Method, device, system and storage medium for detecting coil loose defects of a steel coil

Abstract

This invention provides a method, device, system, and storage medium for detecting loose coil defects in steel coils. The detection method includes: acquiring a steel coil point cloud model from a steel coil point cloud set and a steel coil region from a steel coil image set; extracting edges from the steel coil region to obtain the steel coil edges; projecting the steel coil edges onto the steel coil point cloud model to establish a correlation between the steel coil edges and the steel coil point cloud model; calculating the distance between corresponding edge points in the steel coil point cloud model; and applying a threshold judgment to the distance to determine whether a loose coil defect exists in the steel coil. This technical solution detects loose coil defects in steel coils by fusing 2D images with 3D LiDAR, which can accurately determine the specific distance value of the loose coil, improve the accuracy of defect judgment, and has higher versatility, applicable to the determination of loose coil defects in steel coils of different specifications.

Description

Technical Field

[0001] This invention belongs to the field of automated processing, and in particular relates to a method, equipment, system and storage medium for detecting loose coil defects in steel coils. Background Technology

[0002] In traditional steel mills, during the production of steel coils, gaps may occur between layers if the coils are not tightly wound. This is known as "loose coiling" and is a common defect in finished steel coil manufacturing.

[0003] Loose coils are difficult to detect using image recognition because current 2D image-based detection solutions all use edge operators to extract the steel coil layers and calculate the distance between gaps in the steel coil layers to determine if they are loose. This approach has the following drawbacks:

[0004] 1. 2D cameras can only acquire 2D information, lacking scale information and unable to capture the true scale of the coil. Therefore, it is necessary to estimate the dimensions in the image (usually based on the scale of the image and the actual steel coil). Typically, the distance between the layers of the steel coil needs to exceed 3cm before rework is required. This distance is relatively low compared to the overall steel coil, which has many layers and is very dense.

[0005] 2. Extracting the gap area requires that the gap between the steel coils be obvious and that the depth information at the gap be significantly different from that of the steel coil area;

[0006] 3. When calculating the distance between steel coil layers in image recognition, the steel coil layers extracted from the edges need to be fitted into a circle for calculation. However, when loosening occurs, the supporting layer is usually very irregular, leading to inaccurate distance calculations. Figure 1 As shown.

[0007] As a result, the current automatic detection method for loose coils is not accurate enough, leading to a decrease in yield. There is an urgent need for a new method for detecting loose coil defects in steel coils. Summary of the Invention

[0008] To address the technical problems identified in the prior art, this invention provides a method, device, system, and storage medium for detecting loose coil defects in steel coils.

[0009] The first aspect of this application provides a method for detecting loose coil defects in steel coils. The method includes: acquiring a steel coil point cloud model from a steel coil point cloud set and a steel coil region from a steel coil image set; performing edge extraction on the steel coil region to obtain steel coil edges; projecting the steel coil edges onto the steel coil point cloud model to establish a correlation between the steel coil edges and the steel coil point cloud model; calculating the distance between corresponding edge points in the steel coil point cloud model; and applying a threshold judgment to the distance to determine whether the steel coil has a loose coil defect.

[0010] In the optional scheme of this application, obtaining the steel coil point cloud model in the steel coil point cloud set and the steel coil region in the steel coil image set includes: obtaining the steel coil point cloud set and the steel coil image set; back-projecting the steel coil point cloud set onto the steel coil image set, and determining the steel coil region in the steel coil image set; and segmenting the steel coil point cloud model in the steel coil point cloud set using a 3D point cloud algorithm.

[0011] In an optional scheme of this application, edge extraction of the steel coil region to obtain the steel coil edge includes: reading the steel coil region; performing grayscale processing on the steel coil region to obtain a grayscale image; and performing edge extraction on the grayscale image using an edge operator to obtain the steel coil edge.

[0012] In the optional scheme of this application, projecting the edge of the steel coil onto the steel coil point cloud model and establishing the association between the edge of the steel coil and the steel coil point cloud model includes: jointly calibrating the image acquisition device and the point cloud acquisition device; finding the corresponding edge points by projecting the edge of the steel coil onto the steel coil point cloud model; assigning RBG values ​​to the edge points; and transforming the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device.

[0013] In an optional scheme of this application, calculating the distance between corresponding edge points in the steel coil point cloud model includes: calculating the coordinates of the center point of the steel coil point cloud model; randomly selecting an edge point and calculating the adjacent edge points that are concentric with the center point coordinates of that edge point; calculating the distance between the edge point and its adjacent edge points, until all edge points are traversed.

[0014] In an optional solution of this application, the threshold judgment of distance to determine whether the steel coil has a loose coil defect includes: extracting edge points whose distance exceeds a preset distance threshold as target point clouds; performing statistics on the target point clouds, and judging that the steel coil has a loose coil defect when the target point cloud exceeds a preset number threshold.

[0015] In a second aspect of this application, a detection device for loose coil defects in steel coils is provided. The detection device includes: an image acquisition device for acquiring a set of steel coil images; a point cloud acquisition device for acquiring a set of steel coil point clouds; and a host computer configured to: segment a steel coil region and a steel coil point cloud model from the steel coil image set and the steel coil point cloud set; extract edges from the steel coil region to obtain steel coil edges; project the steel coil edges onto the steel coil point cloud model to establish a relationship between the steel coil edges and the steel coil point cloud model; calculate the distance between corresponding edge points in the steel coil point cloud model; and perform a threshold judgment on the distance to determine whether the steel coil has a loose coil defect.

[0016] In an optional scheme of this application, projecting the edge of the steel coil onto the steel coil point cloud model and establishing the association between the steel coil edge and the steel coil point cloud model includes: jointly calibrating the image acquisition device and the point cloud acquisition device; finding the corresponding edge points by projecting the steel coil edge onto the steel coil point cloud model; assigning RGB values ​​to the edge points; transforming the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device; and extracting the edge of the steel coil region to obtain the steel coil edge includes: reading the steel coil region; performing grayscale processing on the steel coil region to obtain a grayscale image; and extracting the edge of the grayscale image using an edge operator to obtain the steel coil edge. Calculating the distance between corresponding edge points in the steel coil point cloud model includes: calculating the coordinates of the center point of the steel coil point cloud model; randomly selecting edge points and calculating adjacent edge points concentric with the center point coordinates of the edge point; calculating the distance between the edge point and its adjacent edge points until all edge points are traversed; and performing a threshold judgment on the distance to determine whether the steel coil has a loose coil defect includes: extracting edge points whose distance exceeds a preset distance threshold as target point clouds; and statistically analyzing the target point clouds, and determining that the steel coil has a loose coil defect when the number of target point clouds exceeds a preset number threshold.

[0017] In an optional embodiment of this application, the detection system includes: a segmentation module for acquiring a steel coil point cloud model from a steel coil point cloud set and acquiring a steel coil region from a steel coil image set; an edge extraction module for extracting edges from the steel coil region to obtain steel coil edges; a projection module for projecting the steel coil edges onto the steel coil point cloud model to establish a relationship between the steel coil edges and the steel coil point cloud model; a distance calculation module for calculating the distance between corresponding edge points in the steel coil point cloud model; and a judgment module for performing a threshold judgment on the distance to determine whether the steel coil has a loose coil defect.

[0018] Finally, this application also provides a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the detection method described above.

[0019] Beneficial effects:

[0020] The method for detecting loose coil defects in steel coils provided in this invention fuses image recognition and point cloud acquisition. It acquires a steel coil point cloud model from a steel coil point cloud dataset and a steel coil region from a steel coil image dataset. Edge extraction is performed within the steel coil region to obtain the steel coil edges. These edges are then projected onto the steel coil point cloud model to extract corresponding edge points. The distance between these edge points in the point cloud model is calculated, and a threshold is applied to determine the presence of a steel coil defect. This technical solution, by fusing 2D images with 3D LiDAR, accurately determines the specific distance of the loose coil, improving the accuracy of defect detection. It also has greater versatility and can be applied to the determination of loose coil defects in steel coils of different specifications.

[0021] Other features and advantages of the embodiments of the present invention will be described in the following detailed description section. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 A schematic diagram illustrating a common loose coil defect provided in this application;

[0024] Figure 2 A flowchart illustrating the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0025] Figure 3 middle Figure 3 (a) is the corresponding point cloud model of the steel coil. Figure 3 (b) is a simplified schematic diagram of the steel coil edge after edge extraction. Figure 3 (c) is a simplified schematic diagram of the edge points in the corresponding steel coil point cloud model;

[0026] Figure 4 This is a flowchart of step S1 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0027] Figure 5 This is a flowchart of step S2 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0028] Figure 6 This is a flowchart of step S3 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0029] Figure 7 This is a flowchart of step S4 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0030] Figure 8 This is a flowchart of step S5 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0031] Figure 9 This is a connection topology diagram of the steel coil loose coil defect detection device provided in an embodiment of the present invention;

[0032] Figure 10 This is a schematic diagram of a module for detecting loose coil defects in steel coils, provided in an embodiment of the present invention.

[0033] Figure Labels

[0034] 100. Detection equipment; 101. Image acquisition device;

[0035] 102. Point cloud acquisition device; 103. Host computer;

[0036] 200. Detection system; 201. Segmentation module;

[0037] 202. Edge extraction module; 203. Projection module;

[0038] 204. Distance calculation module; 205. Judgment module. Detailed Implementation

[0039] To make the above and other features and advantages of the present invention clearer, the invention will be further described below with reference to the accompanying drawings. It should be understood that the specific embodiments given herein are for the purpose of explanation to those skilled in the art and are exemplary only, not restrictive.

[0040] Continuing from the previous point, in order to solve the technical problems existing in the prior art, the method of judging whether there is loose coiling defect in steel coils is only based on edge extraction technology and distance calculation in image recognition. However, due to the lack of scale information in 2D images, the accuracy is poor when measuring distances. The present invention provides a general inventive concept, namely, a method for detecting loose coiling defects in steel coils that combines 3D point cloud technology.

[0041]

Detection Method

[0042] Please see Figure 2 , Figure 2 A flowchart illustrating the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0043] This invention provides a method for detecting loose coil defects in steel coils, the method comprising the following steps:

[0044] Step S1: Obtain the steel coil point cloud model from the steel coil point cloud set and the steel coil region from the steel coil image set;

[0045] Step S2: Extract the edges of the steel coil area to obtain the edges of the steel coil;

[0046] Step S3: Project the edge of the steel coil onto the point cloud model of the steel coil to establish the relationship between the edge of the steel coil and the point cloud model of the steel coil;

[0047] Step S4: Calculate the distance between corresponding edge points in the steel coil point cloud model;

[0048] Step S5: Perform a threshold judgment on the distance to determine whether the steel coil has a loose coil defect.

[0049] In this embodiment of the invention, it is necessary to first acquire a set of steel coil images and a set of steel coil point clouds. The set of steel coil images is acquired using an image acquisition device, such as a 2D camera, and the set of steel coil point clouds is acquired using a point cloud acquisition device, such as a lidar.

[0050] It should be noted that, in this embodiment of the invention, the steel coil image set and the steel coil point cloud set are interpreted as point clouds and images containing both background and the steel coil target. Step S1, the target extraction step, involves removing the background from the steel coil point cloud set and the steel coil image set, and then obtaining the steel coil point cloud model from the steel coil point cloud set and the steel coil region from the steel coil image set, thereby reducing computation and improving the accuracy of subsequent calculations.

[0051] In steps S2 to S4, the edge extraction function in the two-dimensional image is first used to extract the edge of each coil of steel. Then, the coil edges are projected onto the coil point cloud model to establish a connection between the coil edges and the point cloud model. This allows for the extraction of edge points corresponding to the coil edges from the point cloud model. After extracting the edge points, the distance between the edge points in the point cloud model is calculated, and a threshold is then used to determine any loose coil defects in the steel coil.

[0052] Please see Figure 3 , Figure 3 (a) is the corresponding point cloud model of the steel coil. Figure 3 (b) is a simplified schematic diagram of the steel coil edge after edge extraction. Figure 3 (c) is a simplified schematic diagram of the edge points in the corresponding steel coil point cloud model.

[0053] It is understandable that since the coil only needs to be determined in the radial direction, the axial direction can be ignored. Only the image set and point cloud set of the coil at the end face need to be extracted. That is, during image recognition, the image acquisition device can directly acquire the image set of the coil at the end face. Similarly, when acquiring the point cloud set of the coil, the point cloud model of the coil is reduced to a plane.

[0054] It should be noted that "coil edge" refers to the lines of the coil layers and the end face of the coil. Figure 3 (b) Demonstrates a few edges of the outermost and innermost coils; other coil edges are simplified in this diagram. Using Figure 3 (b) By back-projecting onto the steel coil point cloud model, edge points corresponding to the edges of the steel coil can be extracted from the steel coil point cloud model, such as... Figure 3 (c) Then, steps S4 to S55 are performed to calculate the distance between the corresponding edge points in the steel coil point cloud model; a threshold judgment is made on the distance to determine whether there is a loose coil defect in the steel coil.

[0055] It is understood that the method for detecting loose coil defects in steel coils provided in this embodiment of the invention fuses image recognition and point cloud acquisition. It obtains a steel coil point cloud model from a steel coil point cloud set and a steel coil region from a steel coil image set. Edge extraction is performed within the steel coil region to obtain the steel coil edges. Then, the steel coil edges are projected onto the steel coil point cloud model to obtain the corresponding edge points. The distance between these edge points in the steel coil point cloud model is calculated, and a threshold is applied to the distance to determine whether a steel coil defect exists. This technical solution detects loose coil defects by fusing 2D images with 3D LiDAR, accurately determining the specific distance value of the loose coil, improving the accuracy of defect detection, and possessing greater versatility, applicable to the determination of loose coil defects in steel coils of different specifications.

[0056] Please see Figure 4 , Figure 4 This is a flowchart of step S1 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0057] In a specific scheme, obtaining the steel coil point cloud model from the steel coil point cloud set and the steel coil region from the steel coil image set includes:

[0058] Step S11: Obtain the point cloud set and image set of the steel coil;

[0059] Step S12: Back-project the point cloud of the steel coil onto the image set of the steel coil, and determine the steel coil region in the image set of the steel coil;

[0060] Step S13: Use a 3D point cloud algorithm to segment the steel coil point cloud model from the steel coil point cloud set.

[0061] In one example, step S11 can use OpenCV to read in a PNG image file, i.e., a set of steel coil images, and use PCL to read in the corresponding PCD point cloud file, i.e., a set of steel coil point clouds.

[0062] In a specific scheme for step S12, the coordinate points are first translated to control the back projection range of the point cloud, and the point cloud in the steel coil point cloud set is back-mapped to the pixel coordinate position of the steel coil image set, so that the steel coil image set is converted into a depth map; then, depth filtering is performed based on the depth information of the steel coil image set to obtain the steel coil region.

[0063] In a specific scheme for step S13, a laser radar scans the steel coil along a preset route to obtain a point cloud set of the steel coil; point cloud features are extracted from the point cloud set based on the characteristics of the steel coil; and a clustering segmentation algorithm is used on the steel coil data to obtain a steel coil point cloud model. For example, each point cloud includes RGB information in addition to three-dimensional coordinates (x, y, z), which can be extracted based on the point cloud color. The steel coil point cloud set is first preprocessed by filtering, downsampling, etc., and then clustered and segmented based on the color of the steel coil to obtain the steel coil point cloud model.

[0064] It can be understood that steps S12 and S13 above involve removing the image background and point cloud background from the steel coil point cloud set and the steel coil image set, respectively, and extracting the target foreground image and target foreground point cloud, respectively. This facilitates subsequent calculations.

[0065] Please see Figure 5 , Figure 5 This is a flowchart of step S2 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0066] Step S2 involves edge extraction of the steel coil area to obtain the steel coil edge, including:

[0067] Step S21: Read the steel coil area;

[0068] Step S22: Perform grayscale processing on the steel coil area to obtain a grayscale image;

[0069] Step S23: Extract the edges of the grayscale image using an edge operator to obtain the edges of the steel coil.

[0070] The process involves reading the steel coil area and converting it to a grayscale image. Using an edge operator, if pixel values ​​show a "jump" or significant change at the edges, taking the first derivative at that edge will reveal an extremum. Where the first derivative is an extremum, the second derivative is zero. Based on this principle, edge extraction can be performed to extract the edges of the steel coil.

[0071] Please see Figure 6 , Figure 6 This is a flowchart of step S3 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0072] In step S3, the edge of the steel coil is projected onto the point cloud model of the steel coil, and the association between the edge of the steel coil and the point cloud model of the steel coil is established, including:

[0073] Step S31: Perform joint calibration of the image acquisition device and the point cloud acquisition device, and transform the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device.

[0074] Step S32: Project the edge of the steel coil onto the steel coil point cloud model to find the corresponding edge point.

[0075] It is understandable that the image acquisition device can be set to acquire data in the vertical direction relative to the end face of the steel coil. First, the image acquisition device and the point cloud acquisition device are jointly calibrated to ensure their consistency and the position of the coordinate system. The edge of the steel coil is projected onto the point cloud model, and the relevant edge points are extracted from the region corresponding to the edge of the steel coil in the point cloud model.

[0076] Please see Figure 7 , Figure 7 This is a flowchart of step S4 in the method for detecting loose coil defects in steel coils provided in an embodiment of the present invention;

[0077] In step S4, the calculation of the distance between corresponding edge points in the steel coil point cloud model includes:

[0078] Step S41: Calculate the coordinates of the center point of the steel coil point cloud model;

[0079] Step S42: Randomly select an edge point and calculate the adjacent edge points that are concentric with the coordinates of the edge point relative to the center point;

[0080] Step S43: Calculate the distance between the edge point and its adjacent edge points until all edge points have been traversed.

[0081] With the axial direction of the steel coil as the X direction, the radial direction as the Y direction, and the direction perpendicular to the ground as the Z direction, the center point coordinates of the steel coil point cloud model refer to the center point coordinates of the end face in the steel coil point cloud model. According to the defect characteristics of loose coil, it is only necessary to determine the distance between its coil layers from the end face view. The axial direction of the steel coil can be ignored. Therefore, the coordinate system of the center point coordinate O1 is (x0, z0), which can also be directly marked as the origin coordinates.

[0082] Assuming a steel coil has n layers, randomly select a first random edge point P1 (x1, z1) on the nth layer of the steel coil point cloud model. Calculate P2 (x2, y2, z2) on the n+1th layer that is coaxial with points O1 and P1, and P3 (x3, y3, z3) on the n-1th layer that is coaxial with points O1 and P1. Calculate the distance from point P1 to point P2 and the distance from point P1 to point P3, and so on, until all edge points are traversed.

[0083] It is understandable that, given the high accuracy of point cloud-based coordinate positioning, calculating the distance between edge points on the steel coil point cloud model is more accurate than the current method of using image recognition for fitting calculations. This method can determine whether the spacing of a certain area on the steel coil has exceeded the distance threshold, thus making the judgment of loose coil defects more reliable and improving product yield.

[0084] Please see Figure 8 , Figure 8 This is a flowchart of step S5 in the method for detecting loose coil defects in steel coils provided in this embodiment of the invention; step S5 involves performing a threshold judgment on the distance to determine whether the steel coil has a loose coil defect, including:

[0085] Step S51: Extract edge points whose distance exceeds a preset distance threshold as target point cloud;

[0086] Step S52: Statistically analyze the target point cloud. When the number of target point clouds exceeds a preset threshold, it is determined that the steel coil has a loose coil defect.

[0087] For example, if the distance between points P1 and P2 exceeds a distance threshold, meaning that both P1 and P2 are target point clouds, by performing a quantity count on the target point clouds, it can be determined that there are cases where the distance between layers in a certain area of ​​the steel coil exceeds the distance threshold, thereby increasing the fault tolerance of defects.

[0088] Testing Equipment

[0089] Please see Figure 9 , Figure 9 This is a connection topology diagram of the steel coil loose coil defect detection device 100 provided in an embodiment of the present invention.

[0090] Embodiment 2 of the present invention provides a detection device 100 for loose coil defects in steel coils, comprising:

[0091] Image acquisition device 101 is used to acquire a set of images of steel coils;

[0092] Point cloud acquisition device 102 is used to acquire point cloud data of steel coils;

[0093] The host computer 103 is configured to execute the detection method mentioned in the above method embodiments. By fusing 2D images with 3D LiDAR, the loose coil defect of the steel coil can be detected. This method can accurately determine the specific distance value of the loose coil, improve the accuracy of defect judgment, and has higher versatility, applicable to the judgment of loose coil defects of steel coils of different specifications.

[0094] [Detection System]

[0095] Please see Figure 10 , Figure 10This is a schematic diagram of a module of a steel coil loosening defect detection system 200 provided in an embodiment of the present invention. The detection system 200 includes:

[0096] The segmentation module 201 is used to obtain a steel coil point cloud model from the steel coil point cloud set and to obtain a steel coil region from the steel coil image set.

[0097] Edge extraction module 202 is used to extract the edges of the steel coil area to obtain the edges of the steel coil.

[0098] Projection module 203 is used to project the edge of the steel coil onto the point cloud model of the steel coil, and establish the relationship between the edge of the steel coil and the point cloud model of the steel coil.

[0099] Distance calculation module 204 is used to calculate the distance between corresponding edge points in the point cloud model of the steel coil;

[0100] The judgment module 205 is used to perform a threshold judgment on the distance to determine whether the steel coil has a loose coil defect.

[0101] Each module contains a computer program to implement its corresponding function, and the system can automatically determine the loosening of steel coils.

[0102] Furthermore, those skilled in the art should understand that if all or part of the sub-modules involved in the detection system 200 provided in the embodiments of the present invention are combined or replaced by means of fusion, simple changes, mutual transformation, etc., such as moving the position of each component; or setting the product they constitute as a whole; or having a detachable design; as long as the combined components can form a device / apparatus / system with a specific function, using such a device / apparatus / system to replace the corresponding components of the present invention also falls within the protection scope of the present invention.

[0103] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0104] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for detecting loose coil defects in steel coils, characterized in that, The detection method includes: Obtain the steel coil point cloud model from the steel coil point cloud set and the steel coil region from the steel coil image set; Edge extraction is performed on the steel coil area to obtain the steel coil edge; The edge of the steel coil is projected onto the steel coil point cloud model to establish a relationship between the steel coil edge and the steel coil point cloud model, so as to extract the edge points corresponding to the steel coil edge in the steel coil point cloud model; Calculate the distances between corresponding edge points in the point cloud model of the steel coil; A threshold judgment is performed on the distance to determine whether the steel coil has a loose coil defect; The step of projecting the edge of the steel coil onto the point cloud model of the steel coil, and establishing the association between the edge of the steel coil and the point cloud model of the steel coil, includes: The image acquisition device and the point cloud acquisition device are jointly calibrated to transform the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device; wherein, the image acquisition device is used to acquire a set of steel coil images; and the point cloud acquisition device is used to acquire a set of steel coil point clouds. Project the edge of the steel coil onto the point cloud model of the steel coil to find the corresponding edge point; The calculation of the distance between corresponding edge points in the steel coil point cloud model includes: Calculate the coordinates of the center point of the steel coil point cloud model; Randomly select an edge point and calculate the adjacent edge points that are concentric with the coordinates of the edge point relative to the center point; Calculate the distance between an edge point and its adjacent edge points, and continue this process until all edge points have been traversed.

2. The detection method according to claim 1, characterized in that, The acquisition of the steel coil point cloud model from the steel coil point cloud set and the steel coil region from the steel coil image set includes: Obtain the point cloud set and image set of the steel coil; The point cloud set of the steel coil is back-projected onto the image set of the steel coil, and the steel coil region is determined in the image set of the steel coil. The steel coil point cloud model is segmented from the point cloud set of the steel coil using a 3D point cloud algorithm.

3. The detection method according to claim 1, characterized in that, The step of extracting the edge of the steel coil region to obtain the edge of the steel coil includes: Read the steel coil area; The steel coil area is subjected to grayscale processing to obtain a grayscale image; Edge extraction is performed on the grayscale image using an edge operator to obtain the edge of the steel coil.

4. The detection method according to claim 1, characterized in that, The threshold determination of the distance to determine whether the steel coil has a loose coil defect includes: Extract edge points whose distance exceeds a preset distance threshold as the target point cloud; The target point cloud is statistically analyzed. When the number of target point clouds exceeds a preset threshold, it is determined that the steel coil has a loose coil defect.

5. A detection device for loose coil defects in steel coils, characterized in that, The detection equipment includes: Image acquisition device, used to acquire a set of images of steel coils; Point cloud acquisition device, used to acquire point cloud data of steel coils; The host computer is configured to: The steel coil region and the steel coil point cloud model are segmented from the steel coil image set and the steel coil point cloud set; Edge extraction is performed on the steel coil area to obtain the steel coil edge; The edge of the steel coil is projected onto the steel coil point cloud model to establish a relationship between the steel coil edge and the steel coil point cloud model, so as to extract the edge points corresponding to the steel coil edge in the steel coil point cloud model; Calculate the distances between corresponding edge points in the point cloud model of the steel coil; A threshold judgment is performed on the distance to determine whether the steel coil has a loose coil defect; The step of projecting the edge of the steel coil onto the steel coil point cloud model and establishing the association between the steel coil edge and the steel coil point cloud model includes: jointly calibrating the image acquisition device and the point cloud acquisition device; transforming the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device; and finding the corresponding edge point by projecting the steel coil edge onto the steel coil point cloud model. The calculation of the distance between corresponding edge points in the steel coil point cloud model includes: calculating the coordinates of the center point of the steel coil point cloud model; randomly selecting edge points and calculating adjacent edge points that are concentric with the center point coordinates of the edge points; calculating the distance between the edge points and adjacent edge points, until all edge points are traversed.

6. The detection device according to claim 5, characterized in that, The step of extracting the edge of the steel coil region to obtain the steel coil edge includes: reading the steel coil region; performing grayscale processing on the steel coil region to obtain a grayscale image; and extracting the edge of the grayscale image using an edge operator to obtain the steel coil edge. The threshold judgment of the distance to determine whether the steel coil has a loose coil defect includes: extracting edge points whose distance exceeds a preset distance threshold as target point clouds; performing statistics on the target point clouds, and judging that the steel coil has a loose coil defect when the target point cloud exceeds a preset number threshold.

7. A detection system for loose coil defects in steel coils, characterized in that, The detection system includes: The segmentation module is used to obtain a steel coil point cloud model from the steel coil point cloud set and to obtain a steel coil region from the steel coil image set; the edge extraction module is used to extract the edges of the steel coil region to obtain the steel coil edges. The projection module is used to project the edge of the steel coil onto the steel coil point cloud model, establish the association between the steel coil edge and the steel coil point cloud model, and extract the edge points corresponding to the steel coil edge in the steel coil point cloud model. The distance calculation module is used to calculate the distance between corresponding edge points in the steel coil point cloud model; The judgment module is used to perform a threshold judgment on the distance to determine whether the steel coil has a loose coil defect; The step of projecting the edge of the steel coil onto the steel coil point cloud model and establishing the association between the steel coil edge and the steel coil point cloud model includes: jointly calibrating the image acquisition device and the point cloud acquisition device; transforming the coordinates of the image acquisition device back to the coordinates of the point cloud acquisition device; and finding the corresponding edge points by projecting the steel coil edge onto the steel coil point cloud model; wherein, the image acquisition device is used to acquire a set of steel coil images; and the point cloud acquisition device is used to acquire a set of steel coil point clouds. The calculation of the distance between corresponding edge points in the steel coil point cloud model includes: calculating the coordinates of the center point of the steel coil point cloud model; randomly selecting edge points and calculating adjacent edge points that are concentric with the center point coordinates of the edge points; calculating the distance between the edge points and adjacent edge points, until all edge points are traversed.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, and one or more of the programs are executed by one or more processors to implement the detection method as claimed in any one of claims 1 to 4.

Images