Centering method, system and equipment for steel coil and saddle

Abstract

The invention relates to the field of intelligent positioning or fixing of workpieces, in particular to a centering method, system and equipment for a steel coil and a saddle. A plurality of field images are obtained, wherein the field images comprise steel coils and saddles; extracting a target steel coil mask from the field image; extracting a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set comprises a plurality of steel coil coordinate points; morphological analysis is conducted on the steel coil coordinate points to obtain longitudinal height information of the steel coil and a boundary coordinate point set located at the bottom of the steel coil; determining a target offset according to the longitudinal height information and the boundary coordinate point set; and the saddle and the steel coil are controlled to be centered according to the target offset. Accurate positioning and centering of the steel coil in the transportation and machining process are achieved. According to the method, the centering accuracy can be greatly improved, the production efficiency is improved, the experience dependence degree of manual operation is reduced, and the human error risk is reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent positioning or fixing of workpieces, and in particular to a method, system and equipment for centering a steel coil and a saddle. Background Art

[0002] The transportation route based on the walking beam and saddle is a commonly used logistics solution for the steel coil entrance section. After the steel coil is transferred from the warehouse area to the saddle by the crane, it becomes in motion. The falling position of the steel coil affects the horizontal stability of the steel coil during subsequent transportation. If the position deviates too much from the center line of the saddle, the steel coil will tilt or even collapse, causing a safety accident. It is very necessary to estimate the position of the steel coil after unloading and ensure that the position of the steel coil after unloading is within a reasonable range of the center line of the saddle before the crane unloads the coil. In addition, on-site protective measures make it very inconvenient to perform contact measurement of the steel coil size.

[0003] In traditional steel mills, when aligning steel coils, instructors at a distance usually assess the alignment near the saddle and assist operators in dropping the steel coils. The communication efficiency between the two parties on the real-time situation on site is low, and the instructors’ estimation of the steel coil drop position is subject to subjective error risks, with low accuracy, low assessment speed and efficiency, and can only provide vague adjustment feedback to the operators. In addition, on-site instructors may also face safety risks due to steel coil production accidents, fatigue due to difficult on-site conditions, and reduced efficiency. When steel coils are frequently added and production is fast, there is also the problem of insufficient manpower.

[0004] Based on this, an embodiment of the present invention is proposed. Summary of the invention

[0005] In order to overcome the above technical problems, the present application provides a method, system and equipment for centering a steel coil and a saddle.

[0006] According to a first aspect of the present application, a method for centering a steel coil and a saddle is provided, comprising: acquiring multiple on-site images, wherein the on-site images include a steel coil and a saddle; extracting a target steel coil mask from the on-site images; extracting a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes multiple steel coil coordinate points; performing morphological analysis on the steel coil coordinate points to obtain longitudinal height information of the steel coil, and a boundary coordinate point set located at the bottom of the steel coil; determining a target offset based on the longitudinal height information and the boundary coordinate point set; and controlling the centering of the saddle and the steel coil based on the target offset.

[0007] In a further scheme of the present application, the method of extracting the target steel coil mask from the on-site image includes: extracting the original steel coil mask from the on-site picture; inputting the original steel coil mask to obtain the connected domain of the steel coil and filling the void area of ​​the connected domain of the steel coil; filtering out the boundary burrs of the connected domain of the steel coil by Gaussian smoothing; calculating the area of ​​the connected domain of the steel coil and obtaining the main body mask of the steel coil by extracting the contour of the connected domain with the largest area; performing convex hull boundary calculation on the main body mask of the steel coil and filling it to obtain a smooth steel coil mask, and then obtaining the target steel coil mask through a defect filtering algorithm.

[0008] In a further solution of the present application, the extraction of the steel coil contour point set corresponding to the target steel coil mask includes: finding the steel coil contour according to the target steel coil mask; traversingly searching for the coordinates of the pixel points of the steel coil contour and saving them as a steel coil contour point set.

[0009] In a further scheme of the present application, the morphological analysis of the steel coil coordinate points to obtain the longitudinal height information of the steel coil includes: sorting the steel coil coordinate points according to the size of the longitudinal coordinate values; obtaining extreme value coordinates among the sorted steel coil coordinate points; and calculating the height difference from the extreme value coordinates to the saddle as the longitudinal height information.

[0010] In a further solution of the present application, the extracting of the boundary coordinate point set located at the bottom of the steel coil includes: extracting straight lines from the steel coil coordinate points; extracting the steel coil coordinate points located on the straight line at the bottom of the steel coil as the boundary coordinate point set.

[0011] In a further scheme of the present application, the target offset is determined based on the longitudinal height information and the boundary coordinate point set, including: obtaining the height difference between the boundary coordinate point in the boundary coordinate point set and the extreme value coordinate; mapping the corresponding weight factor according to the height difference; multiplying the horizontal coordinate of the boundary coordinate point in the boundary coordinate point set by the corresponding weight factor to obtain the contribution value of each boundary coordinate point to the lateral position; summing and averaging the contribution values ​​of all pixel points to obtain the final lateral center of the steel coil; calculating the difference between the lateral center of the steel coil and the preset lateral center of the saddle to obtain the target offset.

[0012] In a further solution of the present application, the centering method also includes: obtaining a first distance from the extreme value coordinate in the boundary coordinate point set to other boundary coordinate points; and extracting the largest value in the first distance as the width information of the steel coil.

[0013] In a further solution of the present application, the centering method also includes: obtaining a first distance from the extreme value coordinate in the boundary coordinate point set to other boundary coordinate points; and extracting the largest value in the first distance as the width information of the steel coil.

[0014] In a further solution of the present application, the centering method further includes: when the width information is less than a preset width threshold, stopping the centering and prompting an abnormality.

[0015] According to a second aspect of the present application, there is provided a steel coil and saddle centering system, comprising: an image acquisition module for acquiring multiple on-site images, wherein the on-site images include steel coils and saddles; a first extraction module for extracting a target steel coil mask from the on-site images; a second extraction module for extracting a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes multiple steel coil coordinate points; an acquisition module for performing morphological analysis on the steel coil coordinate points to acquire longitudinal height information of the steel coil and a boundary coordinate point set located at the bottom of the steel coil; a determination module for determining a target offset based on the longitudinal height information and the boundary coordinate point set; and a control module for controlling the alignment of the saddle and the steel coil based on the target offset.

[0016] Finally, the present application also provides a centering device for a steel coil and a saddle, comprising: an image acquisition device, for a user to collect on-site photos; a driving device, connected to the saddle, for adjusting the position of the saddle; and a controller, which is configured to execute the above-mentioned centering method.

[0017] In summary, based on a general inventive concept of an embodiment of the present invention, i.e., a method for centering a steel coil and a saddle, by automatically extracting information such as the target steel coil mask, contour point set, and boundary coordinate point set, and performing morphological analysis and calculation, it is possible to quickly and accurately measure the position and size of the steel coil, use the longitudinal height information and the boundary coordinate point set to determine the target offset, and control the movement of the saddle according to the target offset, so as to achieve precise positioning and centering of the steel coil during transportation and processing. The above method can greatly improve the accuracy of centering, improve production efficiency, reduce the dependence on experience of manual operation, and reduce the risk of human error. Compared with traditional manual operation, automated processing and analysis can complete tasks more quickly and accurately, saving time and labor costs.

[0018] Other features and advantages of the embodiments of the present invention will be described in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flow chart of a method for centering a steel coil and a saddle provided by an embodiment of the present invention;

[0021] Figure 2 A logical topological diagram of a method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0022] Figure 3 A schematic diagram of a scene picture provided by an embodiment of the present invention;

[0023] Figure 4 A specific flow chart of step S30 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0024] Figure 5 A schematic diagram of an original mask of a steel coil provided in an embodiment of the present invention;

[0025] Figure 6 A schematic diagram of a smooth steel coil profile provided by an embodiment of the present invention;

[0026] Figure 7 A specific flow chart of step S30 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0027] Figure 8 A schematic diagram of a steel coil profile provided by an embodiment of the present invention;

[0028] Fig. 9 A specific flow chart of step S40 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0029] Fig.10 Another specific flow chart of step S40 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0030] Fig.11 A schematic diagram of a straight line at the bottom of a steel coil provided by an embodiment of the present invention;

[0031] Fig.12 A specific flow chart of step S50 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0032] Fig.13 A schematic diagram of the longitudinal height information and target offset provided by an embodiment of the present invention;

[0033] Fig.14 A specific flow chart of step S60 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention;

[0034] Fig.15 A schematic diagram of a module of a steel coil and saddle centering system provided by an embodiment of the present invention;

[0035] Fig.16A hardware schematic diagram of a steel coil and saddle centering device provided by an embodiment of the present invention; and

[0036] Fig.17 A schematic diagram of a module of an electronic device provided by an embodiment of the present invention.

[0037] Marking Description

[0038] 100. Centering system;

[0039] 10. Image acquisition module; 20. First extraction module;

[0040] 30. Second extraction module; 40. Acquisition module;

[0041] 50. Determination module; 60. Control module;

[0042] 200. Centering equipment;

[0043] 21. Image acquisition device; 22. Driving device;

[0044] 23. Controller;

[0045] 400. Electronic equipment;

[0046] 401. Processor; 402. Memory. DETAILED DESCRIPTION

[0047] In order to make the above and other features and advantages of the present application more clear, the present application is further described below in conjunction with the accompanying drawings. It should be understood that the specific embodiments given herein are for the purpose of explaining to those skilled in the art and are only exemplary and not restrictive.

[0048] In the following description, many specific details are set forth to provide a thorough understanding of the present application. However, it is apparent to those skilled in the art that specific details need not be adopted to practice the present application. In other cases, well-known steps or operations are not described in detail to avoid blurring the present application.

[0049] The method for centering a steel coil and a saddle provided in the embodiment of the present application can be performed by the data detection device provided in the embodiment of the present application, and the device can be configured in an electronic device.

[0050] refer to Figure 1 , Figure 1 A flow chart of a method for centering a steel coil and a saddle provided in an embodiment of the present invention includes:

[0051] Step S10, acquiring a plurality of on-site images, wherein the on-site images include the steel coil and the saddle;

[0052] Step S20, extracting the target steel coil mask from the scene image;

[0053] Step S30, extracting a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes a plurality of steel coil coordinate points;

[0054] Step S40, performing morphological analysis on the steel coil coordinate points to obtain the longitudinal height information of the steel coil and a set of boundary coordinate points located at the bottom of the steel coil;

[0055] Step S50, determining the target offset according to the longitudinal height information and the boundary coordinate point set;

[0056] Step S60: Control the movement of the saddle to align with the steel coil according to the target offset.

[0057] In the present application, the acquisition of multiple on-site images in step S10 may be performed by stepwise acquisition or by time sequence acquisition, for example, when the steel coil is hoisted to a certain position, the on-site image acquisition begins.

[0058] The purpose of extracting the mask of the target steel coil in the scene image in step S20 is to cover or ignore the area in the image except the steel coil, and only retain the pixel information of the steel coil. By extracting the mask, the complexity of subsequent processing can be reduced, and the attention can be focused on the steel coil, so as to more accurately perform analysis and measurement related to the steel coil. In the optional scheme of the present application, the following methods can be used to extract the mask of the target steel coil: at least one of threshold segmentation, edge detection algorithm or morphological operation.

[0059] In step S30, the boundary contour point set of the steel coil can be obtained by detecting the edge of the steel coil or using an image segmentation algorithm, where the steel coil contour point set includes multiple steel coil coordinate points, and each steel coil coordinate point can be used to feedback the position of the steel coil in the image, including the horizontal coordinate and the vertical coordinate, which can help determine the position relationship of the steel coil relative to the saddle.

[0060] It should be noted that the "steel coil contour point set" refers to all the steel coil coordinate points at the boundary contour of the steel coil, the "longitudinal height information" refers to the longitudinal distance of the steel coil from the saddle on the image, that is, it represents the height of the current steel coil from the saddle, and the "boundary coordinate point set" refers to the boundary coordinate point set located at the bottom part of the steel coil in the steel coil contour point set.

[0061] Step S40 performs morphological analysis on the steel coil coordinate points to obtain the longitudinal height information of the steel coil and identify the boundary coordinate point set at the bottom of the steel coil. The longitudinal height information of the steel coil can be obtained by calculating the height of the steel coil profile or using morphological operations. At the same time, the boundary coordinate points at the bottom of the steel coil are identified for subsequent calculation of the offset. In step S50, the offset is calculated by combining the longitudinal height information of the steel coil and the boundary coordinate points at the bottom to determine the positional relationship of the steel coil relative to the saddle. In step S60, the position of the saddle is adjusted according to the calculated offset so that the steel coil is in the desired center position.

[0062] Based on a general inventive concept of an embodiment of the present invention, i.e., a method for centering a steel coil and a saddle, by automatically extracting information such as the target steel coil mask, contour point set, and boundary coordinate point set, and performing morphological analysis and calculation, it is possible to quickly and accurately measure the position and size of the steel coil, use the longitudinal height information and the boundary coordinate point set to determine the target offset, and control the movement of the saddle according to the target offset to achieve precise positioning and centering of the steel coil during transportation and processing. The above method can greatly improve the accuracy of centering, improve production efficiency, reduce the dependence on experience of manual operation, and reduce the risk of human error. Compared with traditional manual operation, automated processing and analysis can complete tasks more quickly and accurately, saving time and labor costs.

[0063] The present invention is described below by a specific embodiment:

[0064] In the embodiment of the present invention, a comprehensive steel coil feature extraction scheme based on deep learning is adopted, and the EfficientNetV3 model is used as the neural network backbone to collect multiple on-site images (including steel coil hoisting, falling, seating, and removal) from each saddle spectator position on site. The data annotation scheme adopts the method of annotating all pixels belonging to the steel coil in the on-site image. After multiple rounds of training, the model can correctly extract the target steel coil mask belonging to the steel coil part in the on-site image, and adopt contour filling and convex hull smoothing methods to make the extracted target steel coil mask stable and smooth, thereby improving the accuracy rate; then, the bottom position, contour center, bottom straight line and other robust features are extracted from the target steel coil mask as the basis for the longitudinal height information, centering position and width information of the steel coil, and finally the target offset of the control is obtained, and then the target offset is converted into the control amount, and the automation and steel coil centering process is realized by automatically moving the saddle. Among them, each step from image to mask to contour and then to position and size features is an optimizable process, and the design purpose of each link is to transmit stable intermediate results downstream.

[0065] See also Figure 2 , Figure 2The logical topology diagram of the method for centering a steel coil and a saddle provided in an embodiment of the present invention is provided to illustrate the general inventive concept of the embodiment of the present invention. In the figure, the rounded dotted rectangular container represents the input and output of the deep learning model or post-processing process, the rounded solid rectangular container represents the intermediate result of the algorithm, and the rectangular solid container represents the processing method between the intermediate results or between the intermediate results and the input and output.

[0066] Step S10, obtaining multiple on-site images;

[0067] Specifically, an industrial camera is triggered to collect on-site pictures in a step-by-step process. When the crane pulls the steel coil to a fixed position above the saddle, a on-site picture is collected, such as Figure 3 The acquisition angle can be in multiple directions to avoid the influence of noise from the external environment to the greatest extent.

[0068] Combination Figure 2 Also see Figure 4 , Figure 4 The specific flow chart of step S30 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. Step S30 extracts the target steel coil mask from the scene image, including:

[0069] Step S21, extracting the original mask of the steel coil from the on-site picture;

[0070] Step S22, inputting the original mask of the steel coil to obtain the connected domain of the steel coil and filling the empty area of ​​the connected domain of the steel coil;

[0071] Step S23, removing the boundary burrs of the connected domain of the steel coil by Gaussian smoothing;

[0072] Step S24, calculating the area of ​​the connected domain of the steel coil and extracting the contour of the connected domain with the largest area to obtain the main body mask of the steel coil;

[0073] Step S25, performing convex hull boundary calculation and filling on the main body mask of the steel coil to obtain a smooth steel coil mask, and then obtaining a target steel coil mask through a defect filtering algorithm.

[0074] Specifically, in step S31, for the input scene picture, the original mask of the steel coil representing the position and shape of the steel coil is extracted through threshold segmentation, edge detection and connected domain analysis. The original mask of the steel coil is a binary image, in which the steel coil area is marked as the foreground (white) and the non-steel coil area is marked as the background (black). Figure 5 shown.

[0075] The steel coil is usually composed of connected domains, that is, the area formed by connecting adjacent foreground pixels. In step S32, the connected domain analysis is performed on the original mask of the steel coil, each connected domain of the steel coil is identified, and the hole areas in the mask are filled. Filling the holes can ensure the integrity of the connected domain of the steel coil, avoid errors in subsequent processing, and improve accuracy.

[0076] Due to the morphological characteristics of the steel coil, there may be burrs or irregular parts on the boundaries of the connected domain, which will interfere with subsequent calculations and analysis. By applying image filtering algorithms such as Gaussian smoothing, the boundaries of the connected domain of the steel coil can be smoothed, burrs can be removed, and the steel coil mask can be made clearer and more accurate. Then the area of ​​each connected domain of the steel coil is calculated, and the connected domain with the largest area is selected as the main body of the steel coil. The contour of the connected domain is extracted and the corresponding mask, namely the steel coil main body mask, is generated. Then, a convex hull boundary operation is performed on the steel coil main body mask to obtain a smoother steel coil contour. Figure 6 shown.

[0077] Then, the obtained smooth coil mask is filled to ensure that there are no holes inside the mask. Finally, the smooth coil mask is processed by the defect filtering algorithm to obtain the final target coil mask.

[0078] In general, filling the hollow areas of the connected domain of the steel coil can help eliminate the holes inside the steel coil, making the mask more complete and accurate. Filling the hollow areas of the connected domain of the steel coil can make the mask more complete and accurate. Filling the hollow areas of the connected domain of the steel coil can help eliminate the holes inside the steel coil, making the mask more complete and accurate. By performing convex hull boundary filling on the steel coil main body mask, a smooth steel coil mask can be obtained. The smooth steel coil mask is neater and continuous, which is helpful for subsequent defect filtering and analysis. Through the defect filtering algorithm, defects that do not meet the conditions can be eliminated to obtain a clean and accurate target steel coil mask, which can improve the accuracy of steel coil quality assessment and optimization.

[0079] Combination Figure 2 Also see Figure 7 , Figure 7 The specific flow chart of step S30 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. In step S30, a steel coil contour point set corresponding to the target steel coil mask is extracted, including:

[0080] Step S31, searching for the steel coil profile according to the target steel coil mask;

[0081] Step S32, traversally searching for the pixel coordinates of the steel coil contour and saving them as a steel coil contour point set.

[0082] Specifically, an image processing algorithm, such as edge detection or threshold segmentation algorithm, is used to find pixel points on the boundary of the steel coil, and a connected domain analysis is performed on these pixel points. Adjacent pixel points are classified into the same connected domain. According to the shape and size of the connected domain, the connected domain at the non-steel coil boundary contour is filtered out, and the connected domain of the target steel coil is retained to obtain the target steel coil boundary contour, such as Figure 8 The coordinates of these pixel points of the target steel coil boundary contour are extracted and saved as a point set, that is, the point set extracted from the mask is obtained.

[0083] It can be understood that the target steel coil mask is a two-dimensional pixel display of the steel coil area on the entire scene image. For example, the target steel coil mask as the foreground can be 1 pixel, and the other backgrounds can be 0, where each pixel has a position, which can be represented by row and column coordinates. Generally speaking, the upper left corner of the scene image is the origin (0,0), the row index increases from top to bottom, and the column index increases from left to right. First, we use the mask to extract the steel coil contour at the boundary, that is, it represents the boundary of the steel coil, and then traverse each pixel point in the entire scene image. When encountering a pixel point on the steel coil contour, its coordinates are saved in the point set. Finally, all the point coordinates on the steel coil contour are saved in order, which is the steel coil contour point set mentioned above. The steel coil contour point set contains multiple steel coil coordinate points, which actually refers to the coordinate points of the steel coil boundary contour.

[0084] Combination Figure 2 Also see Fig. 9 , Fig. 9 The specific flow chart of step S40 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. In step S40, morphological analysis is performed on the steel coil coordinate points to obtain the longitudinal height information of the steel coil, including:

[0085] Step S41, sorting the steel coil coordinate points according to the size of the longitudinal coordinate values;

[0086] Step S42, obtaining extreme value coordinates from the traversed and sorted steel coil coordinate points;

[0087] Step S43, calculating the height difference from the extreme value coordinate to the saddle as longitudinal height information.

[0088] Specifically, each steel coil coordinate point is a tuple in the form of (x, y). It is sorted according to the y longitudinal coordinate value, the lowest extreme coordinate is obtained, and then the height difference from the extreme coordinate to the saddle is calculated to obtain the longitudinal height information.

[0089] In the embodiment of the present invention, since the height of the saddle surface does not change, the height reference value of the saddle surface can be initially calibrated. It is understandable that the longitudinal height information currently only feeds back the distance between the pixels on the photo.

[0090] Combination Figure 2 Also see Fig.10 , Fig.10 Another specific flow chart of step S40 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. In step S40, a set of boundary coordinate points located at the bottom of the steel coil is extracted, including:

[0091] Step S44, extracting straight lines from the steel coil coordinate points;

[0092] Step S45, extracting the steel coil coordinate points that are on the same straight line as the extreme value coordinates as the boundary coordinate point set.

[0093] It can be understood that in step 44, the least squares method can be optionally used for straight line fitting to obtain all the straight lines that can exist in the steel coil coordinate points. Assume that the steel coil coordinate points are a list of point sets to be processed. First, a straight line function model is defined, where a and b are the parameters to be fitted. Then, the steel coil coordinate points are converted into arrays x and y. Next, the least squares method straight line fitting is performed by calling the function, and the returned fitting parameters are stored. Finally, the fitting parameters a and b are extracted, which are the slope and intercept of the fitted straight line.

[0094] Furthermore, since the intercept of the straight line is used to describe the position of the straight line in the y-axis direction, the straight line representing the bottom of the steel coil can be obtained by simply sorting the intercepts or judging the threshold, such as Fig.11 As shown, all the steel coil coordinate points located on the straight line are added to the boundary coordinate point set.

[0095] Furthermore, whether the steel coil coordinate point is on the straight line at the bottom of the steel coil can be determined by the distance from the steel coil coordinate point to the straight line. Assuming the point to be determined is (x1, y1), by judging the distance from the point to the straight line, if it is less than the preset straight line threshold, it proves that the point is on the straight line, otherwise the return point is not on the straight line.

[0096] Combination Figure 2 Also see Fig.12 , Fig.12 The specific flow chart of step S50 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. In step S50, the target offset is determined according to the longitudinal height information and the boundary coordinate point set, including:

[0097] Step S51, obtaining the height difference between the boundary coordinate point and the extreme value coordinate in the boundary coordinate point set;

[0098] Step S52, mapping the corresponding weight factor according to the height difference;

[0099] Step S53, multiplying the horizontal coordinate of the boundary coordinate point in the boundary coordinate point set by the corresponding weight factor to obtain a contribution value of each boundary coordinate point to the horizontal position;

[0100] Step S54, summing and averaging the contribution values ​​of all pixels to obtain the final transverse center of the steel coil;

[0101] Step S55, calculating the difference between the transverse center of the steel coil and the preset transverse center of the saddle to obtain a target offset.

[0102] It can be understood that in the collected on-site images, due to the field of view of the industrial camera, the straight line at the bottom of the steel coil is unlikely to appear horizontal in the image, so the higher edge will have a larger offset in the horizontal direction of the image, while the lower edge will have a smaller offset. This is because the different field of view angles lead to the near-large-far-small effect under the camera's perspective, which makes a certain correlation between the edge height and the actual lateral position. In the boundary coordinate point set, according to the different heights, the estimated lateral position of the steel coil is calculated by weighted average.

[0103] Specifically, the boundary coordinate points with larger height differences relative to the extreme value coordinates will be given greater weights, thereby better reflecting the contribution of the edge height changes of the steel coil to the lateral position. The boundary coordinate points with smaller height differences have smaller weights. By calculating the lateral center of the steel coil in this weighted average manner, better and more accurate calculation results can be obtained.

[0104] Then, the calculated horizontal difference between the lateral center of the steel coil and the lateral center of the saddle is the current lateral offset of the steel coil.

[0105] Perform industrial camera calibration to obtain the camera's internal parameter matrix and external parameter matrix. Camera calibration can be performed using a checkerboard pattern, and the calibration results include information such as the camera's focal length, principal point, and distortion parameters. Based on the camera's internal and external parameters, the camera projection matrix can be obtained. This converts the vertical height information and target offset in the image into actual data, such as Fig.13 As shown, the crane is controlled to move according to the offset and released according to the calculated height to achieve automatic centering and automatic lowering with the saddle, thereby improving work efficiency and accuracy.

[0106] See also Fig.14 , Fig.14 The specific flow chart of step S60 in the method for centering a steel coil and a saddle provided in an embodiment of the present invention. The centering method further includes:

[0107] Step S61, obtaining a first distance from an extreme value coordinate in a boundary coordinate point set to other boundary coordinate points;

[0108] Step S62, extracting the maximum value in the first distance as the width information of the steel coil;

[0109] Step S63: When the width information is less than the preset width threshold, stop the centering and prompt an abnormality.

[0110] From the above, it can be seen that the boundary coordinate point set represents the coordinates of the pixel points on the straight line at the bottom of the steel coil. The first distance from the extreme value coordinate to the boundary coordinate point in the boundary coordinate point set is calculated, and the largest value is the length of the straight line at the bottom of the steel coil. This length can be used to feedback the width value of the steel coil. When the width is less than the width threshold, it proves that the steel coil has been seriously tilted in its radial direction. At this time, the centering program is stopped to prompt an abnormality, so as to prompt the operator to readjust the crane parameters and lifting position.

[0111] “Other boundary coordinate points” refer to boundary coordinate points other than extreme value coordinates in the entire boundary coordinate point set.

[0112] Fig.15 A schematic diagram of a module of a steel coil and a saddle centering system 100 provided in an embodiment of the present invention. A steel coil and a saddle centering system 100 includes:

[0113] An image acquisition module 10 acquires a plurality of on-site images, wherein the on-site images include the steel coil and the saddle;

[0114] A first extraction module 20 extracts a target steel coil mask from the scene image;

[0115] A second extraction module 30 extracts a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes a plurality of steel coil coordinate points;

[0116] An acquisition module 40 performs morphological analysis on the steel coil coordinate points to obtain the longitudinal height information of the steel coil and a set of boundary coordinate points located at the bottom of the steel coil;

[0117] A determination module 50 determines a target offset according to the longitudinal height information and the boundary coordinate point set;

[0118] The control module 60 controls the movement of the saddle to align with the steel coil according to the target offset.

[0119] Fig.16 A hardware schematic diagram of a steel coil and saddle centering device 200 provided in an embodiment of the present invention. A steel coil and saddle centering device 200 includes:

[0120] Image acquisition device 21, the user collects photos of the scene;

[0121] A driving device 22 connected to the saddle and used for adjusting the position of the saddle;

[0122] The controller 23 is configured to execute the above-mentioned centering method.

[0123] It should be understood that the specific features, operations and details described hereinabove with respect to the method of the present application may also be similarly applied to the device and system of the present application, or vice versa. In addition, each step of the method of the present application described above may be performed by a corresponding component or unit of the device or system of the present application.

[0124] It should be understood that each module / unit of the device of the present application can be implemented in whole or in part by software, hardware, firmware or a combination thereof. Each module / unit can be embedded in the processor of the electronic device or independent of the processor in the form of hardware or firmware, or can be stored in the memory of the electronic device in the form of software for the processor to call to perform 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.

[0125] like Fig.17 As shown, the present application also provides an electronic device 400, which includes a processor 401 and a memory 402 storing computer program instructions. When the processor 401 executes the computer program instructions, each step of the above-mentioned method for centering the steel coil and the saddle is implemented. The electronic device 400 can be a server, a terminal, or any other electronic device with necessary computing and / or processing capabilities.

[0126] In one embodiment, the electronic device 400 may include a processor, a memory, a network interface, a communication interface, etc. connected through a system bus. The processor of the electronic device 400 may be used to provide necessary computing, processing and / or control capabilities. The memory of the electronic device 400 may include a non-volatile storage medium and an internal memory. The non-volatile storage medium may store an operating system, a computer program, etc. The internal memory may provide an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the electronic device 400 may be used to connect and communicate with external devices through a network. When the computer program is executed by the processor, the steps of the method of the present application are executed.

[0127] The present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the above-mentioned method for centering a steel coil and a saddle is implemented.

[0128] Those skilled in the art will appreciate that the method steps of the present application can be completed by instructing related hardware such as the electronic device 400 or a processor through a computer program, and the computer program can be stored in a non-temporary computer-readable storage medium, and the computer program causes the steps of the present application to be executed when it is executed. Depending on the circumstances, any reference to memory, storage, or other media herein 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 disk, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0129] The various technical features described above can be combined arbitrarily. Although all possible combinations of these technical features are not described, any combination of these technical features should be considered to be covered by this specification as long as there is no contradiction in such combination.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for centering a steel coil and a saddle, characterized in that: include: Acquire a plurality of on-site images, wherein the on-site images include the steel coil and the saddle; extracting a target steel coil mask from the scene image; Extracting a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes a plurality of steel coil coordinate points; Performing morphological analysis on the steel coil coordinate points to obtain longitudinal height information of the steel coil and a set of boundary coordinate points located at the bottom of the steel coil; Determine the target offset according to the longitudinal height information and the boundary coordinate point set; The saddle and the steel coil are aligned according to the target offset.

2. The centering method according to claim 1, characterized in that: The step of extracting the target steel coil mask from the on-site image comprises: Extracting the original mask of the steel coil from the on-site picture; Input the original mask of the steel coil to obtain the connected domain of the steel coil and fill the empty area of ​​the connected domain of the steel coil; The boundary burrs of the connected domain of the steel coil are filtered out by Gaussian smoothing; Calculate the area of ​​the connected domain of the steel coil and obtain the main body mask of the steel coil by extracting the contour of the connected domain with the largest area; The convex hull boundary of the steel coil main body mask is calculated and filled to obtain a smooth steel coil mask, and then the target steel coil mask is obtained through the defect filtering algorithm.

3. The centering method according to claim 1, characterized in that: The step of extracting a steel coil contour point set corresponding to the target steel coil mask includes: Find the steel coil outline according to the target steel coil mask; The coordinates of the pixel points of the steel coil contour are searched traversally and saved as a steel coil contour point set.

4. The centering method according to claim 1, characterized in that: The step of performing morphological analysis on the steel coil coordinate points to obtain longitudinal height information of the steel coil includes: Sorting the steel coil coordinate points according to the size of the longitudinal coordinate values; Obtain extreme value coordinates from the traversed and sorted steel coil coordinate points; The height difference from the extreme value coordinate to the saddle is calculated as longitudinal height information.

5. The centering method according to claim 4, characterized in that: The step of extracting a boundary coordinate point set located at the bottom of the steel coil includes: Extracting straight lines from the steel coil coordinate points; The steel coil coordinate points located on the straight line at the bottom of the steel coil are extracted as the boundary coordinate point set.

6. The centering method according to claim 5, characterized in that: The determining of the target offset according to the longitudinal height information and the boundary coordinate point set includes: Obtaining a height difference between a boundary coordinate point in the boundary coordinate point set and the extreme value coordinate; Mapping a corresponding weight factor according to the height difference; Multiplying the horizontal coordinate of the boundary coordinate point in the boundary coordinate point set by the corresponding weight factor to obtain a contribution value of each boundary coordinate point to the horizontal position; The contribution values ​​of all pixels are summed and averaged to obtain the final transverse center of the steel coil; The difference between the transverse center of the steel coil and the preset transverse center of the saddle is calculated to obtain the target offset.

7. The centering method according to claim 5, characterized in that: The centering method further comprises: Obtain a first distance from the extreme value coordinate in the boundary coordinate point set to other boundary coordinate points; The maximum value in the first distance is extracted as the width information of the steel coil.

8. The centering method according to claim 7, characterized in that: The centering method further comprises: When the width information is less than a preset width threshold, the centering is stopped and an abnormality is prompted.

9. A centering system for a steel coil and a saddle, characterized in that: include: An image acquisition module is used to acquire a plurality of on-site images, wherein the on-site images include the steel coil and the saddle; A first extraction module extracts a target steel coil mask from the scene image; A second extraction module extracts a steel coil contour point set corresponding to the target steel coil mask, wherein the steel coil contour point set includes a plurality of steel coil coordinate points; an acquisition module, performing morphological analysis on the steel coil coordinate points to acquire longitudinal height information of the steel coil and a set of boundary coordinate points located at the bottom of the steel coil; A determination module determines a target offset according to the longitudinal height information and the boundary coordinate point set; and a control module controls the alignment of the saddle and the steel coil according to the target offset.

10. A centering device for a steel coil and a saddle, characterized in that: include: Image acquisition device, where users collect photos of the scene; A driving device connected to the saddle and used for adjusting the position of the saddle; A controller, wherein the controller is configured to execute the centering method according to any one of claims 1 to 8.

Images