Inspection of steel elements

Abstract

A method for classifying a steel coil (10) is disclosed, the method comprising the step of aligning the steel coil (10) with an alignment device (20) for approaching at least one specific surface (11) of the steel coil (10), where the alignment comprises at least one of rotation, lifting, lowering and pivoting. The method further comprises: inspecting the at least one surface (11), the inspecting comprising: capturing at least one image (i) of the at least one surface (11); comparing the captured at least one image (i) with a base image (b) stored in a database, the base image representing a specific quality (q0) of the at least one surface (11); determining a quality (q1) of the at least one surface (11) depending on a comparison of the at least one captured image (i) with the base image (b); and classifying the steel coil (10) on the basis of the determined mass (q1).

Description

Technical Field

[0001] This disclosure relates to a method for classifying steel coils and a system for performing the method to classify steel coils. Background Technology

[0002] Core steel is used to manufacture the core of a transformer and is therefore a crucial part of its function. Currently, core steel is primarily delivered in coils, placed on top of wooden pallets and wrapped in paper and plastic. Core steel is subject to numerous quality issues related to internal and external transportation, which can lead to damage to the coils on the bottom slotted side of the pallet. The coils are also exposed to harsh environmental conditions from external weather. Rain, snowmelt, and high humidity can cause the steel to rust, especially if the wrapping around the coil is inadequate or damaged.

[0003] EP1128179A2 describes a method for identifying workpieces punched from sheet metal to determine those that are substantially free of defects. The method includes the steps of: inserting workpieces into a channel of predetermined width to filter those that have passed through the channel; analyzing corresponding images of the workpieces that have passed through the channel to compare these images with a reference workpiece image; and rejecting workpieces that have a portion different from the reference workpiece image.

[0004] CN109995195A describes a post-processing production line for motor stator winding. The motor stator inspection device has a simple structure and can automatically detect whether the enameled wire of the stator is damaged and provide a prompt when the enameled wire of the stator is damaged to reject unqualified products.

[0005] JP2023033132A describes an edge anomaly detection device for a coiled metal strip, the device comprising: an image acquisition unit that acquires an image obtained by imaging an edge portion of a metal strip wound in a coil shape in the axial direction; and an anomaly determination unit that determines the presence / absence of an anomaly in the edge portion based on a smoothed image generated by a smoothing processing unit.

[0006] XP093129959 describes transverse color imaging (LCI) as a measuring device used in the processing of sheet metal and metal strip.

[0007] CN113240628A discloses a method for judging the quality of steel coils. The method includes the following steps: acquiring end images of N steel coils, wherein end defects and defect locations of the steel coils are marked in the end images of the N steel coils; processing defect features to obtain corresponding training samples; and inputting the training samples into a classification model.

[0008] CN218481431U discloses a defect detection device for hot-rolled steel plates used in battery casings. The device includes a detection mechanism and a first detection platform assembly. The detection mechanism includes a laser rangefinder for scanning the shape data of the steel coil's end face; the first detection platform assembly includes a rotary table for adjusting the relative position of the steel coil and the laser rangefinder. Summary of the Invention

[0009] The objective is to provide a method for classifying steel coils to identify low-quality or damaged coils early in the steel coil processing process, while still reducing the risk of personal safety hazards. Another objective is to provide a system for approaching and inspecting steel coils that is more efficient and safer for humans.

[0010] This objective is achieved through the features of the independent claim. Advantageous embodiments are indicated in the dependent claims.

[0011] For example, as claimed in the independent claims, the embodiments of this disclosure address, in whole or in part, the above-mentioned disadvantages in the art. Further embodiments of the methods and systems for sorting steel coils are the subject of the additional claims.

[0012] A method for classifying steel coils is provided, the method comprising the steps of: aligning the steel coil with an alignment device for accessing at least one specific surface of the steel coil, wherein aligning the steel coil includes at least one of: rotation, lifting, lowering, and pivoting. The method further comprises: inspecting the at least one surface, wherein the inspection includes: capturing at least one image of the at least one surface; comparing the captured at least one image with the base image stored in a database, the base image representing a specific quality of the at least one surface; determining the quality of the at least one surface based on the comparison of the captured at least one image with the base image; and classifying the steel coil based on the determined quality.

[0013] Alignment devices are, for example, cranes, coil holders, hand-held trolleys, forklift attachments, coil turners, dedicated coil lifting equipment, or other solutions that lift and lower the coil and / or pivot and / or rotate the coil by approximately 90°, 180°, or any other angle to approach a specific surface of the coil. At least one surface is inspected. For example, the bottom or top surface of a core coil. The captured image is compared to a base image representing a specific quality. The base image is, for example, stored in a database. The base image is, for example, an image of a steel surface that has certain characteristic features defining a specific quality of the coil. The specific quality of the coil is, for example, “acceptable,” “defective,” or “critical / uncertain.” Coils classified as having “acceptable” quality are ready for further processing. Coils classified as having “defective” quality are identified early as low-quality or damaged and returned to the manufacturer for compensation, saving costs. Coils classified as having “critical / uncertain” quality can undergo further inspection. This inspection can be automatic or manual.

[0014] Potential damage leaves traces on the surface of steel components, such as on the laminates of steel coils. These traces can be identified and categorized by scanning the surface of the steel coil. Therefore, images showing the damage (primarily broken parallel lines) are captured. During the visual inspection of the steel coil, the following anomalies were observed: rust, harmful substances, angel hairs, white edges, and bright surfaces can be identified by listing some possible types of damage that suggest traces. Image capture can be done automatically or manually using different devices. It is also possible to take multiple images of different points on one or more surfaces to obtain an appropriate impression of the steel coil's quality.

[0015] Automatically comparing captured images with images from a database improves efficiency. Manual work that must have been done by touching surfaces or manually operating the camera is reduced. Automating the inspection of steel coils involves less subjectivity and is therefore less prone to error. Furthermore, avoiding physical contact with the steel helps prevent potential hazards, as edges can be very sharp.

[0016] According to an embodiment, the capture of the at least one image is provided by a camera.

[0017] Different cameras or devices with sensitive optics in different wavelength ranges can be used for visual inspection. For example, sensors or cameras that operate within the visible or invisible wavelength spectrum. Furthermore, microscopic cameras can be used to detect defects, such as burrs on a surface, more precisely.

[0018] The camera can be connected to a screen, making it easier for the operator to obtain an overall view, for example, of the lower slit side of the steel coil. The camera can be positioned close to the alignment device holding the steel coil so that it is displayed from the side. A better way to obtain a shot of the slit side of the steel coil is to position the camera directly below it. Another aspect is the illumination of the steel coil. Light shining directly onto the material from the camera is not the optimal solution, as the metal will reflect the light back into the camera. Illumination provided by an external source located on the side of the machine is preferred.

[0019] According to an embodiment, the capture of the at least one image is provided via lateral color imaging.

[0020] Another method is to use transverse color imaging (LCI) to directly measure the burr height on the surface of the steel coil. This method requires more manual processing than simply placing a camera on the steel element (such as a coil or sheet). Internal defects in the steel coil can be detected using this method.

[0021] According to an embodiment, the comparison of the at least one image with the base image is provided through artificial intelligence and / or machine learning.

[0022] Combining artificial intelligence (AI) and / or machine learning with cameras can automate, or at least improve, inspection by taking identical images of the steel coil each time, and then using machine learning and AI tools to classify the coil as "defective," "acceptable," or "critical / uncertain." To create a model to be used for this classification, a large number of already classified images are needed. Then, if sufficient data is provided during the creation / training phase, using the model to classify the images and thus the quality of the steel coil is simple, very fast, and highly accurate. These camera inspections can potentially be performed at different locations.

[0023] Potential damage left traces on the laminate. By capturing images of the steel coil surface, images showing the damage (mainly broken parallel lines) were received. These images were fed into an AI system that provided an indication of whether the surface was damaged.

[0024] According to another embodiment, the method includes the steps of unpacking the steel coil and subsequently inspecting at least one surface.

[0025] The critical time to identify low-quality or damaged steel coils is immediately after unpacking. This occurs during external and internal transport, such as delivery from external companies or producers, in preparation for further processing. Therefore, it is important to identify at least one surface of the steel element placed on a wooden pallet (e.g., the bottom side of the core coil). During transport, this bottom side eventually slides and becomes damaged on the wooden pallet.

[0026] The first potential location is directly at the unpacking point. Here, defects will be detected early, saving processing time for the cutting machine and allowing the manufacturer to retrieve the coil. Since the coil is handled "upside down," the biggest challenge in completing inspection at the unpacking point is positioning the camera below the coil to obtain a sufficiently good image. For this purpose, alignment devices will be used, for example, to lift the coil or rotate it approximately 90° or 180°.

[0027] According to another embodiment, the method includes the steps of cutting the steel coil and subsequently inspecting the at least one surface.

[0028] Inspecting the steel coils after they have been cut by a cutting machine for further processing is another alternative. Performing the inspection at this point allows for the detection of errors caused by internal processes. There is no problem as long as the cut steel components (e.g., sheets) are stacked evenly and close to the edges to capture an image of the surface for comparison with a base image.

[0029] After the steel component is cut into sheets or small slices, the cut steel component is placed, for example, at a machine to measure the burr height using LCI (Liquid Crystal Interference). This is a more precise method that can detect internal defects. Alternatively or additionally, visual inspection is performed using a camera.

[0030] According to an embodiment, the inspection of the steel coil occurs before the steel coil is cut.

[0031] Alternatively or additionally, inspections can be performed before cutting the coil. The two methods can also be combined to obtain the best results from both by inspecting the coil before and after cutting. These can be combined to allow for early detection of defect image classification, and LCI can be used as a second opinion in “critical” cases or if there is concern that the coil may have been damaged during internal processes.

[0032] According to an embodiment, a system adapted to perform steps of a method for sorting steel coils includes an alignment device having a lifting device and a rotating device. The system further includes an inspection device having a camera, a data storage device, and a data processing device.

[0033] Alignment devices are, for example, cranes, coil holders, hand-held trolleys, forklift attachments, coil turners, dedicated coil lifting equipment, or other solutions that lift and lower the coil and / or pivot and / or rotate the coil by approximately 90° or 180° or any other angle to approach a specific surface of the coil.

[0034] The inspection apparatus includes a camera, a data storage device, and a data processing device. Different cameras or devices with sensitive optics in different wavelength ranges can be used for visual inspection. For example, sensors or cameras operating within the visible or invisible wavelength spectrum. Furthermore, microscopic cameras can be used to detect defects, such as burrs on a surface, more precisely.

[0035] According to an embodiment, the computer program includes instructions for causing the system to perform the method steps.

[0036] In addition, a computer-readable medium is provided on which a computer program is stored.

[0037] The method for classifying steel coils described above and the computer program for performing the steps of the method are particularly suitable for systems that perform the steps of the method for classifying steel coils.

[0038] Therefore, the features and advantages described in the system and computer program can be used in this method, and vice versa.

[0039] This disclosure includes several aspects of a method for classifying steel coils. Each feature described with respect to one of these aspects is also disclosed herein with respect to another aspect, even if the corresponding feature is not explicitly mentioned in the context of a particular aspect. Attached Figure Description

[0040] The accompanying drawings are included to provide further understanding. In the drawings, elements with the same structure and / or function may be referenced by the same reference numerals. It will be understood that the embodiments shown in the drawings are illustrative representations and are not necessarily drawn to scale.

[0041] Figures 1 to 3 This is a schematic diagram of a system used to classify steel components.

[0042] Figure 4 These are the steps in a method for classifying steel components. Detailed Implementation

[0043] In the following description, the method and apparatus are described primarily in interaction with a steel element, which is specifically designed as a steel coil.

[0044] Figure 1 A system 100 for sorting steel components 10 (e.g., steel coils) is shown. The steel component 10 includes a first surface 11 and a second surface 12. Figure 1In this configuration, the first surface 11 is the bottom side of the steel coil 10, and the second surface 12 is the side surface of the steel coil 10. The steel element 10 is not limited to two surfaces but may have multiple surfaces, including, for example, a top surface. Alternatively or additionally, the steel element 10 is a steel sheet with surfaces and edges. The system 100 includes an alignment device 20 and an inspection device 30. The alignment device 20 includes a housing 21, a lifting device 22, and a rotating device 23. The lifting device 22 and the rotating device 23 may be separate devices or integrated into a single device (e.g., a crane). The alignment device 20 may also be, for example, a crane, a coil holder, a hand-held trolley, an attachment for a forklift, a coil turner, a dedicated coil lifting device, or other solutions that lift and lower the steel element 10 and / or pivot and / or rotate the element 10 by approximately 90° or 180° or any other angle to approach a specific surface of the steel element 10.

[0045] The inspection device 30 includes a camera 31, a data storage device 32, and a data processing device 33. The camera 31 can be any sensor or camera operating in the visible or invisible wavelength spectrum. Furthermore, a microscopic camera can be used to more precisely detect defects, such as burrs on a surface. Figure 1 In this configuration, camera 31 is positioned close to alignment device 30 holding steel element 10 so that steel element 10 is displayed from the side. The captured image is displayed on monitor 34, which makes it easier for the operator to obtain an overall view of the surface of steel element 10.

[0046] Alternatively, the captured image is not displayed at all, but is processed using only the underlying image in a data storage device or computer-based analysis tool. In such cases, an operator is unnecessary, and the system operates entirely automatically.

[0047] Figure 2 Another exemplary embodiment of a system 100 for classifying steel components 10 is shown. System 100 and... Figure 1 The system is the same as in the previous one, except that the camera 31 is positioned directly below the steel element 10 to obtain a better view. In this case, the camera 31 is placed in a recess in the base plate.

[0048] Figure 3 Another exemplary embodiment of a system 100 for classifying steel components 10 is shown. This embodiment is related to... Figure 2Similar. The difference lies in that, in order to place the camera 31 below the steel element 10, a metal fixing device 40 with a flexible arm 42 is used to guide and hold the camera 31 at a specific point on the first surface 11 (e.g., the lower slit side of the core steel coil). The fixing device 40 can be adjusted vertically to compensate for different heights and widths of the steel element 10, and the flexible arm 42 makes it easy to place the camera 31, attached to the end of the fixing device 40, at a specific point. The operator can position the fixing device 40 so that the camera 31 is on the bottom slit side without any part of their body being directly below the suspended steel element 10 to minimize safety risks. After the operator has positioned the camera 31 in contact with the first surface 11, the fixing device 40 will hold the camera 31 against the steel element 10 using a vertical force (equivalent to a rubber band) greater than the relevant gravity. When the fixing device 40 is pulled down, the rubber band is stretched and wants to return to its natural position, thus generating an upward force proportional to the downward vertical distance. The potential vertical force can be adjusted to suit specific conditions, such as roll size. Camera 31 is connected to monitor 34 and software is installed thereon, which the operator uses to perform visual inspections and, if possible, to take magnified images of the surface for further investigation.

[0049] Figure 4 The steps of a method for classifying steel element 10 are shown. In a first step S1, steel element 10 is aligned with alignment device 20. Alignment includes raising and lowering steel element 10 and / or pivoting and / or rotating steel element 10 by approximately 90° or 180° or any other angle to approach a specific surface of steel element 10. In a second step S2, at least one surface 11 of steel element 10 is inspected. The inspection includes: capturing at least one image i of the at least one surface 11; comparing the captured at least one image i with a base image b representing a specific quality q1 of the at least one surface 11; determining the quality q1 of the at least one surface 11 based on the comparison between the captured at least one image i and the base image b; and classifying steel element 10 based on the determined quality q1. It is also possible to repeat the inspection of steel element 10 several times to obtain a definite quality of steel element 10. In a third step S3, steel element 10 is classified based on the determined quality q1. It may be necessary to unpack steel element 10 in a separate step S0 before inspecting steel element 10. Alternatively, the steel component 10 can be cut in a separate step Sc. The cutting can be done before and / or after inspecting the steel component 10.

[0050] While this disclosure is open to various modifications and alternatives, its details have been shown by way of example in the accompanying drawings and will be described in detail. However, it should be understood that this disclosure is not intended to be limited to the specific embodiments described. Rather, it is intended to cover all modifications, equivalents, and alternatives that fall within the scope of this disclosure as defined by the appended claims.

[0051] As stated Figures 1 to 4 The embodiments shown illustrate exemplary models of methods for classifying steel components and systems suitable for performing the steps of those methods; therefore, they do not constitute a complete list of all embodiments of methods and systems for classifying steel components. For example, actual arrangements and methods may differ from the illustrated embodiments in terms of arrangement, apparatus, methods, and signals.

[0052] Figure Labels

[0053] 10 steel coils 11 First Surface 12 Second Surface 20 Alignment Device 21 shell 22 Lifting Device 23 Rotating device 30 Inspection Device 31 cameras 32 data storage devices 33 Data Processing Device 34 monitors 40 Fixtures 42 flexible arms iImage 100 systems for classification

Claims

1. A method for classifying steel coils (10), the method comprising the following steps: - Aligning the steel coil (10) with the alignment device (20) for accessing at least one specific surface (11) of the steel coil (10), wherein the alignment includes at least one of the following: - Rotate, - promote, - Reduce, and - Pivot, - Inspecting at least one surface (11) includes: - Capture at least one image (i) of the at least one surface (11). - Compare at least one captured image (i) with a base image (b) stored in a database, the base image representing a specific quality (q_0) of the at least one surface (11). - Determine the quality (q1) of the at least one surface (11) by comparing it with the base image (b), and - The steel coils (10) are classified based on the determined quality (q1).

2. The method according to claim 1, wherein, The capture of the at least one image (i) is provided by a camera (31).

3. The method according to claim 1 or 2, wherein, The capture of the at least one image (i) is provided by transverse color imaging (LCI).

4. The method according to claim 1, 2 or 3, wherein, The comparison of the at least one image (i) with the base image (b) is provided by artificial intelligence and / or machine learning.

5. The method according to any one of the preceding claims, comprising the following steps: The steel coil (10) is unpacked and then the at least one surface (11) is inspected.

6. The method according to any one of the preceding claims, comprising the following steps: The steel coil (10) is cut, and then the at least one surface (11) is inspected.

7. The method according to any one of the preceding claims, wherein, The inspection of the steel coil (10) takes place before the steel coil (10) is cut.

8. A system (100) suitable for performing the steps of the method for classifying steel coils (10) according to claim 1, the system comprising: - Alignment device (20), which has: - Lifting device (22). - Rotating device (23), and - Inspection device (30), which has: - Camera (31) - Data storage device (32). - Data processing device (33).

9. A computer program comprising instructions for causing the system (100) of claim 8 to perform the steps of the method of claim 1.

10. A computer-readable medium having a computer program according to claim 9 stored thereon.

Images