respective axis of the bobbin is perpendicular to the direction in which the guide rail extends.

[0035] With such a configuration, since the respective image acquisition devices are usually mounted on different sides and the different sides sandwich a guide rail extending between the sides in the longitudinal direction, each end face in the longitudinal direction of the bobbin can be better inspected. It is also possible to perform rotation of the yarn bobbin during image acquisition to acquire images at different regions of the yarn bobbin.

[0036] The corresponding bobbin holder can be configured as a ring with a fixing area, in which the yarn bobbin can be placed. The yarn bobbin should be placed so that a relative movement of the yarn bobbin is prevented. This can be a movement in the circumferential direction, or this can be a movement in the translational direction of the yarn bobbin. It is beneficial to prevent relative movement in both the circumferential direction and the translational direction.

[0037] The fixing area can be defined by two parallel extension rods extending in the direction of the ring center axis of the ring defining the bobbin holder. The two parallel extension rods can have a gap between them to hold the bobbin at its peripheral surface. With such a configuration, it is easy to place yarn bobbins with different diameters in the same holder.

[0038] The respective ring of the bobbin holder may have a swivel hinge mounted to the guide rail via a movable element. In a particular inspection system, it is advantageous that a plurality of such bobbin holders are arranged one after another along the guide rail, and it may be advantageous that the bobbin holders are adapted to move together along the guide rail. This means that the respective bobbin holders may each be connected to an adjacent bobbin holder by a chain or cable or other coupling, so that when the chain or cable or other coupling is pulled along the respective guide rail, the bobbin holders move through the system at a defined distance from each other.

[0039] It is further advantageous that the system comprises a darkroom having two openings through which the guide rail extends from the entrance and the exit. One or more image acquisition devices may be disposed within the darkroom. When such a darkroom is used, ambient light is prevented from interfering with image acquisition and from generating artifacts.

[0040] However, it is not mandatory to use such a darkroom, in particular when the image acquisition device does not operate in the visible spectrum but in the infrared or X-ray spectrum.

[0041] Advantageously, the corresponding image acquisition device comprises an illumination section for illuminating a region of the yarn bobbin for defect inspection and a camera device for acquiring an image of the region. Via the illumination section, the corresponding region can be illuminated, and via the camera, the illuminated region can be photographed.

[0042] Additionally, the system may include image processing and adjustment modules.

[0043] Via the image processing and adjustment module, each acquired image can be filtered. The corresponding filtering can provide better data quality, wherein the corresponding defect type is easier to detect.

[0044] It has proven to be advantageous to use different image acquisition devices for each different defect type. This is because different image acquisition conditions can be preferred for different defect types. On the one hand, different filters can be applied for different defect types using the image processing and adjustment module. On the other hand and / or in addition, the respective camera devices used are sensitive to different wavelengths and / or the respective illumination units irradiate radiation with different wavelengths and / or with different wavelength ranges and / or color temperatures.

[0045] In particular, the illumination is operated at a specific wavelength or color temperature depending on the type of defect to be detected.

[0046] Depending on the type of defects to be detected, different filters can also be used in the image processing and adjustment modules.

[0047] The respective image acquisition device used may be one or more of the following: an image acquisition device for inelasticity defects, an image acquisition device for damaged tube defects, an image acquisition device for wear defects, an image acquisition device for color defects and an image acquisition device for diameter defects.

[0048] Corresponding defect types have already been explained in the previous section. Corresponding differences in the image acquisition devices may be that they are operated at different wavelengths and / or different color temperatures and / or the corresponding camera devices may have different sensitivities.

[0049] In particular, for the image acquisition device for inelastic defects, it is considered to be beneficial that the illumination part has a ring form and the camera device is arranged in the central area of ​​the ring. The ring-shaped illumination part is installed so that it irradiates the corresponding light beam to the corresponding end face in the longitudinal direction of the yarn bobbin. Therefore, the corresponding central axis of the ring of the illumination part can be arranged perpendicular to the extension direction of the guide rail.

[0050] In particular, it is considered to be beneficial to arrange two image acquisition devices for inelastic defect detection sandwiching the guide rail. Advantageously, the two image acquisition devices are aligned centrally relative to each other, and the center of the alignment is orthogonal to the extension direction of the guide rail.

[0051] Furthermore, it is considered to be advantageous to provide at least one further image acquisition device upstream and downstream of the image acquisition device for inelastic defects. Such image acquisition devices may be image acquisition devices for wear defects, color defects, diameter defects, etc.

[0052] The system may also be configured to automatically sort yarn bobbins fed to the system according to the type of defect detected.

[0053] The system may also be configured to periodically store data acquired during defect detection in a database after confirmation by a person (operator) that the defect actually detected corresponds to the defect type assigned to the yarn bobbin from the system.

[0054] In view of this, it is beneficial to have a display unit and markings on the superimposed image. Thus, the respective person operating the inspection system can instantly feed the database and enter it with more data (wider data set) to have a better assignment to defect types.

[0055] Typically, the system is configured to perform defect assignment to different defect types via artificial intelligence, and / or in particular the system works with deep learning.

[0056] In addition to the above inspection system, a method for inspecting yarn bobbins and assigning different defect types to yarn bobbins is provided. The method uses an application and a database for defect assignment. The corresponding aforementioned aspects described with respect to the system can also be provided for the corresponding inspection method.

[0057] The method may be an automated method performed on a computer.

[0058] Furthermore, a computer-readable application program can be provided which automatically performs a corresponding bobbin inspection and / or controls a corresponding system.

[0059] In the following, some embodiments of the invention are described with reference to the accompanying drawings.

[0060] In this article,

[0061] Figure 1 A schematic diagram of a portion of a system including an inspection area disposed in a darkroom is shown.

[0062] Figure 2 The layout of the inspection area within the darkroom is shown in detail.

[0063] Figure 3a , Figure 3b , Figure 3c and Figure 3d shows that an inelastic defect is detected and a corresponding image with a marker superimposed is generated ( Figure 3a , Figure 3b , Figure 3c , Figure 3d The corresponding right side of the image).

[0064] Figure 4a and Figure 4b An image generated by wear defect detection is shown. On the right side of the figure, wear defect areas are assigned by markings to areas on the end side in the longitudinal direction of the respective yarn bobbin.

[0065] Figure 5a , Figure 5b and Figure 5c Shows the damaged bobbin defect and the Figure 5a-5c The corresponding marks on the corresponding right side are superimposed onto the acquired image.

[0066] Figure 6 A schematic configuration of the entire system is shown.

[0067] exist Figure 6On the right side of the figure, a computer 1 and a display 2 are shown. The computer 1 may have a database 3 stored in a storage area and an application 4 running on the computer. An image acquisition device 5 is connected to the computer 1, and data can be exchanged between the image acquisition device 5 and the computer 1. Via the interface, data can be input into the corresponding database 3. This is done by Figure 6 The box "Data Input" in FIG. 4 is schematically shown. The image acquisition device 5 and / or the tube selection component for selecting the corresponding yarn tube after the defect determination and / or the conveyance of the yarn tube 12 along the guide rail 8 described later can be controlled by a control module in the computer 1. The control module can also be integrated in the application 4. However, in Figure 6 In the figure, the control module is depicted as a separate independent module in the computer 1.

[0068] exist Figure 6 On the left side of the figure, a schematic configuration of a darkroom 6 is shown, in which, in a specific case, an image acquisition device 5 is installed. A guide rail 8 extends from an entrance opening 9 of the darkroom 6 to an exit opening 10 (e.g. Figure 1 (also in Chinese).

[0069] First, the yarn bobbin 12 to be evaluated is placed on a holder 11, which is not shown in the figure (but is Figure 1 and Figure 2 The holder 11 is mounted on the guide rail 8. Then, the yarn bobbin 12 mounted on the holder 11 is moved along the guide rail 8 into the inspection area in the darkroom 6. At the inspection area, the image acquisition device 5 acquires an image of the corresponding bobbin. In certain cases, it may be beneficial to capture an image from a corresponding end face that delimits the yarn bobbin in the longitudinal direction of the yarn bobbin. The end face is Figure 6 13. However, the inspection is not limited to the corresponding end faces and can also be performed at the corresponding outer peripheral surface. In addition, when the corresponding image acquisition device operates at a wavelength or wavelength region that penetrates the yarn bobbin 12, the inspection of the inside of the yarn bobbin can be performed.

[0070] However, in certain cases, the respective image acquisition device comprises a camera device 14 and an illumination section 15. Via the illumination section 15, radiation having a specific wavelength or a wave comprising a specific wavelength band or having a specific color temperature is illuminated toward the respective yarn bobbin 12.

[0071] An image of the illuminated yarn tube 12 is captured with a corresponding camera device 14. The corresponding image data are transmitted to the computer 1. In a corresponding image processing and adjustment module 16, which can be part of the computer 1, corresponding filtering rules can be applied depending on the corresponding defects to be detected. The corresponding filtered data are fed to the application 4. In the application 4, it is decided which of the multiple defect types is provided at the evaluated yarn tube 12. After the corresponding tube inspection, the yarn tube 12 is moved out of the darkroom 6 through the outlet opening 10 and the yarn tube 12 can be selected and the corresponding yarn tube 12 without defects or without significant defects can be assigned to the corresponding batch of yarn tubes to be shipped to the customer.

[0072] The corresponding image acquisition device 5 may have a wear detection function, a tube detection function, a color detection function, a diameter detection function and an inelastic zone detection function. With the corresponding functions, the corresponding defects can be detected.

[0073] exist Figure 6 In the particular case shown, for one image acquisition device 5 shown therein, the corresponding functions are shown as alternatives or can exist simultaneously. The corresponding functions can be provided in one image acquisition device, however, separate image acquisition devices can be provided, wherein each image acquisition device includes a camera device 14 and a corresponding lighting section 15. In each respective image acquisition device, the corresponding camera device and the lighting section can have different specific configurations related to the illumination wavelength or wavelength region or sensitivity. This will be explained in the following section.

[0074] However, image data are generated by the image acquisition device 5 , which are fed to the application 4 .

[0075] In the application 4, the corresponding data are compared with the data sets stored in the database 3. The database 3 is built up by feeding the database 3 with data of the corresponding defect type. To this end, in a first step of starting the corresponding inspection system, the yarn bobbins 12 with the corresponding defect type are inspected with the system and the corresponding data are assigned to the specific defect type. To this end, the data can have an identifier that identifies the specific defect. This feeding of yarn bobbins with different defect types can be carried out for different defects. For example, 10 to 100 yarn bobbins with inelastic defects are fed into the system and the corresponding data derived with the image acquisition device are fed into the database.

[0076] Another 10 to 100 yarn bobbins with broken bobbin defects are used to generate bobbin defect data in the database. The same operation can be performed for sample bobbins with wear defects, bobbins with different specific diameters, and bobbins with specific colors.

[0077] This training of the database and the addition of further data to the database can be carried out regularly after a specified period of time or a specified number of yarn tubes inspected, or simultaneously during the inspection of the respective yarn tubes to be evaluated. To this end, it is necessary to match the respective identified defects (which are identified by the system) and to see whether this defect corresponds to the respective defect that should be identified.

[0078] This can be done by a human.

[0079] This can be carried out after the corresponding yarn bobbin comes out from the outlet opening 10 and the corresponding person (operator) looks at the defect and decides whether this defect corresponds to the defect identified by the system identification. In this case, the corresponding data are fed to the database as the data corresponding to this defect type.

[0080] This assignment can also be done with the aid of a display device, which a person views. In order to help the operator specify whether a respective detected defect corresponds to a respective defect actually present in the respective yarn bobbin, a colored marking or at least a marking can be provided which is superimposed on the acquired image.

[0081] The acquired image may be any image, which may be an X-ray image that can inspect not only the surface of the yarn bobbin but also the interior of the yarn bobbin. However, a corresponding mark may be generated by a plurality of pixels corresponding to the area where the defect is detected.

[0082] Thus, by comparing an image without pixels and an image with pixels when the images are placed adjacent to each other on the screen, the operator can confirm whether the detected defect corresponds to the type of defect assumed to be detected, and in this case, the corresponding data can be assigned to the database and become an identifier characteristic of the defect type in question.

[0083] Through this configuration, the corresponding database 3 and application 4 can be trained.

[0084] Application 4 is an application that uses artificial knowledge and searches for similarities between the acquired images and images in the database with specific defects. In order to have a good database, it is beneficial to have sample yarn bobbins that are very different in terms of defect location, degree of defect or size of defect for a specific defect, such as inelasticity. Therefore, there is a strong diversity in the differences between the aspects of a specific defect type. Therefore, it is easier to make a more reliable assignment of the acquired images of the bobbins to be evaluated. The wider the diversity of specific defect aspects (e.g., the diversity of inelasticity defects), the more reliable the generated data and suitability for use with the application.

[0085] Figure 1 Shows Figure 6 An example of a configuration is shown on the left side of the system. Figure 1 A darkroom 6 is shown enclosing the inspection area. A corresponding guide rail 8 extends from an entrance opening 9 to an exit opening 10. The corresponding opening may have a corresponding opening area of ​​from / to 22, which is 150% larger than the cross-sectional area of ​​the corresponding yarn bobbin perpendicular to its longitudinal axis. Such a reduced size opening prevents too much light from entering the inspection area.

[0086] Also available in Figure 2 8. Suspended therefrom is a ring 17 corresponding to the bobbin holder 11. At the lower side of the ring 17, opposite to the side on which the rotary hinge 19 is provided, two parallel-extending rods 18 are provided, between which a gap 20 is provided. The respective outer peripheral surface of each yarn bobbin 12 can be fixed in the gap 20. Thus, a translational movement of the bobbin can be prevented. However, the yarn bobbin itself can still rotate relative to its center axis.

[0087] The respective ring 17 is mounted via a swivel hinge 19 to a movable element arranged at the underside of the guide rail 8, which movable element is not visible in the figure, and the respective ring can be Figure 1 and Figure 2 The rotations are shown in the order shown.

[0088] That is, the bobbin direction (axial direction) corresponds to the longitudinal direction of the guide rail when the respective yarn bobbin 12 enters the darkroom 6. In the inspection area, the bobbin is rotated (e.g. automatically by a control module), in a specific case by about 90°, so that the front and rear faces of the respective yarn bobbin 12 can be inspected. However, the rotation of the yarn bobbin can also be performed during image acquisition in the inspection area.

[0089] exist Figure 1 and Figure 2 In the particular configuration shown, the elements with reference numerals 23a, 23b correspond to the illumination part of the image acquisition device for identifying elastic defects of the yarn bobbin. The elements with reference numerals 24a, 24b correspond to the camera device of said respective image acquisition device for determining elastic defects. The respective camera device 24a, 24b is arranged in a central area along the central axis of the illumination part 23a, 23b. The illumination part is defined by a ring and illuminates the respective area of ​​the yarn bobbin with a specific wavelength or wavelength band or light temperature.

[0090] Considering the good ability to detect inelastic defects, the corresponding illumination wavelength and / or the corresponding sensitivity and camera device are selected. In this specific case, for inelastic defects, two image processing and image acquisition devices are arranged on opposite sides of the corresponding yarn bobbin 12 and sandwich the guide rail 8.

[0091] In the particular case of this example, downstream of this image acquisition device for inelastic defect detection, an image acquisition device for wear detection and an image acquisition device for color defect detection are arranged. The image acquisition device for color defect detection has the reference numeral 5b. The image acquisition device for wear defect detection has the reference numeral 5c. Each device comprises a corresponding illumination section and a camera section. In particular, the corresponding illumination for wear detection has a plate-like shape, and the camera is arranged to observe the end face of the corresponding yarn bobbin through the hole of the plate-like shape, so that a plane wave illumination is provided.

[0092] The illumination section for color defect detection has a bell-like shape so that the illumination beam illuminates the sample at different angles. Furthermore, in the case of the image acquisition device for color defect detection, the corresponding camera device 14 is held concentrically.

[0093] After the respective measurement has been performed and the respective holder 11 has been moved through the examination area, the respective holder 11 is rotated back to the original orientation and then the respective holder 11 is moved out of the system through the exit opening 10 .

[0094] At this point, the yarn bobbins are removed from the holders and bobbin selection is performed, for example, in an automated manner, such as Figure 6 This automatic selection can also be controlled by the control module.

[0095] 3 , 4 and 5 respectively, it is shown how the corresponding data are processed in the application 4 so that the corresponding markings are generated. For this purpose, the application 4 has a defect marking module 26 in addition to the defect assignment module 25 .

[0096] In the defect assignment module 25, the respective defect assignment is done by correspondingly matching the acquired image data with the data in the database 3 via artificial intelligence. Via the respective defect marking module 26, a marking is generated, which in the present case may be a specific set of pixels corresponding to the area where the respective defect is detected.

[0097] The mark is superimposed on the image and can be shown on the display 2 (eg Figure 3a , Figure 3b , Figure 3c and Figure 3d 3 to 5 are images without a mark and the corresponding area of ​​the image with a mark (the left image in FIGS. 3 to 5 is an image without a mark and the corresponding right image is an image with a mark superimposed, the right image being at the same position and with the same magnification as the image without a mark). Therefore, when the user views the display 2, he can see at which position in the corresponding area in the original image the corresponding defect is detected and how the defect looks.

[0098] Figure 3a to Figure 3d The corresponding defect shown in corresponds to an inelastic defect. An inelastic defect has specific characteristics, e.g. Figure 3c This can be best seen in Figure 3c As can be seen in the area below (a corresponding mark is shown in this area of ​​the right-hand image) on the left side of FIG. 1 , the corresponding yarn does not extend in the circumferential portion, but has a wave. From this wave form it can be inferred that no elasticity is provided in the core at said portion of the yarn, since the yarn is loosely suspended there, which gives rise to the appearance of a wave form.

[0099] The inventors have demonstrated for the first time that specific inelastic defects can be determined by means of artificial intelligence and that corresponding labels can be assigned to these defects.

[0100] For example, by Figure 3a to Figure 3d and Figure 4a and Figure 4b By comparison, the image acquisition device for acquiring inelastic defects can have a higher resolution than the image acquisition device for wear defects, wherein Figure 4a and Figure 4b Resulting images obtained by an image acquisition device for wear detection are shown (left and right images of FIG. 4 , with and without markings).

[0101] However, the abrasion test is performed in the same way and Figure 4a , Figure 4b In the figure, the left side shows each image without the mark superimposed on the image, and the right side shows the mark superimposed on the image. In this case, the mark is red, while the mark showing the inelastic defect is yellow.

[0102] Therefore, although in the present case, each defect image is shown separately on the corresponding screen, these images can also be overlapped, and the defects can be marked with different colors.

[0103] Figure 5a , Figure 5b and Figure 5c Different defects of the tube itself are shown. Moreover, in the case of the figures, the images depicted on the left have superimposed markings, while the images on the right do not have markings. Figure 5a In the case of , the bobbin is broken and the break line is shown as a mark. Figure 5b In the example, part of the bobbin is missing.

[0104] exist Figure 5c In the case of a 3.0-inch 4-inch 2.0 ...4.0

[0105] Since the acquired image is shown on the left and the image with the corresponding marking superimposed is shown on the right, and since the two images are shown next to each other on the display 2, when the two images have the same size and are visualized next to each other, the respective user can identify whether the respective defect corresponds to the assumed defect automatically identified by the system. In this case, the user assigns the respective data to be entered into the database and the data can be used to improve the database and thus for subsequent defect determinations.

[0106] The present invention is not limited to the specific configuration described above, but in the broadest sense can generally be applied to determine defects in yarn bobbins by matching acquired data with pre-stored data in a database for corresponding defect types.

[0107] In particular, the present invention relates to systems and methods using artificial intelligence and using deep learning, which are trained with further data during processing.

[0108] In a specific configuration, corresponding defects can be detected automatically and a sorting of the yarn bobbins into different batches can take place.

[0109] In addition to the specific system, a corresponding inspection method and / or an application program can also be provided, with which the corresponding method can be carried out on a computer.

[0110] Although in a specific embodiment it is described that the corresponding application and the database are provided on one computer. It may also be the case that the database is provided remote from the application and the application is an application executable on a PDA or another remote terminal device, wherein the application is an application executable on the Internet. The corresponding control of the application may also be performed remote from the corresponding facility where the corresponding inspection of the yarn bobbin is performed. The corresponding inactivation of the different elements may be accomplished by wireless communication or alternatively via a local area network.

[0111] Reference numerals list

[0112] 1. Computer

[0113] 2 Displays

[0114] 3 Database

[0115] 4 Applications

[0116] 5 Image acquisition device

[0117] 5b Image acquisition device for color defect detection

[0118] 5c Image acquisition device for wear detection

[0119] 6 Darkroom

[0120] 8 rails

[0121] 9 Entrance openings

[0122] 10 Exit openings

[0123] 11 Retainer

[0124] 12 yarn bobbins

[0125] 13 End face

[0126] 14 Camera Device

[0127] 15 Lighting Department

[0128] 16 Image processing and adjustment modules

[0129] 17 rings

[0130] 18 shots

[0131] 19 Rotating hinge

[0132] 20 gaps

[0133] 23a, 23b are lighting parts of an image acquisition device for identifying elastic defects of a yarn bobbin.

[0134] 24a, 24b Camera device of an image acquisition device for identifying elastic defects of a yarn bobbin

[0135] 25 Defect Allocation Module

[0136] 26 Defect marking module

Claims

1. A yarn bobbin inspection system, comprising: An image acquisition device (5) for acquiring an image of the yarn bobbin (12), a database (3) in which at least a first data set about a first defect type is stored, said first data set being generated by using a sample of yarn bobbins (12) having the first defect type, and An application program (4) for determining whether a yarn bobbin (12) inspected by means of the image acquisition device (5) has the first defect type.

2. The inspection system according to claim 1, characterized in that: The inspection system is configured to store in the database (3) at least a second data set regarding a second defect type, the second data set being generated by using a sample of yarn bobbins (12) having the second defect type.

3. The inspection system according to claim 2, characterized in that: The application (4) is configured to determine whether the yarn bobbin (12) inspected by means of the image acquisition device (5) has the first defect type and / or the second defect type.

4. The inspection system according to any one of claims 1 to 3, characterized in that: The defect type is selected from the group consisting of: an inelastic defect, a damaged tube defect, a wear defect, a color defect, and a diameter defect.

5. The inspection system according to claim 3 or 4, characterized in that: The inspection system is configured to store one or more further data sets in the database (3), each further data set relating to at least one further defect type, the respective data set of the further defect type being generated by using samples of yarn bobbins (12) having the further defect type.

6. The inspection system according to any one of claims 2 to 5, characterized in that: The application (4) is configured to determine the type of defects of the yarn bobbin (12) inspected by means of the image acquisition device (5).

7. The inspection system according to any one of claims 1 to 6, characterized in that: The application (4) has a defect assignment module (25) which assigns the defect type of the yarn bobbin (12) inspected by the image acquisition device (5) to the yarn bobbin (12), and A defect marking module (26) generates data for providing a marking which can be superimposed on an image of the yarn bobbin (12) acquired by the image acquisition device (5).

8. The inspection system according to claim 7, characterized in that: The marking is defined by a cluster of pixels whose size and form correspond to the area of ​​the defect.

9. The inspection system according to any one of claims 7 or 8, characterized in that: The system is configured to use different colors for the marking for different defect types.

10. The inspection system according to any one of claims 7 to 9, characterized in that: An overlay image is visualized on a display (2) in which the marking is superimposed on an image of the corresponding yarn bobbin (12) acquired by means of the image acquisition device (5).

11. The inspection system according to any one of claims 1 to 10, characterized in that: The system further comprises a bobbin holder (11) mounted on a rail (8) and movable along the rail through an inspection area, which is the area where image acquisition is performed.

12. The inspection system according to claim 11, characterized in that: The bobbin holder (11) is adapted to rotate the yarn bobbin (12) along an axis perpendicular to the bobbin's central axis.

13. The inspection system according to any one of claims 11 or 12, characterized in that: The bobbin holder (11) consists of a ring (17) having a fixing area in which the yarn bobbin (12) can be placed to prevent relative movement of the yarn bobbin (12) within the bobbin holder (11) in at least one of a circumferential direction and / or a translational direction.

14. The inspection system according to claim 13, characterized in that: The fixing area is defined by two parallel extending rods (18) extending in the direction of the ring center axis of the ring (17) with a gap (20) between the rods, the gap (20) retaining the yarn bobbin (12) at its outer peripheral surface.

15. The inspection system according to any one of claims 11 to 14, characterized in that: The ring (17) has a swivel hinge (19) which is mounted to the guide rail (8) via a movable element.

16. The inspection system according to any one of claims 11 to 15, characterized in that: A plurality of bobbin holders (17) are arranged one after another along the guide rail (8).

17. The inspection system according to any one of claims 11 to 16, characterized in that: The bobbin holder (17) is adapted to move together along the guide rail (8).

18. The inspection system according to any one of claims 1 to 17, characterized in that: The system further comprises a darkroom (6) having two openings (9, 10) through which the guide rail (8) extends from an inlet to an outlet.

19. The inspection system according to claim 18, characterized in that The image acquisition device (5) is arranged in the dark room (6).

20. The inspection system according to any one of claims 1 to 19, characterized in that: The image acquisition device (5) comprises an illumination unit (15) for illuminating a region of the yarn bobbin (12) for defect detection and a camera device (14) for acquiring an image of the region.

21. The inspection system according to any one of claims 1 to 20, characterized in that: The system also includes an image processing and adjustment module (16).

22. Inspection system according to claims 1 to 21, characterized in that The system is configured to use different image acquisition devices (5) to determine defects, depending on the type of defect to be detected.

23. The inspection system according to claim 20, characterized in that The illumination section (15) is operated at a specific wavelength and / or color temperature depending on the type of defect to be detected.

24. The inspection system according to claim 21, characterized in that The image processing and conditioning module (16) is adapted to process and condition the image data using different filters depending on the type of defect to be detected.

25. Inspection system according to claims 18 to 24, characterized in that At least one of the following types of image acquisition devices is arranged in the darkroom (6): an image acquisition device for inelastic defects, an image acquisition device for damaged tube defects, an image acquisition device for wear defects, an image acquisition device for color defects, and an image acquisition device for diameter defects.

26. The inspection system according to claim 25, characterized in that The image acquisition device for inelastic defects comprises an illumination portion (15) in the form of a ring and a camera device (14) arranged in the central area of ​​the ring.

27. The inspection system according to claim 25 or 26, characterized in that: The two image acquisition devices for inelastic defects are arranged to sandwich the guide rail (8).

28. An inspection system according to any one of claims 25 to 27, characterized in that: A further image acquisition device is arranged upstream or downstream of the image acquisition device for the inelastic defect.

29. The inspection system according to any one of claims 1 to 28, characterized in that: The system is configured to automatically sort yarn bobbins (12) fed to the system according to the type of defects detected.

30. The inspection system according to any one of claims 1 to 29, characterized in that: The system is configured to periodically store data acquired during defect detection in the database after confirmation by a human that the detected defect corresponds to a defect type assigned to the defect from the system.

31. An inspection system according to any one of claims 1 to 29, characterized in that: The system is configured to perform defect assignment to different defect types by artificial intelligence and / or the system is configured such that the system operates using deep learning.

32. A method for inspecting yarn bobbins and assigning different defect types to said yarn bobbins, wherein: The method uses an application and a database for defect allocation.