A method, system and device for detecting pinch roll wrap angle

Abstract

The application provides a kind of pinch roll winding angle detection method and detection system, belong to steel processing and manufacturing field.The pinch roll winding angle detection method provided in the application is based on the pre-uncoiling image containing steel coil and uncoiler, respectively determine the first reference point in the pre-uncoiling image located at the center of the end face of the steel coil and the second reference point perpendicular to the first reference point along the vertical direction and located on the pinch roll;Based on the pre-uncoiling image, the steel coil contour feature data is extracted, and the strip head endpoint is determined according to the steel coil contour feature data;The first reference point, the second reference point and the strip head endpoint are fitted, and the pinch roll winding angle is output.The pinch roll winding angle detection method provided in the application can reduce the accident risk during the detection of the pinch roll winding angle, and improve the detection efficiency of the pinch roll winding angle.

Description

Technical Field

[0001] This application relates to the field of steel processing and manufacturing, and in particular to a method and system for detecting the winding angle of a pinch roll. Background Technology

[0002] Uncoiling refers to the process of unfolding a coiled steel strip into a flat surface. An uncoiler is used to unwind and straighten the coil, making it a flat strip that facilitates the next step in steel processing and production.

[0003] During the uncoiling process of steel coils, the pinch rolls of the uncoiler pull the strip head (strip end) into the uncoiler's inlet guide plate. Before pulling it out, it is usually necessary to set a suitable pinch roll winding angle to form the correct tension on the strip, ensure a good coil shape, and avoid various stress deformations in the strip during the pulling process. An incorrect pinch roll winding angle can easily cause the strip to deform and become unusable during winding, leading to production losses and safety accidents. Due to the irregular shape of the strip head and the complex working conditions at the uncoiling site, the pinch roll winding angle is usually measured manually in the current technology. Manual measurement has a high risk of accidents and is inefficient.

[0004] In view of the above, this invention is proposed. Summary of the Invention

[0005] This application provides a method and system for detecting the winding angle of a pinch roll, in order to solve the technical problems of accident risk and low efficiency caused by manually measuring the winding angle of a pinch roll in the prior art.

[0006] The first aspect of this application provides a method for detecting the winding angle of a pinch roll, comprising the following steps:

[0007] Based on the pre-uncoiling image containing the steel coil and the uncoiler, a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point perpendicular to the first reference point in the vertical direction and located on the pinch roll are determined respectively.

[0008] Steel coil contour feature data is extracted from pre-uncoiled images, and the lead end point is determined based on the steel coil contour feature data.

[0009] By fitting the first reference point, the second reference point, and the lead end point together, the winding angle of the pinch roll is output.

[0010] In some implementations, extracting the steel coil contour feature data based on the pre-uncoiled image, and determining the lead end point based on the steel coil contour feature data includes:

[0011] Based on the pre-uncoiled image, the outline feature data of the steel coil is extracted, and the longitudinal section outline feature data of the head facing the steel coil axis is obtained based on the outline feature data of the steel coil.

[0012] The longitudinal section profile feature data with the head is masked to obtain the longitudinal section mask data;

[0013] The endpoints facing the pinch rolls in the extracted longitudinal section mask data are identified as the lead end points.

[0014] In some implementations, extracting the steel coil contour feature data based on the pre-uncoiled image, and determining the lead end point based on the steel coil contour feature data includes:

[0015] Based on the pre-uncoiled image, the profile feature data of the steel coil is extracted. Based on the profile feature data of the steel coil, the profile feature data of the longitudinal section of the strip facing the steel coil axis and the profile feature data of the transverse end face of the strip facing the pinch roll are obtained.

[0016] The longitudinal section profile feature data and the transverse end face profile feature data with the head are respectively masked to obtain longitudinal section mask data and transverse end face mask data.

[0017] The intersection point closest to the second reference point in the intersection region of the extracted longitudinal section mask data and transverse end face mask data is determined as the leading endpoint.

[0018] In some implementations, the first reference point, the second reference point, and the lead end point are fitted together to output the pinch roll winding angle, including:

[0019] Determine the reference line for the leading angle based on the leading endpoint and the first reference point;

[0020] The reference line for the angle of the pinch roller is determined based on the first reference point and the second reference point;

[0021] The winding angle of the pinch roll is calculated and output based on the reference line of the lead angle and the reference line of the pinch roll angle.

[0022] In some implementations, calculating the output pinch roll take-up angle based on the lead angle reference line and the pinch roll angle reference line includes:

[0023] The horizontal angle of the band is obtained based on the slope of the band angle reference line;

[0024] The horizontal angle of the pinch roll is obtained based on the slope of the pinch roll angle reference line;

[0025] The difference between the horizontal angle of the lead roller and the horizontal angle of the pinch roller is calculated to output the winding angle of the pinch roller.

[0026] In some implementations, calculating the output pinch roll take-up angle based on the lead angle reference line and the pinch roll angle reference line includes:

[0027] The winding angle of the pinch roll is output based on the angle between the lead angle reference line and the pinch roll angle reference line.

[0028] In some implementations, the first reference point, the second reference point, and the lead end point are fitted together to output the pinch roll winding angle, including:

[0029] Obtain the coordinate parameters of the first reference point, the second reference point, and the endpoint in the preset coordinate system;

[0030] The pinch roll winding angle is output based on the obtained coordinate parameters through fitting calculation.

[0031] In some embodiments, based on a pre-uncoiling image including the steel coil and the uncoiler, determining a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point perpendicular to the first reference point and located on the pinch roll includes:

[0032] Extract the parameters of the first reference point and the second reference point, and determine the first reference point and the second reference point in the pre-unrolled image;

[0033] The parameters of the first reference point and the parameters of the second reference point are stored in a priori database.

[0034] In some embodiments, the method for detecting the winding angle of the pinch rollers further includes:

[0035] The pinch roll take-up angle is compared with a preset threshold. Based on whether the pinch roll take-up angle exceeds the preset threshold, the accuracy of the pinch roll take-up angle is determined; and / or

[0036] The winding angles of multiple continuously output pinch rollers are curve-fitted to obtain a fitting curve. The accuracy of the winding angle of the pinch rollers is determined based on the growth trend of the fitting curve.

[0037] In some embodiments, the method for detecting the winding angle of the pinch roll further includes: acquiring a pre-unwinding image based on an image acquisition device oblique to the axis of the pinch roll.

[0038] A second aspect of this application provides a detection system for the winding angle of a pinch roll, used in applying the above-described method for detecting the winding angle of a pinch roll, including:

[0039] The image processing module is used to extract the outline feature data of the steel coil based on the pre-uncoil image containing the steel coil and the uncoiler, and to determine the first reference point, the second reference point, and the head end point;

[0040] The data processing module is used to fit the first reference point, the second reference point and the lead end point together, and output the winding angle of the pinch roll.

[0041] A third aspect of this application provides a detection device for the take-up angle of a pinch roll, including a host computer and an image display device; the image display device is used to visualize at least one of the following in a pre-unwinding image: steel coil contour feature data, a strip head angle reference line, a pinch roll angle reference line, and the pinch roll take-up angle; the host computer is configured to:

[0042] Based on the pre-uncoiling image containing the steel coil and the uncoiler, a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point perpendicular to the first reference point in the vertical direction and located on the pinch roll are determined respectively.

[0043] Steel coil contour feature data is extracted from pre-uncoiled images, and the lead end point is determined based on the steel coil contour feature data.

[0044] By fitting the first reference point, the second reference point, and the lead end point together, the winding angle of the pinch roll is output.

[0045] A fourth aspect of this application provides a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the detection method described above.

[0046] Compared with the prior art, the method and system for detecting the winding angle of the pinch roll provided in this application have at least the following advantages:

[0047] Beneficial effects:

[0048] The method for detecting the pinch roll winding angle provided in this application is based on a pre-uncoiling image containing a steel coil and an uncoiler. It determines a first reference point located at the center of the steel coil end face in the pre-uncoiling image and a second reference point perpendicular to the first reference point and located on the pinch roll. Determining the positions of the first and second reference points in the pre-uncoiling image provides a data basis for subsequently outputting the pinch roll winding angle. Based on the pre-uncoiling image, it extracts the steel coil contour feature data and determines the strip end point. By extracting the steel coil contour feature data from the pre-uncoiling image, the shape of the strip end on the steel coil can be accurately determined, thus identifying the position of the strip end point in the pre-uncoiling image. By fitting the first reference point, the second reference point, and the strip end point together, the pinch roll winding angle can be output using the position data of the first reference point, the second reference point, and the strip end point in the pre-uncoiling image. Therefore, by using automated image processing technology, the winding angle of the pinch rolls can be detected quickly and accurately, thus avoiding the time-consuming and labor-intensive traditional manual measurement methods, thereby improving the unwinding efficiency of steel coils and effectively reducing the risk of accidents.

[0049] Other advantages and features of the method and system for detecting the take-up angle of the pinch roller provided in this application will be further described in subsequent specific embodiments. Attached Figure Description

[0050] To more clearly illustrate the specific embodiments of this application or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A logic flowchart of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0052] Figure 2 A logic flowchart of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0053] Figure 3 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0054] Figure 4 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0055] Figure 5 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0056] Figure 6 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0057] Figure 7 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0058] Figure 8 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0059] Figure 9 A flowchart illustrating the steps of a method for detecting the take-up angle of a pinch roller provided in an embodiment of this application;

[0060] Figure 10 A structural block diagram of a pinch roll take-up angle detection system provided for embodiments of this application;

[0061] Figure 11 A structural block diagram of a device for detecting the take-up angle of the pinch roll provided in an embodiment of this application;

[0062] Figure 12 A schematic diagram illustrating the detection of the pinch roller winding angle provided in an embodiment of this application;

[0063] Figure 13 A schematic diagram of longitudinal section mask data provided for embodiments of this application;

[0064] Figure 14 A schematic diagram of longitudinal section mask data and transverse end face mask data provided for embodiments of this application;

[0065] Figure 15 A schematic diagram of the field during the detection of the pinch roller winding angle provided for an embodiment of this application;

[0066] Figure 16 A schematic diagram of the field during the detection of the pinch roller winding angle provided for an embodiment of this application;

[0067] Figure 17 This is a schematic diagram of the field during the detection of the winding angle of the pinch rollers, provided for an embodiment of this application.

[0068] Explanation of reference numerals in the attached figures:

[0069] 1. Detection system; 11. Image acquisition module; 12. Image processing module; 13. Data processing module;

[0070] 14. Data storage module; 15. Image display module;

[0071] 1000. Detection equipment; 1001. Host computer; 1002. Image display device; 1003. Image acquisition device;

[0072] a. Uncoiler; b. Steel coil; c. Pinch roll; d. Leading end; e. Second reference point; f. First reference point; g. Leading end point; h. Longitudinal section mask data; j. Transverse end face mask data. Detailed Implementation

[0073] To make the above and other features and advantages of this application clearer, the application is further described below with reference to the accompanying drawings. It should be understood that the specific embodiments given herein are for the purpose of explanation to those skilled in the art, and are exemplary only, not restrictive.

[0074] In the following description, numerous specific details are set forth to provide a thorough understanding of this application. In other instances, well-known steps or operations have not been described in detail to avoid obscuring this application.

[0075] The method for detecting the winding angle of the pinch roll provided in this application embodiment can be executed by the detection system 1 for detecting the winding angle of the pinch roll provided in this application embodiment. The detection system 1 can be configured in the detection device 1000 for detecting the winding angle of the pinch roll.

[0076] As for the foregoing, the inventive concept of at least one embodiment of this application is to provide a method for detecting the winding angle of a pinch roll. By using a deep learning algorithm and semantic segmentation technology, the outline feature data of the steel coil is accurately extracted from the pre-uncoiling image, and the lead end point in the pre-uncoiling image is determined. Combined with a first reference point located at the center of the end face of the steel coil and a second reference point perpendicular to the first reference point and located on the pinch roll, the winding angle of the pinch roll is obtained by fitting the lead end point, the first reference point, and the second reference point. This achieves fast and accurate automated detection of the winding angle of the pinch roll, thereby improving the uncoiling efficiency of the steel coil and effectively reducing the risk of accidents.

[0077] Based on the above overall concept, and referring to Figures 1 to 17 One embodiment of this application provides a method for detecting the winding angle of a pinch roller, comprising the following steps:

[0078] S100. Based on the pre-uncoiling image containing the steel coil and the uncoiler, determine a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point that is perpendicular to the first reference point in the vertical direction and located on the pinch roll.

[0079] S200. Extract the steel coil contour feature data based on the pre-uncoiled image, and determine the head end point based on the steel coil contour feature data;

[0080] S300: Fit the first reference point, the second reference point and the lead end point together to output the pinch roll winding angle.

[0081] It should be noted that, as Figure 12 As shown, the pre-uncoiling image in this embodiment includes a steel coil and an uncoiler. Since the steel coil is usually supported on the uncoiler and the lead is facing the pinch roll during pre-uncoiling, the pre-uncoiling image can have the steel coil lead image information and at least one end face image information, and at the same time, it can have the image information of the middle part located between the two ends of the pinch roll, so as to facilitate the determination of the lead end point, the first reference point and the second reference point in the pre-uncoiling image.

[0082] It needs to be clarified that the reference Figure 12 As shown, since the steel coil is horizontal rather than vertical in the pre-uncoiling state, the end face is visible as circular or near-circular. In this embodiment, the first reference point is located at the center of the end face of the steel coil, that is, the center of the end face of the steel coil; the second reference point is perpendicular to the first reference point in the vertical direction and is located on the pinch roll; the strip end point refers to the outermost point of the strip end facing the pinch roll; the pinch roll winding angle refers to the angle between the straight line connecting the first reference point and the second reference point and the straight line connecting the first reference point and the strip end point.

[0083] In step S100 of this embodiment, determining the first reference point and the second reference point based on the pre-uncoiling image helps to determine the relative position of the steel coil and the pinch roll in the pre-uncoiling image, providing a data basis for subsequent fitting and calculation of the pinch roll winding angle. In step S200, the steel coil contour feature data can be extracted from the acquired pre-uncoiling image, and the geometric features of the steel coil strip head can be obtained based on the steel coil contour features, thereby accurately determining the strip head end point. Through step S300, the first reference point, the second reference point, and the strip head end point are fitted together, thereby quickly and accurately obtaining the pinch roll winding angle, thus realizing a high degree of automation in pinch roll winding angle detection, greatly improving the steel coil uncoiling efficiency and reducing the risk of accidents.

[0084] It should be understood that by fitting the first reference point, the second reference point, and the end point of the strip together, the pinch roll winding angle is output. The pinch roll winding angle can be calculated based on an algorithm model using coordinate parameters. For example, the pinch roll winding angle can be calculated using an algorithm model based on the coordinate parameters of the first reference point, the second reference point, and the end point of the strip in a preset coordinate system. Alternatively, it can be calculated using an algorithm model based on image processing technology. For example, image processing technology can be used to obtain relevant features from the pre-unwinding image, that is, the algorithm model can be used to identify and extract the connecting straight lines between the first reference point, the second reference point, and the end point of the strip, and then the included angle between these straight lines can be calculated.

[0085] In some embodiments of this application, the pre-uncoiling image can be obtained using an image acquisition device 1003. Because the shape of the strip is irregular and its geometric features are complex, compared to using sensing devices to acquire the geometric features of the steel coil and pinch rolls, the image acquisition device 1003 can more accurately acquire these complex geometric features, ensuring that the calculated pinch roll winding angle is the optimal pinch angle value. Furthermore, the pre-uncoiling image can be a real-time on-site image acquired by the image acquisition device 1003, or it can be a processed image obtained from the real-time on-site image acquired by the image acquisition device 1003 after further processing such as image enhancement, image filtering, image segmentation, image compression, image registration, and image fusion.

[0086] In some embodiments, reference is made to Figure 4 and Figure 13 As shown, step S200 includes:

[0087] S211. Extract steel coil contour feature data based on pre-uncoiled image, and obtain longitudinal section contour feature data of the head facing the steel coil axis based on the steel coil contour feature data.

[0088] S212. Perform masking on the longitudinal section profile feature data with the head to obtain longitudinal section mask data;

[0089] S213. Extract the endpoints in the longitudinal section mask data that face the pinch roll direction and determine them as the lead end points.

[0090] It should be noted that this embodiment uses a preset algorithm model to extract the outline features of the pre-uncoiled image. The preset algorithm model can use ResNet, MobileNet series, ConvNeXt, OCRNet, GCNet, Unet, FCN or EfficientNetV3 model as the neural network backbone. In this embodiment, EfficientNetV3 model can be optionally used as the neural network backbone. The semantic information of each pixel in the pre-uncoiled image is extracted through deep learning, and each pixel is classified into the steel coil area or the pinch roll area through semantic segmentation technology to extract the outline feature data of the steel coil.

[0091] In step S211 of this embodiment, the steel coil contour feature data is extracted from the pre-uncoiled image using a preset algorithm model. Based on semantic segmentation, the longitudinal section contour feature data of the strip head, facing the axial direction of the steel coil and located on the end face of the steel coil, is accurately divided from the steel coil contour feature data to precisely locate and identify the geometric features of the irregular strip head facing the axial direction of the steel coil. Further, in step S212, the longitudinal section contour feature data of the strip head is masked to highlight the data in the longitudinal section area, removing irrelevant or interfering information and retaining the key longitudinal section contour feature data, thus achieving data cleaning and optimization, and improving data quality and reliability. In step S213, after obtaining the longitudinal section mask data, the leftmost or rightmost endpoint of the longitudinal section mask data can be extracted and determined as the strip head endpoint based on the direction of the strip head towards the pinch roll, thereby achieving precise positioning of the strip head endpoint. For example, when the strip head faces right, the rightmost endpoint of the longitudinal section mask data can be extracted and determined as the strip head endpoint; when the strip head faces left, the leftmost endpoint of the longitudinal section mask data can be extracted and determined as the strip head endpoint.

[0092] In another embodiment of this application, reference is made to... Figure 5 and Figure 14 As shown, step S200 includes:

[0093] S221. Extract the profile feature data of the steel coil based on the pre-uncoiled image, and obtain the profile feature data of the longitudinal section of the strip facing the steel coil axis and the profile feature data of the transverse end face of the strip facing the pinch roll based on the profile feature data of the steel coil.

[0094] S222. Perform masking processing on the longitudinal section profile feature data and the transverse end face profile feature data with the head respectively to obtain longitudinal section mask data and transverse end face mask data.

[0095] S223. The intersection point closest to the second reference point in the intersection region of the extracted longitudinal section mask data and transverse end face mask data is determined as the leading endpoint.

[0096] It should be noted that, in the embodiments of this application, the longitudinal profile feature of the strip head refers to the profile feature of the side of the strip head, which is located on the end face of the steel coil; the profile feature of the transverse end face of the strip head refers to the profile feature of the side of the strip head facing the pinch roll.

[0097] In step S221 of this embodiment, the steel coil outline features are extracted from the pre-uncoiled image using a preset algorithm model. Based on semantic segmentation, the longitudinal section outline feature data and the transverse end face outline feature data of the strip head are separately divided from the steel coil outline features to accurately locate and identify the geometric features of the irregular strip head facing the axial side of the steel coil and the side facing the pinch roll. Further, in step S212, the longitudinal section outline features and the transverse end face outline features of the strip head are masked respectively, so that the longitudinal section region and the transverse end face outline features of the strip head are highlighted, thereby obtaining the longitudinal section mask data. Based on the cross-sectional mask data, masking is performed separately to remove irrelevant or interfering information, while retaining the key longitudinal section profile feature data and cross-sectional profile feature data of the strip, thus achieving data cleaning and optimization and improving data quality and reliability. Furthermore, highlighting the longitudinal section region and the cross-sectional profile feature of the strip helps to highlight the key intersection area, which is beneficial for the accurate calibration of the strip end point. In step S223, by extracting the intersection points of the longitudinal section mask data and the cross-sectional mask data, the strip end point can be accurately calibrated, thereby achieving precise positioning of the strip end point.

[0098] It is understood that the EfficientNetV3 model optionally used in this application embodiment, as a convolutional neural network model, incorporates various optimization strategies, including automatic model scaling, depth and width variability, etc., to provide better performance and efficiency. When using the EfficientNetV3 model for contour feature extraction, a pre-trained EfficientNetV3 model can be used as the backbone of the neural network, and then fine-tuned or further trained according to task requirements. By inputting pre-unwound image data, the EfficientNetV3 model can learn feature representations at different levels, including low-level edge and texture features and high-level semantic information. The longitudinal section mask data and transverse end face mask data obtained after processing by the EfficientNetV3 model can be a binary image, such as... Figure 15 As shown, Figure 15The longitudinal section mask data is marked in green, and the transverse end face mask data is marked in red. This mask data better represents the position and shape of the sides and ends of the steel coil, providing an accurate data foundation for subsequent processing steps. After extensive testing, the detection method designed in this application, using the EfficientNetV3 model to extract the steel coil contour feature data, achieves an accuracy rate of 99.26%, with extremely low error, which helps improve the accuracy of the pinch roll winding angle.

[0099] In another embodiment of this application, reference is made to... Figure 6 and Figure 12 Step S300 includes:

[0100] S310. Determine the leading angle reference line based on the leading endpoint and the first reference point;

[0101] S320. Determine the pinch roller angle reference line based on the first reference point and the second reference point;

[0102] S330: Calculate and output the winding angle of the pinch roll based on the reference line of the lead angle and the reference line of the pinch roll angle.

[0103] In this embodiment, step S310 determines the lead angle reference line based on the lead end point and the first reference point. The lead end point and the first reference point can be connected by a straight line to obtain the lead angle reference line, thereby determining the actual direction of the lead and providing a baseline for subsequent angle calculation. Step S320 determines the pinch roll angle reference line based on the first reference point and the second reference point. The first reference point and the second reference point can be connected by a straight line. Since the second reference point is located in the vertical direction of the first reference point, the pinch roll angle reference line is a vertical straight line segment, which is used as the reference normal line when fitting the pinch roll angle. Step S330 calculates and outputs the pinch roll winding angle based on the lead angle reference line and the pinch roll angle reference line. A straight line fitting algorithm can be used to directly fit the angle value between the two straight line segments from the lead angle reference line and the pinch roll angle reference line, which helps to improve the detection efficiency of the pinch roll winding angle.

[0104] In another embodiment of this application, reference is made to... Figure 7 Step S330 includes:

[0105] S3301. Obtain the horizontal angle of the head based on the slope of the head angle reference line;

[0106] S3302. Obtain the horizontal angle of the pinch roll based on the slope of the pinch roll angle reference line;

[0107] S3303: Calculate the difference between the horizontal angle of the belt head and the horizontal angle of the pinch roll to output the winding angle of the pinch roll.

[0108] In this embodiment, reference Figure 12 A coordinate system can be established by taking the first reference point at the center of the steel coil end face in the pre-uncoiling image as the origin, the straight line connecting the first and second reference points as the Y-axis, and the straight line extending horizontally from the first reference point as the X-axis. Further, combining deep learning algorithms, the slope of the aforementioned lead angle reference line is fitted to obtain the lead horizontal angle (the angle between the lead angle reference line and the X-axis), and the slope of the pinch roll angle reference line is fitted to obtain the pinch roll horizontal angle (the angle between the pinch roll angle reference line and the X-axis). By combining the lead horizontal angle and the pinch roll horizontal angle and calculating the difference, the angle value between the lead angle reference line and the pinch roll angle reference line can be obtained, i.e., the pinch roll winding angle, thereby further improving the detection accuracy of the pinch roll winding angle.

[0109] It should be noted that, after identifying the slope of the lead angle reference line and / or the pinch roll angle reference line using the deep learning algorithm in this embodiment, the horizontal angle of the lead and / or the horizontal angle of the pinch roll can be calculated using the arctangent function, providing an accurate data basis for the subsequent calculation of the pinch roll winding angle. For example, the formula for calculating the horizontal angle of the lead can be: θ1 = arctan(k1), where θ1 represents the horizontal angle of the lead and k1 represents the slope of the lead angle reference line; the formula for calculating the horizontal angle of the pinch roll can be: θ2 = arctan(k2), where θ2 represents the horizontal angle of the pinch roll and k2 represents the pinch roll angle reference line. Specifically, the formula for calculating the pinch roll winding angle can be θ = θ1 - θ2, where θ1 represents the horizontal angle of the lead, θ2 represents the horizontal angle of the pinch roll, and θ represents the pinch roll winding angle.

[0110] In some embodiments, reference Figure 12 The preset coordinate system can also be a coordinate system established with the first reference point at the center of the end face of the steel coil in the pre-uncoiled image as the origin, the straight line connecting the first reference point and the second reference point as the Y-axis, and the straight line extending horizontally from the first reference point as the X-axis. The X-axis is positive in the left direction, and the Y-axis is positive in the direction above. Alternatively, it can be combined with a deep learning algorithm to obtain the slope based on the aforementioned obtained lead angle reference line to obtain the vertical angle of the lead (the angle between the lead angle reference line and the Y-axis). Since the Y-axis is established based on the straight line connecting the first reference point and the second reference point (i.e., the pinch roll angle reference line), the angle value between the lead angle reference line and the pinch roll angle reference line can be directly obtained.

[0111] Therefore, the embodiments of this application use deep learning algorithms to obtain the slope of the head angle reference line and / or the pinch roll angle reference line based on image processing technology, which can realize the automatic calculation of the pinch roll winding angle to provide more accurate angle data, avoid errors introduced by human factors, and ensure the accuracy and reliability of the detected pinch roll winding angle.

[0112] In another embodiment of this application, step S330 includes:

[0113] S3311. Output the winding angle of the pinch roll based on the angle between the lead angle reference line and the pinch roll angle reference line.

[0114] Understandably, deep learning algorithms can also be used to directly identify the angle between the two line segments based on the lead angle reference line and the pinch roll angle reference line determined in the pre-unwinding image, thereby obtaining the pinch roll winding angle and further improving the detection accuracy of the pinch roll winding angle.

[0115] In another embodiment of this application, step S330 includes:

[0116] S331. Obtain the coordinate parameters of the first reference point, the second reference point, and the endpoint in the preset coordinate system respectively;

[0117] S332. Based on the obtained coordinate parameters, fit and calculate to output the winding angle of the pinch roll.

[0118] It should be understood that in step S31, the preset coordinate system can be a coordinate system established with the first reference point at the center of the end face of the steel coil in the pre-uncoiled image as the origin, the straight line connecting the first reference point and the second reference point as the Y-axis, and the straight line extending horizontally from the first reference point as the X-axis, wherein the X-axis is positive in the left direction and the Y-axis is positive in the direction above. By obtaining the coordinate parameters of the first reference point, the second reference point and the end point of the lead in the preset coordinate system respectively, and using a deep learning algorithm to fit and calculate, the pinch roll winding angle can also be output, so as to further improve the detection accuracy of the pinch roll winding angle.

[0119] In another embodiment of this application, step S100 includes:

[0120] S110. Extract the first reference point parameters and the second reference point parameters, and determine the first reference point and the second reference point in the pre-unrolled image; wherein the first reference point parameters and the second reference point parameters are stored in a priori database.

[0121] Understandably, since the uncoiler and the center of the coil end face are often in a fixed position in actual scenarios, the position parameters of the first reference point and the second reference point can be preset and stored in the database of the data storage module 14. Therefore, when detecting the winding angle of the pinch roll, it is not necessary to identify the first reference point and the second reference point in the pre-uncoil image in real time. During detection, it is only necessary to extract the first reference point parameters and the second reference point parameters from the database and match them with the center of the coil end face in the pre-uncoil image and the corresponding features on the pinch roll in the vertical direction along the center of the coil end face. The matching of the first reference point parameters and the second reference point parameters with the features of the pre-uncoil image can be achieved by image processing technology to make the detection method more efficient.

[0122] In some embodiments of this application, the calibration of the first reference point and the second reference point can also be performed in real time based on the pre-unwinding image, and determined in real time in the pre-unwinding image in combination with image processing technology.

[0123] Therefore, the method for detecting the pinch roll winding angle provided in this application can obtain the pinch roll winding angle in real time and accurately, making the winding angle control more convenient and precise. It is easy to set the appropriate winding angle in a timely manner to form the correct tension on the strip steel, ensure a good winding shape, guarantee the high yield of strip steel, improve the safety of steel coil uncoiling, and promote the efficiency of steel coil uncoiling.

[0124] In some embodiments of this application, reference is made to Figure 9 To improve the accuracy of the output pinch roll winding angle, the detection method for the pinch roll winding angle also includes:

[0125] S400. Compare the pinch roll winding angle with a preset threshold, and determine whether the pinch roll winding angle is accurate based on whether the pinch roll winding angle is greater than the preset threshold; and / or

[0126] S500: Perform curve fitting on multiple continuously output pinch roll winding angles to obtain a fitting curve, and determine whether the pinch roll winding angle is accurate based on the growth trend of the fitting curve.

[0127] In one embodiment of this application, step S400 compares the output pinch roller winding angle with a preset threshold. Based on whether the pinch roller winding angle exceeds the preset threshold, if the winding angle exceeds the preset threshold, it can be determined that the pinch roller winding angle is inaccurate. The preset threshold can be set according to actual needs. For example, if the actual pinch roller winding angle is usually less than 50°, the preset threshold can be set to 50°. If the output pinch roller winding angle is greater than 50°, it can be determined that the pinch roller winding angle is not accurate enough. At this time, a warning can be issued, or the pinch roller winding angle can be masked and the detection of the pinch roller winding angle can be repeated to further improve the detection accuracy of the pinch roller winding angle.

[0128] In one embodiment of this application, step S500 involves curve fitting based on the continuous output of multiple pinch roll winding angles to obtain a fitted curve. The accuracy of the pinch roll winding angle is determined based on the growth trend of the fitted curve. For example, if five pinch roll winding angles are continuously output, curve fitting is performed on the five pinch roll winding angles to obtain a fitted curve. If the fitted curve shows a continuous increase, it can be determined that the output pinch roll winding angle is accurate. If the fitted curve shows a discontinuous increase, it can be determined whether the pinch roll winding angle is accurate or not. At this time, an early warning can be issued or the multiple pinch roll winding angles can be masked and the pinch roll winding angle detection can be performed again to further improve the detection accuracy of the pinch roll winding angle.

[0129] In another embodiment of this application, the method for detecting the winding angle of the pinch roll further includes: acquiring a pre-unwinding image based on an image acquisition device 1003 that is inclined to the axis of the pinch roll.

[0130] It is understandable that the pre-uncoiling image is acquired by the image acquisition device 1003, which is angled towards the axis of the pinch roll. This can clearly obtain multi-directional information of the pinch roll and the steel coil (including side and top view information), thereby improving the accuracy and reliability of the measurement and ensuring the accuracy and comprehensiveness of subsequent inspections. Moreover, the image acquisition device 1003, which is arranged at the axis of the pinch roll, can reduce the obstruction caused by the equipment structure or the steel coil itself, ensuring the clarity and integrity of the acquired image, which is beneficial to subsequent processing and analysis.

[0131] The detection method also includes visualizing at least one of the following in the pre-uncoiling image: the coil outline features, the lead angle reference line, the pinch roll angle reference line, and the pinch roll winding angle. (See reference...) Figures 9 to 11 As shown, this visualization can intuitively display relevant parameters, helping operators to better understand and grasp the situation on the production site.

[0132] It should be noted that, Figure 9 The left side shows the actual pre-unrolled image captured, and the right side shows the visualized image after masking. Figure 10 and Figure 11 The red box in the image represents the area detected by the algorithm, the yellow line represents the reference line for the pinch roller angle, the red line represents the reference line for the lead angle, the yellow text represents the pixel area of ​​the longitudinal section contour feature of the lead and the size of the lead feature, and the red text represents the output pinch roller winding angle.

[0133] Based on the method for detecting the take-up angle of the pinch roll in the above embodiments, another embodiment of this application provides a detection system 1 for the take-up angle of the pinch roll, referring to... Figure 10 As shown, it includes:

[0134] Image processing module 12 is used to extract the outline features of the steel coil based on a pre-uncoiled image containing the steel coil and the uncoiler, and to determine the first reference point, the second reference point, the end point of the lead, the lead angle reference line and the pinch roll angle reference line.

[0135] Data processing module 13 is used to fit the first reference point, the second reference point and the end point of the belt head, and output the winding angle of the pinch roll.

[0136] In another embodiment of this application, reference is made to... Figure 10 As shown, the detection system 1 also includes:

[0137] Image acquisition module 11 is used to acquire pre-uncoil images containing steel coils and an uncoiler;

[0138] The data storage module 14 is equipped with a database for pre-storing the first reference point parameters and the second reference point parameters, so that the image processing module 12 can obtain them and determine them in the pre-unrolled image.

[0139] The image display module 15 is used to visualize at least one of the steel coil outline features, the lead angle reference line, the pinch roll angle reference line, and the pinch roll winding angle in the pre-unwinding image.

[0140] In this embodiment, the deep learning algorithm model is integrated into the image processing module 12 and the data processing module 13 to handle actual data processing, feature extraction, and angle calculation. Furthermore, the algorithm model can also be called by other modules via API, making the detection system 1 more flexible and scalable.

[0141] The pinch roll winding angle detection system 1 of this application enables unmanned detection of the pinch roll winding angle, preventing safety accidents when manually measuring the winding angle on site. At the same time, the pinch roll winding angle detected by this invention is more accurate and efficient than manual measurement, making it suitable for widespread application in steel coil plants to ensure that the winding angle before uncoiling the steel coil meets the operating standards and to guarantee a high yield of strip steel.

[0142] Based on the method for detecting the winding angle of the pinch roll in the above embodiments, another embodiment of this application also provides a device 1000 for detecting the winding angle of the pinch roll, see reference. Figure 11 As shown, the detection equipment 1000 includes a host computer 1001, an image display device 1002, and an image acquisition device 1003 obliquely aligned with the steel coil shaft center. The image acquisition device 1003 is used to acquire pre-uncoiling images. The image display device 1002 is used to visualize at least one of the following in the pre-uncoiling images: the steel coil outline features, the lead angle reference line, the pinch roll angle reference line, and the pinch roll winding angle. For specific visualization effects, please refer to [reference needed]. Figures 15-17 As shown; wherein, the host computer 1001 is configured as follows:

[0143] Based on the pre-uncoiling image containing the steel coil and the uncoiler, a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point perpendicular to the first reference point in the vertical direction and located on the pinch roll are determined respectively.

[0144] Steel coil contour feature data is extracted from pre-uncoiled images, and the lead end point is determined based on the steel coil contour feature data.

[0145] By fitting the first reference point, the second reference point, and the lead end point together, the winding angle of the pinch roll is output.

[0146] The pinch roll winding angle detection device 1000 of this application enables unmanned detection of the pinch roll winding angle, preventing safety accidents when manually measuring the winding angle on site. At the same time, the pinch roll winding angle detected by this invention is more accurate and efficient than manual measurement, making it suitable for widespread application in steel coil plants to ensure that the winding angle before uncoiling the steel coil meets the operating standards and to guarantee a high yield of strip steel.

[0147] It should be understood that the specific features, operations, and details described herein with respect to the methods of this application can also be similarly applied to the apparatus and system of this application, or vice versa. Furthermore, each step of the methods of this application described above can be performed by a corresponding component or unit of the apparatus or system of this application.

[0148] It should be understood that the various modules / units of the device of this application can be implemented wholly or partially through software, hardware, firmware, or a combination thereof. Each module / unit can be embedded in the processor of the electronic device in hardware or firmware form or independent of the processor, or it can be stored in the memory of the electronic device in software form for the processor to call to execute the operation of each module / unit. Each module / unit can be implemented as an independent component or module, or two or more modules / units can be implemented as a single component or module.

[0149] Based on the method for detecting the winding angle of the pinch roller in the above embodiments, another embodiment of this application also provides a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the detection method as described above.

[0150] Those skilled in the art will understand that the method steps of this application can be performed by a computer program instructing related hardware, such as electronic devices or processors. The computer program can be stored in a non-transitory computer-readable storage medium, and its execution causes the steps of this application to be performed. Depending on the context, any reference herein to memory, storage, or other media may include non-volatile or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state drive, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0151] The technical features described above can be combined arbitrarily. Although not all possible combinations of these technical features are described, any combination of these technical features should be considered to be covered by this specification, provided that such combination does not contain contradictions.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for detecting the winding angle of a pinch roller, characterized in that, Includes the following steps: Based on a pre-uncoiling image containing a steel coil and an uncoiler, a first reference point located at the center of the end face of the steel coil in the pre-uncoiling image and a second reference point perpendicular to the first reference point in the vertical direction and located on the pinch roll are determined respectively. Based on the pre-uncoiled image, extract the outline feature data of the steel coil, and determine the head end point according to the outline feature data of the steel coil; By fitting the first reference point, the second reference point and the head end point together, the winding angle of the pinch roll is output. The pre-unwinding image is acquired by an image acquisition device based on the obliquely aligned pinch roll axis. The step of extracting steel coil contour feature data based on the pre-uncoiled image and determining the lead end point based on the steel coil contour feature data includes: Based on the pre-uncoiled image, extract the outline feature data of the steel coil, and obtain the longitudinal section outline feature data of the strip head facing the steel coil axis and the transverse end face outline feature data of the strip head facing the pinch roll according to the outline feature data of the steel coil. The longitudinal section profile feature data and the transverse end face profile feature data with the head are respectively masked to obtain longitudinal section mask data and transverse end face mask data. The intersection point closest to the second reference point in the intersection region of the longitudinal section mask data and the transverse end face mask data is determined as the head endpoint.

2. The method for detecting the winding angle of the pinch roller according to claim 1, characterized in that, By fitting the first reference point, the second reference point, and the lead end point together, the output pinch roll winding angle includes: Determine the head angle reference line based on the head endpoint and the first reference point; Determine the pinch roll angle reference line based on the first reference point and the second reference point; The winding angle of the pinch roll is calculated and output based on the head angle reference line and the pinch roll angle reference line.

3. The method for detecting the winding angle of the pinch roller according to claim 2, characterized in that, The winding angle of the pinch roll is calculated and output based on the head angle reference line and the pinch roll angle reference line, including: The belt head horizontal angle is obtained based on the slope of the belt head angle reference line; the pinch roll horizontal angle is obtained based on the slope of the pinch roll angle reference line; the difference between the belt head horizontal angle and the pinch roll horizontal angle is calculated to output the pinch roll winding angle; or The winding angle of the pinch roll is output based on the angle between the head angle reference line and the pinch roll angle reference line.

4. The method for detecting the winding angle of the pinch roller according to claim 1, characterized in that, By fitting the first reference point, the second reference point, and the lead end point together, the output pinch roll winding angle includes: Obtain the coordinate parameters of the first reference point, the second reference point, and the head endpoint in the preset coordinate system, respectively. Based on the obtained coordinate parameters, the winding angle of the pinch roller is calculated and output.

5. The method for detecting the winding angle of the pinch roller according to claim 1, characterized in that, Based on a pre-uncoiling image including a steel coil and an uncoiler, a first reference point located at the center of the end face of the steel coil and a second reference point perpendicular to the first reference point and located on the pinch roll are determined in the pre-uncoiling image, including: Extract the first reference point parameters and the second reference point parameters, and determine the first reference point and the second reference point in the pre-unrolled image; The first reference point parameters and the second reference point parameters are stored in a priori database.

6. The method for detecting the winding angle of the pinch roller according to claim 1, characterized in that, Also includes: The winding angle of the pinch roller is compared with a preset threshold. Based on whether the winding angle of the pinch roller is greater than the preset threshold, it is determined whether the winding angle of the pinch roller is accurate. and / or The winding angles of the pinch rollers, which are continuously output, are curve-fitted to obtain a fitting curve. The accuracy of the winding angle of the pinch rollers is determined based on the growth trend of the fitting curve.

7. A detection system for the winding angle of a pinch roller, characterized in that, A method for detecting the take-up angle of the pinch roll as described in any one of claims 1 to 6, comprising: The image processing module is used to extract the outline feature data of the steel coil based on the pre-uncoil image containing the steel coil and the uncoiler, and to determine the first reference point, the second reference point, and the head end point; The data processing module is used to fit the first reference point, the second reference point and the head end point together, and output the winding angle of the pinch roll.

Images