A method, device, equipment and medium for detecting a bending state of a rolling mill tapping

Abstract

The application discloses a rolling mill tapping bending state detection method and device, equipment and medium, relates to computer technology field, including determining each single frame image by using video stream, generating reference object image, carrying out coordinate calculation, obtaining reference object contour pixel coordinates and rolling piece contour pixel coordinates; generating reference object axis pixel coordinate point set and rolling piece axis pixel coordinate point set, generating rolling piece axis fitting equation and reference object axis fitting equation; calculating the rolling piece axis fitting equation, obtaining the bending measure, carrying out coordinate conversion on the reference object axis pixel coordinate point set, obtaining the plane size coordinates, calculating the rolling piece chord height according to the plane size coordinates and the bending measure; based on the rolling piece axis fitting equation, the reference object axis fitting equation, the bending measure and the rolling piece chord height, the rolling mill tapping bending state detection result is generated. The application can reduce the labor intensity of artificial visual inspection, reduce the production safety accident rate, improve the rolling mill tapping product quality and efficiency.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, apparatus, equipment and medium for detecting the bending state of steel produced from a rolling mill. Background Technology

[0002] During steel rolling, the bending state of the steel sections at the tapping point significantly impacts production safety and product surface quality. Excessive bending can cause chipping at the ends of the rolled piece, damage to the rolls, and, in high-speed rolling, even lead to the piece slipping out of the rolling mill. Currently, most steel rolling mills rely on manual visual inspection, while some have installed surveillance cameras. This not only involves high labor intensity but also poses a significant risk of quality and safety accidents.

[0003] As can be seen from the above, how to reduce the labor intensity of manual visual inspection, reduce the rate of production safety accidents, and improve the product quality and efficiency of steel produced by rolling mills are problems that need to be solved in this field. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a method, apparatus, equipment, and medium for detecting the bending state of steel produced from a rolling mill, which can reduce the labor intensity of manual visual inspection, reduce the rate of production safety accidents, and improve the product quality and efficiency of steel produced from the rolling mill. The specific solution is as follows:

[0005] In a first aspect, this application discloses a method for detecting the bending state of steel exiting a rolling mill, comprising:

[0006] The video stream of steel being produced from the rolling mill is acquired, and each single frame image is determined using the video stream. A reference object image is generated based on each single frame image. Coordinate calculations are performed on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour.

[0007] A set of reference object axis pixel coordinate points and a set of rolling mill axis pixel coordinate points are generated based on the reference object contour pixel coordinate points and the rolling mill contour pixel coordinate points. A fitting equation for the rolling mill axis and a fitting equation for the reference object axis are generated based on the reference object axis pixel coordinate point set and the rolling mill axis pixel coordinate point set.

[0008] The fitting equation of the centerline of the rolled piece is calculated to obtain the bending metric, and the coordinate conversion of the set of pixel coordinate points of the reference object axis is performed to obtain the planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric.

[0009] Based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece, the bending state detection result of the steel output from the rolling mill is generated.

[0010] Optionally, acquiring the video stream of steel exiting the rolling mill and using the video stream to determine each single frame image includes:

[0011] The video stream of steel being discharged from the rolling mill is acquired using a preset image measurement gateway;

[0012] The video stream is cropped based on the region of interest range setting to obtain each of the single-frame images.

[0013] Optionally, generating a reference object image based on each of the single-frame images includes:

[0014] The reference object image is generated by calculating the pixel coordinates of each single frame image using an image similarity calculation algorithm and based on a preset reference object image template.

[0015] Optionally, the step of calculating coordinates for the reference object image and the single-frame image respectively to obtain the reference object contour pixel coordinates and the rolled piece contour pixel coordinates includes:

[0016] A closing-opening operation is performed on the reference image and the rolled piece image in the single frame image to obtain the reference image and the rolled piece image after the operation.

[0017] The reference image and the rolled piece image after the operation are binarized, and a contour extraction algorithm is used to extract the contour of the binarized image to obtain the pixel coordinates of the reference image contour and the pixel coordinates of the rolled piece contour.

[0018] Optionally, generating the set of reference object axis pixel coordinate points and the set of rolling mill axis pixel coordinate points based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates includes:

[0019] A binary image of the reference object and a binary image of the rolled piece are generated based on the pixel coordinates of the reference object's outline and the pixel coordinates of the rolled piece's outline, respectively.

[0020] The median calculation method is used to calculate the binary image of the reference object and the binary image of the rolled piece to obtain the set of pixel coordinate points of the axis of the reference object and the set of pixel coordinate points of the axis of the rolled piece.

[0021] Optionally, generating the fitting equation for the centerline of the rolled piece and the fitting equation for the centerline of the reference object based on the set of pixel coordinate points of the reference object's axis and the set of pixel coordinate points of the rolled piece's axis includes:

[0022] Define the equations for the centerline curves of the reference object and the centerline curves of the rolled piece, respectively.

[0023] The least squares fitting algorithm is used to substitute the set of pixel coordinates of the reference object axis and the set of pixel coordinates of the rolling mill axis into the curve equations of the reference object axis and the rolling mill axis, respectively, to fit and generate various coefficients, so as to obtain the fitting equations of the reference object axis and the rolling mill axis.

[0024] Optionally, the calculation of the fitting equation for the centerline of the rolled piece to obtain the bending metric includes:

[0025] The tangent point is calculated by sliding through all pixels on the fitted equation of the centerline of the rolled piece.

[0026] Based on the tangent point, a tangent point equation is generated, the angle between the tangent point equation and the fitting equation of the central axis of the reference object is calculated, and the angle value is used as the curvature measure.

[0027] Secondly, this application discloses a bending state detection device for steel exiting a rolling mill, comprising:

[0028] The coordinate generation module is used to acquire the video stream of steel output from the rolling mill, determine each single frame image using the video stream, generate a reference object image based on each single frame image, and perform coordinate calculations on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour.

[0029] The fitting equation generation module is used to generate a set of reference object axis pixel coordinate points and a set of rolling mill axis pixel coordinate points based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates, and to generate a fitting equation for the rolling mill axis and a fitting equation for the reference object axis based on the set of reference object axis pixel coordinate points and the set of rolling mill axis pixel coordinate points.

[0030] The calculation module is used to calculate the fitting equation of the centerline of the rolled piece to obtain the bending metric, and to perform coordinate transformation on the set of pixel coordinate points of the reference axis to obtain the planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric.

[0031] The detection result generation module is used to generate the bending state detection result of the rolling mill exiting the steel based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece.

[0032] Thirdly, this application discloses an electronic device, including:

[0033] Memory, used to store computer programs;

[0034] A processor is used to execute the computer program to implement the aforementioned method for detecting the bending state of steel produced from a rolling mill.

[0035] Fourthly, this application discloses a computer storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the steps of the aforementioned disclosed method for detecting the bending state of steel produced from a rolling mill.

[0036] As can be seen, this application provides a method for detecting the bending state of steel exiting a rolling mill, including acquiring a video stream of steel exiting the rolling mill, determining each single-frame image using the video stream, generating a reference object image based on each single-frame image, performing coordinate calculations on the reference object image and the single-frame image respectively to obtain the reference object contour pixel coordinates and the rolled piece contour pixel coordinates; generating a set of reference object axis pixel coordinate points and a set of rolled piece axis pixel coordinate points based on the reference object contour pixel coordinates and the rolled piece contour pixel coordinates, and generating a fitting equation for the centerline of the rolled piece and a fitting equation for the centerline of the reference object based on the set of reference object axis pixel coordinate points and the set of rolled piece axis pixel coordinate points; calculating the fitting equation for the centerline of the rolled piece to obtain a bending metric, and performing coordinate transformation on the set of reference object axis pixel coordinate points to obtain planar dimension coordinates; calculating the chord height of the rolled piece based on the planar dimension coordinates and the bending metric; and generating a bending state detection result of steel exiting the rolling mill based on the fitting equation for the centerline of the rolled piece, the fitting equation for the centerline of the reference object, the bending metric, and the chord height of the rolled piece. This application generates a reference image from a single frame image determined by a video stream, and then performs coordinate calculations to obtain the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolling mill axis. It then generates fitting equations for the centerline of the rolling mill and the centerline of the reference object, calculates the bending metric and the chord height of the rolling mill, and generates bending state detection results for the steel mill exiting the mill. This reduces the labor intensity of manual visual inspection. Based on the bending state detection results for the steel mill exiting the mill, it can not only reduce the rate of production safety accidents, but also improve the product quality and efficiency of the steel mill exiting the mill. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0038] Figure 1 This application discloses a flowchart of a method for detecting the bending state of steel exiting a rolling mill.

[0039] Figure 2 This application discloses another method for detecting the bending state of steel exiting a rolling mill;

[0040] Figure 3This application discloses a system architecture diagram for detecting the bending state of steel exiting a rolling mill.

[0041] Figure 4 This is a diagram of a measurement result display window disclosed in this application;

[0042] Figure 5 This application discloses a measurement area calibration window diagram;

[0043] Figure 6 This is a window diagram illustrating the definition of affine transformation parameters disclosed in this application;

[0044] Figure 7 This is a flowchart of an algorithm parameter configuration window method disclosed in this application;

[0045] Figure 8 Flowchart of another algorithm parameter configuration window method disclosed in this application;

[0046] Figure 9 This is a schematic diagram of the bending state detection device for steel output from a rolling mill disclosed in this application;

[0047] Figure 10 This application provides a structural diagram of an electronic device. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] During steel rolling, the bending state of the steel sections exiting the rolling mill significantly impacts production safety and product surface quality. Excessive bending can cause chipping at the ends of the rolled piece, damage to the rolls, and, in high-speed rolling, even lead to the piece slipping out of the rolling mill. Currently, most steel rolling mills rely on manual visual inspection, while some have surveillance cameras. This not only involves high labor intensity but also increases the risk of quality and safety accidents. Therefore, reducing the labor intensity of manual visual inspection, decreasing the accident rate, and improving the product quality and efficiency of rolling mill output are pressing issues that need to be addressed in this field.

[0050] See Figure 1 As shown in the figure, this invention discloses a method for detecting the bending state of steel produced from a rolling mill, which specifically includes:

[0051] Step S11: Obtain the video stream of steel exiting the rolling mill, determine each single frame image using the video stream, generate a reference object image based on each single frame image, and perform coordinate calculations on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour.

[0052] In this embodiment, after generating the reference image, a closing-opening operation is performed on the reference image and the rolled piece image in the single frame image to obtain the processed reference image and the processed rolled piece image; the processed reference image and the processed rolled piece image are then subjected to image binarization processing, and a contour extraction algorithm is used to extract the contour of the binarized image to obtain the contour pixel coordinates of the reference image and the contour pixel coordinates of the rolled piece.

[0053] Specifically, after generating the reference image, the reference image and the rolled piece image are subjected to closure-opening operations according to the erosion kernel and expansion kernel, respectively, to form the processed reference image and the processed rolled piece image. Then, the processed reference image and the processed rolled piece image are subjected to image binarization processing. At the same time, the contour extraction algorithm is used to extract the contours of the binary processed images, thereby obtaining the contour pixel coordinates of the measured reference object and the rolled piece entity. These two sets of coordinates are recorded to obtain the contour pixel coordinates of the reference object and the contour pixel coordinates of the rolled piece.

[0054] Step S12: Generate a set of reference object axis pixel coordinates and a set of rolling mill axis pixel coordinates based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates, and generate a fitting equation for the rolling mill axis and a fitting equation for the reference object axis based on the reference object axis pixel coordinates and the rolling mill axis pixel coordinates.

[0055] In this embodiment, binary images of the reference object and the rolled piece are generated based on the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour, respectively. The median calculation method is used to calculate the binary images of the reference object and the rolled piece to obtain the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolled piece axis. Then, the equations of the reference object axis curve and the rolled piece axis curve are defined respectively. Using a least squares fitting algorithm, the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolled piece axis are substituted into the equations of the reference object axis curve and the rolled piece axis curve, respectively, to fit and generate various coefficients, thereby obtaining the fitting equations of the reference object axis curve and the rolled piece axis curve.

[0056] Specifically, based on the two contour pixel coordinates, two binary images of the same size as the single-frame image are generated. These binary images contain only contour information. Then, the median calculation method is used to calculate the set of central axis pixel coordinates {(x...}) for each entity contour in the contour images. my m )},x m The x-coordinate of the central axis is y. m This represents the ordinate value of the centerline. Using the image width and length as the X and Y axes respectively, and the X-axis as the scan axis, calculate the x-coordinate of the contour points on the X-axis. p Given, the y-axis value of the corresponding entity contour point is {y p The average value of}, i.e., x m =x p y m =∑{y p} / N, sequentially scan the X-axis to calculate the set of points along the central axis, where x p Let y be the x-coordinate pixel value of any point on the contour. p Let x be the ordinate pixel value of any point on the contour. p y p Let N be any coordinate point on the contour, and let N be the x-coordinate. p Total number of contour points at coordinates.

[0057] In this embodiment, the equations of the centerline curves of the rolled piece and the reference object are defined respectively. The former is generally defined as a quadratic curve, i.e., y = a*x. 2 +b*x+c, where x is the pixel x-coordinate, y is the pixel x-coordinate, a is the quadratic coefficient, b is the linear coefficient, and c is the constant term. The latter is generally defined as a straight line equation, i.e., y=d*x+f, where d is the linear coefficient and f is the constant term. According to the least squares fitting algorithm, the pixel positions of the rolled piece and the reference object are substituted into the fitting equation of the reference object's centerline and the fitting equation of the rolled piece's centerline, respectively, to generate the coefficients of each term, thereby obtaining the fitting equation of the centerline of the rolled piece and the reference object.

[0058] Step S13: Calculate the fitting equation of the centerline of the rolled piece to obtain the bending metric, and perform coordinate conversion on the set of pixel coordinate points of the reference axis to obtain the planar dimension coordinates. Calculate the chord height of the rolled piece based on the planar dimension coordinates and the bending metric.

[0059] In this embodiment, all pixels on the fitted equation of the centerline of the rolled piece are traversed by sliding to calculate the tangent point. Based on the tangent point, a tangent point equation is generated. The angle between the tangent point equation and the fitted equation of the centerline of the reference object is calculated, and the angle is used as the curvature metric. The set of pixel coordinate points of the reference object's centerline is converted to obtain the planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the curvature metric.

[0060] Specifically, the process involves sliding through all pixels along the central axis of the workpiece at intervals of β. Each slide calculates the average value of α points as the tangent point, generating the tangent equation. Simultaneously, the angle between the tangent and the reference axis is calculated and recorded. This angle is used as one of the indicators of bending; a larger angle indicates greater bending. However, excessively large values ​​of α can distort the bending degree. A conversion relationship is generated based on the actual spatial dimensions of the reference object and the image pixel dimensions, converting image coordinates to planar coordinates. The fitted axis of the reference object is translated to the point within the first image intersecting the workpiece. Based on the coordinate conversion relationship in the planar coordinates and the measurement angle, the chord height at the end of the workpiece is calculated: H = L * tan(θ), where H is the chord height, L is the projection of the tangent of the central axis at the end of the workpiece onto the translated reference axis, and θ is the angle between the two axes.

[0061] Step S14: Generate the bending state detection result of the rolling mill exiting the steel based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece.

[0062] In this embodiment, a video stream of steel exiting the rolling mill is acquired, and each single frame image is determined using the video stream. A reference object image is generated based on each single frame image. Coordinate calculations are performed on the reference object image and the single frame image to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour. A set of pixel coordinate points for the reference object axis and a set of pixel coordinate points for the rolled piece axis are generated based on the pixel coordinates of the reference object axis and the rolled piece axis. A fitting equation for the centerline of the rolled piece and a fitting equation for the centerline of the reference object axis are generated based on the pixel coordinates of the reference object axis and the set of pixel coordinate points of the rolled piece axis. The fitting equation for the centerline of the rolled piece is calculated to obtain a bending metric, and coordinate conversion is performed on the set of pixel coordinate points for the reference object axis to obtain planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric. The bending state detection result of steel exiting the rolling mill is generated based on the fitting equation for the centerline of the rolled piece, the fitting equation for the centerline of the reference object axis, the bending metric, and the chord height of the rolled piece. This application generates a reference image from a single frame image determined by a video stream, and then performs coordinate calculations to obtain the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolling mill axis. It then generates fitting equations for the centerline of the rolling mill and the centerline of the reference object, calculates the bending metric and the chord height of the rolling mill, and generates bending state detection results for the steel mill exiting the mill. This reduces the labor intensity of manual visual inspection. Based on the bending state detection results for the steel mill exiting the mill, it can not only reduce the rate of production safety accidents, but also improve the product quality and efficiency of the steel mill exiting the mill.

[0063] See Figure 2 As shown in the figure, this invention discloses a method for detecting the bending state of steel produced from a rolling mill, which specifically includes:

[0064] Step S21: Use a preset image measurement gateway to acquire the video stream of steel output from the rolling mill, and crop the video stream based on the region of interest range setting value to obtain each single frame image.

[0065] Before acquiring the video stream, this application requires the design and construction of detection equipment, including: setting up cameras at a downward angle of 45° to 90° within a horizontal distance of 1 to 20 meters and a vertical distance of 1 to 10 meters from the mill exit; and constructing an image measurement gateway, a shape detection server, and a data storage server. The image measurement gateway is responsible for acquiring the video stream and transmitting network data, including real-time video stream data transmission; the shape detection server is responsible for receiving and measuring video data; and the data storage server is responsible for backing up and storing historical data. The specific system architecture of this application is as follows: Figure 3 As shown, it includes an image measurement gateway, an industrial control gateway, a client, a shape detection server, a data storage server, an industrial controller, an L1 sequential control system, and detection equipment, including but not limited to surveillance cameras and infrared cameras.

[0066] In this embodiment, after acquiring the video stream of steel output from the rolling mill, the video stream is read frame by frame according to the video stream reading protocol and the downsampling setting frequency to obtain a single frame image after reading. Then, the single frame image after reading is cropped according to the pre-given region of interest range setting value to obtain the cropped single frame image. The cropped single frame image is used as the image for the next step of processing.

[0067] Step S22: Calculate the reference object pixel coordinates of each single frame image using an image similarity calculation algorithm and based on a preset reference object image template to generate a reference object image. Calculate the coordinates of the reference object image and the single frame image respectively to obtain the reference object contour pixel coordinates and the rolled piece contour pixel coordinates.

[0068] In this embodiment, based on a preset threshold range of rolled piece pixels, the set of rolled piece pixel coordinates is filtered and extracted, and an image containing only the rolled piece is generated. Simultaneously, based on a given reference object image template, the pixel coordinates of the reference object are extracted according to an image similarity calculation algorithm, and a reference object image is generated. The specific implementation process of the image similarity calculation algorithm is as follows: using the reference object image as the kernel template, in the reference object calibration area, the convolution value between the convolution kernel and its covered pixel area is calculated by convolution traversal. After the traversal, the area with the largest convolution value is taken as the actual reference object area, i.e., MAX({∑(h(z0-z)*i(z))}), where h(z) is the convolution kernel, i(z) is the image area of ​​the same size as the convolution kernel, z is the pixel area, and z0 is the convolution stride. Then, the set of pixel coordinates of this area is extracted from the cropped single-frame image and filled to generate an image containing only the reference object, finally obtaining the reference object image.

[0069] Step S23: Generate a set of reference object axis pixel coordinates and a set of rolling mill axis pixel coordinates based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates, and generate a fitting equation for the rolling mill axis and a fitting equation for the reference object axis based on the reference object axis pixel coordinates and the rolling mill axis pixel coordinates.

[0070] Step S24: Calculate the fitting equation of the centerline of the rolled piece to obtain the bending metric, and perform coordinate conversion on the set of pixel coordinate points of the reference axis to obtain the planar dimension coordinates. Calculate the chord height of the rolled piece based on the planar dimension coordinates and the bending metric.

[0071] Step S25: Generate the bending state detection result of the rolling mill exiting the steel based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece.

[0072] In addition, this application can also implement the design of software functions, including a measurement result display window, measurement area calibration, algorithm parameter configuration, and communication parameter configuration.

[0073] (1) Measurement Result Display Window: This window is the portal to the entire system application and the main operation interface, including the menu bar, measurement information display bar, image and measurement process display bar, and log bar. I. Menu Bar: i. Run and Stop buttons are used to start and stop the measurement process; ii. Single Frame Capture button is used to capture a single image captured by the current camera; iii. Clicking the Parameter Configuration button will pop up the "Algorithm Parameter Configuration" window; iv. Clicking the Communication Configuration button will pop up the "Communication Parameter Configuration" window; v. Clicking the Measurement Calibration button will pop up the "Measurement Area Calibration" window; vi. Clicking the Exit button will close the program window and exit the program, releasing memory. II. Measurement Information Display Bar: Used to display measurement process status information and comprehensive measurement result information, including: current frame number, minimum goodness of fit value of the centerline, maximum curvature, and maximum curvature chord information. III. Image and Measurement Process Display Bar: Used to display real-time images from the camera and additional measurement processing information, including: centerline of the reference object and the rolled piece entity, measurement auxiliary line information, curvature evaluation display and alarm. IV. Log Bar: Displays software operation records and measurement process event information. Finally, based on the workpiece centerline fitting equation, the reference object centerline fitting equation, bending measurement, and workpiece chord height, it generates the bending state detection results at the mill exit. The main interface of the detection results display window is as follows: Figure 4 As shown.

[0074] (2) Measurement area calibration: The measurement area calibration window is as follows Figure 5 As shown, the calibration process refers to defining the reference line and calculating the transformation relationship between pixel coordinates and the world coordinate system. After clicking the "Measurement Area Calibration" button on the main interface, the calibration process is as follows: I. Define the reference line rectangular area: Clicking this button will bring up a rectangle drawing tool in the measurement image display area. Simply click to select the opposite corner of the rectangle in the image to generate the corresponding rectangular area, draw it, and save it. II. Define the reference line: Clicking this button will bring up a line segment drawing tool in the measurement image display area. Draw a reference straight line that matches the actual scene within the defined reference line rectangular area. Simply click to select two endpoints in the image to generate the corresponding reference line segment, draw it, and save it. III. Define the measurement rectangular area: Clicking this button will bring up a rectangle drawing tool in the measurement image display area. Simply click to select the opposite corner of the rectangle in the image to generate the corresponding rectangular area, draw it, and save it. IV. Define affine transformation parameters: Clicking this button will bring up the "Affine Transformation Parameter Definition" window. In this window, you can modify the transformation parameters and select the world coordinate relative to the survey coordinate raster file. The affine transformation parameter definition window is shown below. Figure 6As shown. V. One-click calibration: Clicking this button will invoke the calibration algorithm to calculate calibration parameters, including the calculation of the affine transformation matrix. The affine transformation matrix represents the transformation relationship between pixel coordinates and world coordinates, and is formally expressed as AXP = YW, where A is the affine transformation matrix, XP is the pixel coordinate position, and YW is the world coordinate position. The specific algorithm can be solved using the least squares method.

[0075] (3) Algorithm parameter configuration: The algorithm parameter configuration window is as follows Figure 7 and Figure 8 As shown, in this module, the constants and methods involved in the algorithm can be dynamically modified and configured, thereby improving the adaptability of the algorithm. After clicking the "Parameter Configuration" button in the main interface menu bar, the algorithm parameter configuration window will pop up.

[0076] (4) Communication parameter configuration: In this module, the camera IP address, database address and PLC trigger signal address can be modified and configured. After clicking the "Communication Configuration" button in the main interface menu bar, the communication parameter configuration window will pop up.

[0077] This application develops detection algorithms and system software, with algorithm testing conducted using recorded video. Image data is measured using on-site surveillance cameras, eliminating the need for mechanical installations and sensor hardware costs. The camera is disconnected from the original video network, and a video switch is set up to connect the camera to this network. The newly built video switch is then connected to the original video switch, thus avoiding increasing the bandwidth pressure on the original video network. Furthermore, this application enables joint system debugging and on-site trial operation, allowing for iterative revisions of the detection algorithm and functional displays.

[0078] In this embodiment, a video stream of steel exiting the rolling mill is acquired, and each single frame image is determined using the video stream. A reference object image is generated based on each single frame image. Coordinate calculations are performed on the reference object image and the single frame image to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour. A set of pixel coordinate points for the reference object axis and a set of pixel coordinate points for the rolled piece axis are generated based on the pixel coordinates of the reference object axis and the rolled piece axis. A fitting equation for the centerline of the rolled piece and a fitting equation for the centerline of the reference object axis are generated based on the pixel coordinates of the reference object axis and the set of pixel coordinate points of the rolled piece axis. The fitting equation for the centerline of the rolled piece is calculated to obtain a bending metric, and coordinate conversion is performed on the set of pixel coordinate points for the reference object axis to obtain planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric. The bending state detection result of steel exiting the rolling mill is generated based on the fitting equation for the centerline of the rolled piece, the fitting equation for the centerline of the reference object axis, the bending metric, and the chord height of the rolled piece. This application generates a reference image from a single frame image determined by a video stream, and then performs coordinate calculations to obtain the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolling mill axis. It then generates fitting equations for the centerline of the rolling mill and the centerline of the reference object, calculates the bending metric and the chord height of the rolling mill, and generates bending state detection results for the steel mill exiting the mill. This reduces the labor intensity of manual visual inspection. Based on the bending state detection results for the steel mill exiting the mill, it can not only reduce the rate of production safety accidents, but also improve the product quality and efficiency of the steel mill exiting the mill.

[0079] See Figure 9 As shown in the figure, an embodiment of the present invention discloses a bending state detection device for steel output from a rolling mill, which may specifically include:

[0080] The coordinate generation module 11 is used to acquire the video stream of steel output from the rolling mill, determine each single frame image using the video stream, generate a reference object image based on each single frame image, and perform coordinate calculations on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour.

[0081] The fitting equation generation module 12 is used to generate a set of reference object axis pixel coordinate points and a set of rolling mill axis pixel coordinate points based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates, and to generate a fitting equation for the rolling mill axis and a fitting equation for the reference object axis based on the set of reference object axis pixel coordinate points and the set of rolling mill axis pixel coordinate points.

[0082] The calculation module 13 is used to calculate the fitting equation of the centerline of the rolled piece to obtain the bending metric, and to perform coordinate conversion on the set of pixel coordinate points of the reference axis to obtain the planar dimension coordinates, and to calculate the chord height of the rolled piece based on the planar dimension coordinates and the bending metric.

[0083] The detection result generation module 14 is used to generate the bending state detection result of the rolling mill exiting the steel based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece.

[0084] In this embodiment, a video stream of steel exiting the rolling mill is acquired, and each single frame image is determined using the video stream. A reference object image is generated based on each single frame image. Coordinate calculations are performed on the reference object image and the single frame image to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour. A set of pixel coordinate points for the reference object axis and a set of pixel coordinate points for the rolled piece axis are generated based on the pixel coordinates of the reference object axis and the rolled piece axis. A fitting equation for the centerline of the rolled piece and a fitting equation for the centerline of the reference object axis are generated based on the pixel coordinates of the reference object axis and the set of pixel coordinate points of the rolled piece axis. The fitting equation for the centerline of the rolled piece is calculated to obtain a bending metric, and coordinate conversion is performed on the set of pixel coordinate points for the reference object axis to obtain planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric. The bending state detection result of steel exiting the rolling mill is generated based on the fitting equation for the centerline of the rolled piece, the fitting equation for the centerline of the reference object axis, the bending metric, and the chord height of the rolled piece. This application generates a reference image from a single frame image determined by a video stream, and then performs coordinate calculations to obtain the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolling mill axis. It then generates fitting equations for the centerline of the rolling mill and the centerline of the reference object, calculates the bending metric and the chord height of the rolling mill, and generates bending state detection results for the steel mill exiting the mill. This reduces the labor intensity of manual visual inspection. Based on the bending state detection results for the steel mill exiting the mill, it can not only reduce the rate of production safety accidents, but also improve the product quality and efficiency of the steel mill exiting the mill.

[0085] In some specific embodiments, the coordinate generation module 11 may specifically include:

[0086] The video stream acquisition module is used to acquire the video stream of steel output from the rolling mill using a preset image measurement gateway;

[0087] The cropping module is used to crop the video stream based on a set value for the region of interest range to obtain each of the single-frame images.

[0088] In some specific embodiments, the coordinate generation module 11 may specifically include:

[0089] The coordinate calculation module is used to calculate the reference object pixel coordinates of each single frame image using an image similarity calculation algorithm and based on a preset reference object image template, so as to generate a reference object image.

[0090] In some specific embodiments, the coordinate generation module 11 may specifically include:

[0091] The closing-opening operation module is used to perform closing-opening operations on the reference image and the rolled piece image in the single frame image to obtain the reference image and the rolled piece image after the operation.

[0092] The image binarization processing module is used to perform image binarization processing on the processed reference object image and the processed rolled piece image, and to use a contour extraction algorithm to extract the contour of the binarized image to obtain the contour pixel coordinates of the reference object and the contour pixel coordinates of the rolled piece.

[0093] In some specific embodiments, the fitting equation generation module 12 may specifically include:

[0094] The binary image generation module is used to generate a binary image of the reference object and a binary image of the rolled piece based on the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour, respectively.

[0095] The coordinate point set calculation module is used to calculate the binary image of the reference object and the binary image of the rolled piece using the median calculation method, so as to obtain the pixel coordinate point set of the reference object axis and the pixel coordinate point set of the rolled piece axis.

[0096] In some specific embodiments, the fitting equation generation module 12 may specifically include:

[0097] The centerline curve equation definition module is used to define the centerline curve equations of the reference object and the rolled piece, respectively.

[0098] The centerline curve equation generation module is used to substitute the set of pixel coordinate points of the reference object axis and the set of pixel coordinate points of the rolling mill axis into the centerline curve equation of the reference object and the centerline curve equation of the rolling mill respectively according to the least squares fitting algorithm, and to fit and generate various coefficients to obtain the centerline fitting equation of the reference object and the centerline fitting equation of the rolling mill.

[0099] In some specific embodiments, the computing module 13 may specifically include:

[0100] The tangent point calculation module is used to slide through all the pixels on the fitted equation of the centerline of the rolled piece to calculate the tangent point;

[0101] The bending metric calculation module is used to generate a tangent point equation based on the tangent point, calculate the angle between the tangent point equation and the fitting equation of the central axis of the reference object, and use the angle value as the bending metric.

[0102] Figure 10This is a schematic diagram of an electronic device provided in an embodiment of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the bending state detection method for steel tapping from a rolling mill, as disclosed in any of the foregoing embodiments.

[0103] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0104] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored on it include operating system 221, computer program 222 and data 223, etc., and the storage method can be temporary storage or permanent storage.

[0105] The operating system 221 manages and controls the various hardware devices on the electronic device 20 and the computer program 222 to enable the processor 21 to perform calculations and processing on the data 223 in the memory 22. It can be Windows, Unix, Linux, etc. The computer program 222, in addition to including a computer program capable of performing the bending state detection method for rolling mill steel output disclosed in any of the foregoing embodiments, may further include computer programs capable of performing other specific tasks. The data 223 may include data received by the bending state detection device for rolling mill steel output from external devices, as well as data collected by its own input / output interface 25.

[0106] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0107] Furthermore, this application also discloses a computer-readable storage medium storing a computer program. When the computer program is loaded and executed by a processor, it implements the steps of the bending state detection method for steel tapping from a rolling mill disclosed in any of the foregoing embodiments.

[0108] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0109] The present invention provides a detailed description of a method, apparatus, equipment, and storage medium for detecting the bending state of steel produced from a rolling mill. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for detecting the bending state of steel exiting a rolling mill, characterized in that, include: The video stream of steel being produced from the rolling mill is acquired, and each single frame image is determined using the video stream. A reference object image is generated based on each single frame image. Coordinate calculations are performed on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour. A set of reference object axis pixel coordinate points and a set of rolling mill axis pixel coordinate points are generated based on the reference object contour pixel coordinate points and the rolling mill contour pixel coordinate points. A fitting equation for the rolling mill axis and a fitting equation for the reference object axis are generated based on the reference object axis pixel coordinate point set and the rolling mill axis pixel coordinate point set. The fitting equation of the centerline of the rolled piece is calculated to obtain the bending metric, and the coordinate conversion of the set of pixel coordinate points of the reference object axis is performed to obtain the planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric. Based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending metric, and the chord height of the rolled piece, the bending state detection result of the steel exiting the mill is generated; The process of generating a reference object image based on each of the single-frame images includes: using an image similarity calculation algorithm and based on a preset reference object image template to calculate the reference object pixel coordinates of each single-frame image to generate a reference object image. The bending metric is obtained by calculating the fitting equation of the centerline of the rolled piece, including: sliding through all pixels on the fitting equation of the centerline of the rolled piece to calculate the tangent point; generating the tangent point equation based on the tangent point; calculating the angle between the tangent point equation and the fitting equation of the centerline of the reference object; and using the angle as the bending metric.

2. The method for detecting the bending state of steel exiting a rolling mill according to claim 1, characterized in that, The process of acquiring the video stream of steel output from the rolling mill and using the video stream to determine each single frame image includes: The video stream of steel being discharged from the rolling mill is acquired using a preset image measurement gateway; The video stream is cropped based on the region of interest range setting to obtain each of the single-frame images.

3. The method for detecting the bending state of steel exiting a rolling mill according to claim 1, characterized in that, The step of calculating coordinates for the reference object image and the single-frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour includes: A closing-opening operation is performed on the reference image and the rolled piece image in the single frame image to obtain the reference image and the rolled piece image after the operation. The reference image and the rolled piece image after the operation are binarized, and a contour extraction algorithm is used to extract the contour of the binarized image to obtain the pixel coordinates of the reference image contour and the pixel coordinates of the rolled piece contour.

4. The method for detecting the bending state of steel exiting a rolling mill according to claim 1, characterized in that, The step of generating a set of reference object axis pixel coordinate points and a set of rolled piece axis pixel coordinate points based on the reference object contour pixel coordinates and the rolled piece contour pixel coordinates includes: A binary image of the reference object and a binary image of the rolled piece are generated based on the pixel coordinates of the reference object's outline and the pixel coordinates of the rolled piece's outline, respectively. The median calculation method is used to calculate the binary image of the reference object and the binary image of the rolled piece to obtain the set of pixel coordinate points of the axis of the reference object and the set of pixel coordinate points of the axis of the rolled piece.

5. The method for detecting the bending state of steel exiting a rolling mill according to claim 1, characterized in that, The process of generating the fitting equations for the centerline of the rolled piece and the centerline of the reference object based on the set of pixel coordinate points of the reference object's axis and the set of pixel coordinate points of the rolled piece's axis includes: Define the equations for the centerline curves of the reference object and the centerline curves of the rolled piece, respectively. The least squares fitting algorithm is used to substitute the set of pixel coordinates of the reference object axis and the set of pixel coordinates of the rolling mill axis into the curve equations of the reference object axis and the rolling mill axis, respectively, to fit and generate various coefficients, so as to obtain the fitting equations of the reference object axis and the rolling mill axis.

6. A device for detecting the bending state of steel produced from a rolling mill, characterized in that, include: The coordinate generation module is used to acquire the video stream of steel output from the rolling mill, determine each single frame image using the video stream, generate a reference object image based on each single frame image, and perform coordinate calculations on the reference object image and the single frame image respectively to obtain the pixel coordinates of the reference object contour and the pixel coordinates of the rolled piece contour. The fitting equation generation module is used to generate a set of reference object axis pixel coordinate points and a set of rolling mill axis pixel coordinate points based on the reference object contour pixel coordinates and the rolling mill contour pixel coordinates, and to generate a fitting equation for the rolling mill axis and a fitting equation for the reference object axis based on the set of reference object axis pixel coordinate points and the set of rolling mill axis pixel coordinate points. The calculation module is used to calculate the fitting equation of the centerline of the rolled piece to obtain the bending metric, and to perform coordinate transformation on the set of pixel coordinate points of the reference axis to obtain the planar dimension coordinates. The chord height of the rolled piece is calculated based on the planar dimension coordinates and the bending metric. The detection result generation module is used to generate the bending state detection result of the rolling mill exiting the steel based on the fitting equation of the centerline of the rolled piece, the fitting equation of the centerline of the reference object, the bending measure, and the chord height of the rolled piece; The process of generating a reference object image based on each of the single-frame images includes: using an image similarity calculation algorithm and based on a preset reference object image template to calculate the reference object pixel coordinates of each single-frame image to generate a reference object image. The bending metric is obtained by calculating the fitting equation of the centerline of the rolled piece, including: sliding through all pixels on the fitting equation of the centerline of the rolled piece to calculate the tangent point; generating the tangent point equation based on the tangent point; calculating the angle between the tangent point equation and the fitting equation of the centerline of the reference object; and using the angle as the bending metric.

7. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the bending state detection method for steel output from a rolling mill as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, Used to store a computer program; wherein, when the computer program is executed by a processor, it implements the bending state detection method for steel output from a rolling mill as described in any one of claims 1 to 5.

Images