A method and system for locating surface defects of a steel coil slitting line

Abstract

The present invention relates to the technical field of surface defect detection and identification, and discloses a method and system for locating surface defects of a steel coil longitudinal cutting line, comprising: a camera deployment image acquisition module, an AI detection image processing module, an application server analysis module, a front-end information receiving module, and an audible and visual alarm information sending module. The method and system determine the origin of the defect position coordinates and the pixel coordinate value information of the transverse and longitudinal offsets of the steel coil, and perform segmentation processing on the acquired image to obtain a defect identification result. Based on the origin of the defect position coordinates and the pixel coordinate value information of the transverse and longitudinal offsets of the steel coil and the defect identification result, the transverse and longitudinal positions of the defect are calculated, the absolute position information of the defect is analyzed, a printer, a labeling machine, and an audible and visual alarm are triggered to work, and defect information is displayed at the front end, thereby improving and enhancing the work efficiency of personnel currently employed to confirm surface defects of the steel coil longitudinal cutting line on site.

Description

Technical Field

[0001] The present invention relates to the technical field of surface defect detection and identification, and more particularly to a method and system for locating surface defects of a steel coil slitting line. Background Art

[0002] In the steel manufacturing and processing industries, slitting steel coils is an important finishing process used to cut wide steel coils into narrow strips of varying widths. However, during the slitting process, various defects may appear on the coil surface, such as scratches, roller marks, oxidation spots, and edge cracks. These defects can affect the quality of the final product. Therefore, positioning technology based on surface defect detection and identification is of great significance for the timely detection and treatment of surface defects on the steel coil slitting line.

[0003] However, in the positioning detection of surface defects on the longitudinal section of steel coils, the defect position information fed back by the existing detection model is not very accurate. It is given based on the coordinate position information of the defect in the defect image itself. Although on-site staff can obtain the position information of the longitudinal section surface defects in a timely and rapid manner, they cannot quickly locate the specific position of the defect in the steel coil from the position information of the longitudinal section surface defects, which has a certain impact on the rapid confirmation of steel coil defects, reduces the confirmation efficiency of the steel coil longitudinal section surface defects, prolongs the confirmation time of the steel coil longitudinal section surface defects, increases the confirmation cost of the steel coil longitudinal section surface defects, and causes certain economic losses and waste of manpower to the enterprise. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a surface defect positioning system for a steel coil longitudinal cutting line to solve the problems existing in the above-mentioned background technology.

[0005] The present invention provides the following technical solution: a surface defect positioning system for a steel coil slitting line, comprising: a camera-deployed image acquisition module, an AI detection image processing module, an application server analysis module, a front-end information receiving module, and an acoustic and visual alarm information sending module;

[0006] The camera deploys an image acquisition module, including a camera distance determination unit and an image information acquisition unit, which measures and determines the horizontal and vertical field of view distances of the camera and the distance values ​​of the image overlap areas between cameras, obtains the horizontal and vertical resolution information of the camera, and the camera performs image information acquisition and transmits it to the AI ​​detection server;

[0007] The AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit. Based on the edge-finding algorithm, it determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and performs segmentation processing on the collected image to obtain the defect recognition result;

[0008] The application server analysis module calculates the horizontal and vertical positions of the defect based on the pixel coordinate value information of the origin of the defect position coordinates and the horizontal and vertical offsets of the steel coil and the defect recognition results, and analyzes the absolute position information of the defect;

[0009] The front-end information receiving module stores the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front-end;

[0010] The sound and light alarm information sending module receives trigger instructions from the AI ​​detection image processing module and the application server analysis module, and triggers the printer, labeling machine, and sound and light alarm to work.

[0011] Preferably, the camera deployment image acquisition module includes a camera distance determination unit and an image information acquisition unit, and the specific contents are as follows:

[0012] The camera distance determination unit: before the camera captures the image, the camera position is fixed, the actual horizontal field of view distance of each camera is measured, the encoder trigger frequency is preset, and the fixed actual vertical field of view distance of the camera is obtained; the horizontal and vertical field of view distances of the camera and the distance value of the image overlap area between the cameras are measured and determined, the horizontal and vertical resolution information of the camera is obtained, and the camera configuration operation is completed;

[0013] The image information acquisition unit: The camera that collects images performs image information acquisition, wherein each camera is connected to the corresponding AI server and transmits the collected image information to the AI ​​detection server for processing.

[0014] Preferably, the AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit, and the specific contents are as follows:

[0015] The edge-finding algorithm analysis unit: The AI ​​detection server receives the image information collected by the camera, determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil according to the edge-finding algorithm, and obtains the edge pixel coordinate value of the edge;

[0016] The image segmentation processing unit performs segmentation processing on the collected image, and obtains the type of defect, defect identification image and defect pixel coordinate value in the segmented image based on the segmented image in combination with the edge finding algorithm.

[0017] Preferably, the specific content of the edge-finding algorithm analysis unit is as follows:

[0018] Step S1: Obtain an image captured by a camera, and perform edge detection and judgment based on the number of image frames of the image. The specific judgment content is:

[0019] The image captured by the camera is acquired. If there is a missing or blank image, an image missing or blank image alarm is triggered. The information is transmitted to the application server analysis module to trigger an alarm instruction. The image is then edge-finding processed and judged based on the number of image frames: the first frame is identified and edge-finding processing is performed on the horizontal and vertical blank areas of the first frame. The second to tenth frames are identified and edge-finding processing is performed on the horizontal blank areas of the second to tenth frames. After the eleventh frame, edge-finding adjustment is performed every ten frames.

[0020] Step S2: The image transmitted in step 1 is subjected to grayscale processing, binarization processing, median filtering processing, and contour detection in sequence, and edge extraction is performed based on the contour size and the image size ratio;

[0021] The grayscale processing is used to convert the color defect image captured by the camera into a grayscale image;

[0022] The binarization process: after grayscale processing is performed, converting the camera grayscale image converted by the grayscale processing into a camera image;

[0023] The median filter process: after performing the binarization process, performing an operation of removing noise from the camera image after the binarization process, while maintaining the details and edge information of the image while removing the noise;

[0024] Contour detection: After performing median filtering, the denoised camera image is subjected to recognition of the outer boundary of the steel coil in the image, i.e., the contour of the steel coil is identified, and the edge contours of the steel coil and the machine tool are identified by detecting the position where the grayscale value changes exceed a preset threshold in the image;

[0025] The edge extraction is performed based on the ratio of the outline size to the image size: after performing the outline detection process, the horizontal and vertical blank areas of the edge outlines of the steel coil and the machine tool are identified through the edge outlines of the steel coil and the machine tool;

[0026] Step S3: Obtain the actual coordinate origin of the defect position based on the horizontal and vertical blank areas of the identified steel coil and machine tool edge contours:

[0027] If there is no blank area in both horizontal and vertical directions, the point where the upper left corner of the camera and the steel coil coincide is used as the origin of the horizontal and vertical coordinates;

[0028] If there is a blank area in the horizontal direction and no blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the horizontal field of view edge of the upper left corner of the steel coil is obtained. ;

[0029] If there is no blank area in the horizontal direction and there is a blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the longitudinal field of view edge of the upper left corner of the steel coil is obtained. ;

[0030] If there are blank areas in both the horizontal and vertical directions, the upper left corner of the camera is used as the coordinate origin, and the coordinate values ​​of the horizontal and vertical edge pixels of the upper left corner of the steel coil are obtained respectively. and .

[0031] Preferably, the specific contents of the image segmentation processing unit are as follows:

[0032] Step S1: Segment the captured image, obtain defect recognition results, and save the corresponding batch and defect type;

[0033] Step S2: Obtain the horizontal and vertical field of view of the camera being operated through the steel coil batch information and , and the camera's horizontal and vertical resolution information and ;

[0034] Step S3: Combine the edge finding algorithm to obtain the relative coordinate information of the operating camera defect , and the pixel coordinate values ​​of the edges of the horizontal and vertical fields of view of the steel coil and ;

[0035] Step S4: Calculate the horizontal and vertical per-pixel meter value based on the horizontal and vertical field of view distances of the camera and the horizontal and vertical resolutions of the camera. The calculation formula for the horizontal per-pixel meter value is: ,in The horizontal value per pixel in meters is calculated as follows: ,in Indicates the horizontal value of meters per pixel.

[0036] Preferably, the application server analysis module calculates the horizontal and vertical positions of the defect and analyzes the absolute position information of the defect. The specific contents are as follows:

[0037] Based on the recognition results of the horizontal and vertical blank areas of the steel coil and machine tool edge contours, the distance value of the overlapping area between the cameras is collected. The horizontal and vertical position coordinates of the cameras are calculated by combining the pixel coordinate values ​​of the horizontal and vertical field of view edges of the steel coil and the horizontal and vertical pixel per meter values:

[0038] If there are no blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel meters, the vertical resolution, and the overlap area distance.

[0039] If there is a blank area in the horizontal direction but no blank area in the vertical direction, the horizontal and vertical positions of the camera are calculated based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap area distance, and the horizontal edge blank area length.

[0040] If there is no blank area in the horizontal direction but there is a blank area in the vertical direction, the horizontal and vertical positions of the camera are calculated based on the pixel coordinates of the defect, the horizontal and vertical pixel meters, the vertical resolution, the overlap area distance, and the vertical edge blank area length.

[0041] If there are blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap distance, and the length of the blank areas at the horizontal and vertical edges.

[0042] The horizontal and vertical position information of the corresponding frame camera is saved, an alarm instruction is triggered and transmitted to the sound and light alarm information sending module, and the horizontal and vertical position information represents the absolute position information of the defect.

[0043] Preferably, the front-end information receiving module saves the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front end. The defect information includes: defect identification results, corresponding batches, corresponding defect types and absolute position information of the defects.

[0044] Preferably, the sound and light alarm information sending module receives the trigger instructions issued by the edge-finding algorithm analysis unit in the AI ​​detection image processing module and the trigger instructions issued by the application server analysis module, triggering the printer, labeling machine, and sound and light alarm to work, wherein the printer is used to print labels with defect information, the labeling machine is used to stick the printed labels on the defective steel coils, and the sound and light alarm is used to send sound and light alarm signals.

[0045] A method for locating surface defects of a steel coil slitting line comprises the following steps:

[0046] Step S01: Measure and determine the horizontal and vertical field of view distances of the cameras and the distance values ​​of the image overlap areas between the cameras, and obtain the horizontal and vertical resolution information of the cameras;

[0047] Step S02: Determine the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and segment the collected image to obtain the defect recognition result;

[0048] Step S03: Calculate the horizontal and vertical positions of the defect based on the coordinate origin of the defect position and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil and the defect recognition result, and analyze the absolute position information of the defect;

[0049] Step S04: Save the defect identification information and display the defect information on the front end;

[0050] Step S05: receiving a trigger instruction, triggering the printer, labeling machine, and sound and light alarm to work.

[0051] The technical effects and advantages of the present invention are as follows:

[0052] The present invention greatly improves the accuracy of the position information of the surface defects of the steel coil longitudinal section line on the basis of ensuring the accurate positioning of the position information of the surface defects of the steel coil longitudinal section line, shortens the confirmation time of the surface defects of the steel coil longitudinal section line, reduces the labor cost of confirming the surface defects of the steel coil longitudinal section line, and improves the confirmation speed of the surface defects of the steel coil longitudinal section line;

[0053] Based on the real-time absolute position information of surface defects on the coil slitting line, on-site verification personnel can reach the exact verification location immediately, eliminating the previous situation of running back and forth for defect verification, reducing the workload of verification personnel and improving the on-site working environment for verification personnel;

[0054] This effectively reduces the time required to confirm surface defects on the coil slitting line, avoiding delays in confirmation, the accumulation of confirmation issues, and omissions caused by untimely confirmation of the location of surface defects on the coil slitting line. It also enhances the enthusiasm of on-site confirmation personnel, improves the work atmosphere at the confirmation site, and increases the accuracy and efficiency of on-site defect confirmation. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a schematic diagram of the structure of a surface defect positioning system for a steel coil longitudinal cutting line.

[0056] Figure 2 The figure is a flow chart of a method for locating surface defects on a steel coil longitudinal cutting line. DETAILED DESCRIPTION

[0057] The technical solutions of the present invention will be described clearly and completely below in conjunction with the accompanying drawings of the present invention. In addition, the forms of the various structures described in the following embodiments are merely examples. The surface defect positioning method and system for a steel coil longitudinal cutting line involved in the present invention are not limited to the various structures described in the following embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] like Figure 1 As shown, the present invention provides a surface defect positioning system for a steel coil slitting line, comprising: a camera deployment image acquisition module, an AI detection image processing module, an application server analysis module, a front-end information receiving module, and an acoustic and visual alarm information sending module;

[0059] The camera deploys an image acquisition module, including a camera distance determination unit and an image information acquisition unit, which measures and determines the horizontal and vertical field of view distances of the camera and the distance values ​​of the image overlap areas between cameras, obtains the horizontal and vertical resolution information of the camera, and the camera performs image information acquisition and transmits it to the AI ​​detection server;

[0060] The AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit. Based on the edge-finding algorithm, it determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and performs segmentation processing on the collected image to obtain the defect recognition result;

[0061] The application server analysis module calculates the horizontal and vertical positions of the defect based on the pixel coordinate value information of the origin of the defect position coordinates and the horizontal and vertical offsets of the steel coil and the defect recognition results, and analyzes the absolute position information of the defect;

[0062] The front-end information receiving module stores the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front-end;

[0063] The sound and light alarm information sending module receives trigger instructions from the AI ​​detection image processing module and the application server analysis module, and triggers the printer, labeling machine, and sound and light alarm to work.

[0064] In this embodiment, it should be specifically explained that the camera deployment image acquisition module includes the camera distance determination unit and the image information acquisition unit. The specific contents are as follows:

[0065] The camera distance determination unit: before the camera captures the image, the camera position is fixed, the actual horizontal field of view distance of each camera is measured, the encoder trigger frequency is preset, and the fixed actual vertical field of view distance of the camera is obtained; the vertical field of view is determined according to the encoder trigger condition, the horizontal and vertical field of view distances of the camera and the distance value of the image overlap area between the cameras are measured and determined, the horizontal and vertical resolution information of the camera is obtained, and the camera configuration operation is completed;

[0066] The image information acquisition unit: The camera that collects images performs image information acquisition, wherein each camera is connected to the corresponding AI server and transmits the collected image information to the AI ​​detection server for processing.

[0067] In this embodiment, it should be specifically explained that the AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit. Specifically, the contents are as follows:

[0068] The edge-finding algorithm analysis unit: The AI ​​detection server receives the image information collected by the camera, determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil according to the edge-finding algorithm, and obtains the edge pixel coordinate value of the edge;

[0069] The image segmentation processing unit performs segmentation processing on the collected image, and obtains the type of defect, defect identification image and defect pixel coordinate value in the segmented image based on the segmented image in combination with the edge finding algorithm.

[0070] In this embodiment, it should be specifically explained that the specific contents of the edge-finding algorithm analysis unit are as follows:

[0071] Step S1: Obtain an image captured by a camera, and perform edge detection and judgment based on the number of image frames of the image. The specific judgment content is:

[0072] The image captured by the camera is acquired. If there is a missing or blank image, an image missing or blank image alarm is triggered. The information is transmitted to the application server analysis module to trigger an alarm instruction. The image is then edge-finding processed and judged based on the number of image frames: the first frame is identified and edge-finding processing is performed on the horizontal and vertical blank areas of the first frame. The second to tenth frames are identified and edge-finding processing is performed on the horizontal blank areas of the second to tenth frames. After the eleventh frame, edge-finding adjustment is performed every ten frames.

[0073] Step S2: The image transmitted in step 1 is subjected to grayscale processing, binarization processing, median filtering processing, and contour detection in sequence, and edge extraction is performed based on the contour size and the image size ratio;

[0074] Grayscale processing: used to convert the color defect image captured by the camera into a grayscale image, in order to improve the image quality, make the image display effect clearer, and facilitate subsequent image processing;

[0075] The binarization process: after grayscale processing is performed, the grayscale image converted by the grayscale processing is converted into a camera image, wherein the camera image only contains two colors, black and white, so as to simplify the image data, highlight the image features, and facilitate subsequent processing;

[0076] The median filter process is to remove noise from the binarized camera image after the binarization process, while maintaining the image details and edge information while removing the noise to avoid edge blurring.

[0077] Contour detection: After performing median filtering, the denoised camera image is subjected to recognition of the outer boundaries of the steel coil in the image, i.e., the contour of the steel coil is identified. These boundaries are generally composed of a series of connected points, which can describe the overall shape of the captured steel coil and are locations where there are significant changes in brightness in the image. These locations are generally the dividing lines or overlapping lines between the steel coil and the machine tool. The edge contours of the steel coil and the machine tool are identified by detecting locations in the image where the grayscale value changes exceed a preset threshold, as the locations where there are significant changes are generally located at the contours or boundaries of the steel coil or the machine tool.

[0078] The edge extraction is performed based on the ratio of the outline size to the image size: after performing the outline detection process, the horizontal and vertical blank areas of the edge outlines of the steel coil and the machine tool are identified through the edge outlines of the steel coil and the machine tool;

[0079] Step S3: Obtain the actual coordinate origin of the defect position based on the horizontal and vertical blank areas of the identified steel coil and machine tool edge contours:

[0080] If there is no blank area in both horizontal and vertical directions, the point where the upper left corner of the camera and the steel coil coincide is used as the origin of the horizontal and vertical coordinates;

[0081] If there is a blank area in the horizontal direction and no blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the horizontal field of view edge of the upper left corner of the steel coil is obtained. ;

[0082] If there is no blank area in the horizontal direction and there is a blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the longitudinal field of view edge of the upper left corner of the steel coil is obtained. ;

[0083] If there are blank areas in both the horizontal and vertical directions, the upper left corner of the camera is used as the coordinate origin, and the coordinate values ​​of the horizontal and vertical edge pixels of the upper left corner of the steel coil are obtained respectively. and .

[0084] In this embodiment, it should be specifically explained that the specific contents of the image segmentation processing unit are as follows:

[0085] Step S1: Segment the captured image, obtain defect recognition results, and save the corresponding batch and defect type;

[0086] Step S2: Obtain the horizontal and vertical field of view of the camera being operated through the steel coil batch information and , and the camera's horizontal and vertical resolution information and ;

[0087] Step S3: Combine the edge finding algorithm to obtain the relative coordinate information of the operating camera defect , and the pixel coordinate values ​​of the edges of the horizontal and vertical fields of view of the steel coil and , the defect relative coordinate information represents the pixel coordinate value of the defect in the image;

[0088] Step S4: Calculate the horizontal and vertical per-pixel meter value based on the horizontal and vertical field of view distances of the camera and the horizontal and vertical resolutions of the camera. The calculation formula for the horizontal per-pixel meter value is: ,in The horizontal value per pixel in meters is calculated as follows: ,in Indicates the horizontal value of meters per pixel.

[0089] In this embodiment, it should be specifically explained that the application server analysis module calculates the horizontal and vertical positions of the defect and analyzes the absolute position information of the defect. The specific content is as follows:

[0090] Based on the recognition results of the horizontal and vertical blank areas of the steel coil and machine tool edge contours, the distance value of the overlapping area between the cameras is collected. The horizontal and vertical position coordinates of the cameras are calculated by combining the pixel coordinate values ​​of the horizontal and vertical field of view edges of the steel coil and the horizontal and vertical per pixel meter values. When there are two cameras, the specific contents of the horizontal and vertical position coordinate calculation of camera 1 and camera 2 are as follows:

[0091] If there are no blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, and the overlap area distance:

[0092] The calculation formula for the lateral position of the n-th frame defect image of camera 1 is: , the calculation formula for the longitudinal position is ,in Indicates the horizontal position, Indicates the abscissa value in the relative coordinates of the camera defect, Indicates the horizontal value per pixel in meters. Indicates the vertical position, n indicates the number of image frames, Indicates the vertical resolution information of the camera. Indicates the horizontal value per pixel in meters. The ordinate value in the relative coordinates representing the camera defect;

[0093] The calculation formula for the lateral position of the n-th frame defect image of camera 2 is: , the calculation formula for the longitudinal position is ,in Indicates the distance value of the overlapping area;

[0094] If there is a blank area in the horizontal direction but no blank area in the vertical direction, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap distance, and the horizontal edge blank area length:

[0095] The calculation formula for the lateral position of the n-th frame defect image of camera 1 is: , the calculation formula for the longitudinal position is ,in Indicates the length of the horizontal edge blank area;

[0096] The calculation formula for the lateral position of the n-th frame defect image of camera 2 is: , the calculation formula for the longitudinal position is ,in Indicates the horizontal field of view distance of the camera;

[0097] If there is no blank area in the horizontal direction but there is a blank area in the vertical direction, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap distance, and the vertical edge blank area length:

[0098] The calculation formula for the lateral position of the n-th frame defect image of camera 1 is: , the calculation formula for the longitudinal position is ,in Indicates the length of the blank area on the vertical edge;

[0099] The calculation formula for the lateral position of the n-th frame defect image of camera 2 is: , the calculation formula for the longitudinal position is ;

[0100] If there are blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap distance, and the length of the blank areas at the horizontal and vertical edges:

[0101] The calculation formula for the lateral position of the n-th frame defect image of camera 1 is: , the calculation formula for the longitudinal position is ;

[0102] The calculation formula for the lateral position of the n-th frame defect image of camera 2 is: , the calculation formula for the longitudinal position is ;

[0103] The horizontal and vertical position information of the corresponding frame camera is saved, an alarm instruction is triggered and transmitted to the sound and light alarm information sending module, and the horizontal and vertical position information represents the absolute position information of the defect.

[0104] In this embodiment, it should be specifically explained that the front-end information receiving module saves the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front end. The defect information includes: defect identification results, corresponding batches, corresponding defect types and absolute position information of the defects.

[0105] In this embodiment, it should be specifically explained that the sound and light alarm information sending module receives the trigger instructions issued by the edge-finding algorithm analysis unit in the AI ​​detection image processing module and the trigger instructions issued by the application server analysis module, triggering the printer, labeling machine, and sound and light alarm to work, wherein the printer is used to print labels with defect information, the labeling machine is used to stick the printed labels on the defective steel coils, and the sound and light alarm is used to send sound and light alarm signals.

[0106] like Figure 2 As shown, in this embodiment, it should be specifically explained that a method for locating surface defects of a steel coil longitudinal cutting line includes the following steps:

[0107] Step S01: Measure and determine the horizontal and vertical field of view distances of the cameras and the distance values ​​of the image overlap areas between the cameras, and obtain the horizontal and vertical resolution information of the cameras;

[0108] Step S02: Determine the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and segment the collected image to obtain the defect recognition result;

[0109] Step S03: Calculate the horizontal and vertical positions of the defect based on the coordinate origin of the defect position and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil and the defect recognition result, and analyze the absolute position information of the defect;

[0110] Step S04: Save the defect identification information and display the defect information on the front end;

[0111] Step S05: receiving a trigger instruction, triggering the printer, labeling machine, and sound and light alarm to work.

[0112] In this embodiment, it should be specifically explained that the difference between this embodiment and the prior art lies in that this embodiment greatly improves the accuracy of the position information of the surface defects of the steel coil slit line while ensuring the accurate positioning of the position information of the surface defects of the steel coil slit line, shortens the time for confirming the surface defects of the steel coil slit line, reduces the labor cost for confirming the surface defects of the steel coil slit line, and improves the speed of confirming the surface defects of the steel coil slit line;

[0113] Based on the real-time absolute position information of surface defects on the coil slitting line, on-site verification personnel can reach the exact verification location immediately, eliminating the previous situation of running back and forth for defect verification, reducing the workload of verification personnel and improving the on-site working environment for verification personnel;

[0114] This effectively reduces the time required to confirm surface defects on the coil slitting line, avoiding delays in confirmation, the accumulation of confirmation issues, and omissions caused by untimely confirmation of the location of surface defects on the coil slitting line. It also enhances the enthusiasm of on-site confirmation personnel, improves the work atmosphere at the confirmation site, and increases the accuracy and efficiency of on-site defect confirmation.

[0115] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0116] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A surface defect location system for a steel coil slitting line, characterized by: include: The camera deploys an image acquisition module, an AI detection image processing module, an application server analysis module, a front-end information receiving module, and an audio-visual alarm information sending module; The camera deploys an image acquisition module, including a camera distance determination unit and an image information acquisition unit, which measures and determines the horizontal and vertical field of view distances of the camera and the distance values ​​of the image overlap areas between cameras, obtains the horizontal and vertical resolution information of the camera, and the camera performs image information acquisition and transmits it to the AI ​​detection server; The AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit. Based on the edge-finding algorithm, it determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and performs segmentation processing on the collected images to obtain defect recognition results. The AI ​​detection image processing module includes an edge-finding algorithm analysis unit and an image segmentation processing unit. The specific contents are as follows: The edge-finding algorithm analysis unit: The AI ​​detection server receives the image information collected by the camera, determines the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil according to the edge-finding algorithm, and obtains the edge pixel coordinate value of the edge; The image segmentation processing unit is configured to perform segmentation processing on the acquired image, and obtain the defect type, defect identification image, and defect pixel coordinate values ​​in the segmented image based on the segmented image and in combination with an edge-finding algorithm; The specific contents of the edge-finding algorithm analysis unit are as follows: Step S1: Obtain an image captured by a camera, and perform edge detection and judgment based on the number of image frames of the image. The specific judgment content is: The image captured by the camera is acquired. If there is a missing or blank image, an image missing or blank image alarm is triggered. The information is transmitted to the application server analysis module to trigger an alarm instruction. The image is then edge-finding processed and judged based on the number of image frames: the first frame is identified and edge-finding processing is performed on the horizontal and vertical blank areas of the first frame. The second to tenth frames are identified and edge-finding processing is performed on the horizontal blank areas of the second to tenth frames. After the eleventh frame, edge-finding adjustment is performed every ten frames. Step S2: The image transmitted in step 1 is subjected to grayscale processing, binarization processing, median filtering processing, and contour detection in sequence, and edge extraction is performed based on the contour size and the image size ratio; The grayscale processing is used to convert the color defect image captured by the camera into a grayscale image; The binarization process: after grayscale processing is performed, converting the camera grayscale image converted by the grayscale processing into a camera image; The median filter process: after performing the binarization process, performing an operation of removing noise from the camera image after the binarization process, while maintaining the details and edge information of the image while removing the noise; Contour detection: After performing median filtering, the denoised camera image is subjected to recognition of the outer boundary of the steel coil in the image, i.e., the contour of the steel coil is identified, and the edge contours of the steel coil and the machine tool are identified by detecting the position where the grayscale value changes exceed a preset threshold in the image; The edge extraction is performed based on the ratio of the outline size to the image size: after performing the outline detection process, the horizontal and vertical blank areas of the edge outlines of the steel coil and the machine tool are identified through the edge outlines of the steel coil and the machine tool; Step S3: Obtain the actual coordinate origin of the defect position based on the horizontal and vertical blank areas of the identified steel coil and machine tool edge contours: If there is no blank area in both horizontal and vertical directions, the point where the upper left corner of the camera and the steel coil coincide is used as the origin of the horizontal and vertical coordinates; If there is a blank area in the horizontal direction and no blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the horizontal field of view edge of the upper left corner of the steel coil is obtained. ; If there is no blank area in the horizontal direction and there is a blank area in the vertical direction, the upper left corner of the camera is used as the coordinate origin, and the pixel coordinate value of the longitudinal field of view edge of the upper left corner of the steel coil is obtained. ; If there are blank areas in both the horizontal and vertical directions, the upper left corner of the camera is used as the coordinate origin, and the coordinate values ​​of the horizontal and vertical edge pixels of the upper left corner of the steel coil are obtained respectively. and ; The application server analysis module calculates the horizontal and vertical positions of the defect based on the pixel coordinate value information of the origin of the defect position coordinates and the horizontal and vertical offsets of the steel coil and the defect recognition results, and analyzes the absolute position information of the defect; The front-end information receiving module stores the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front-end; The sound and light alarm information sending module receives trigger instructions from the AI ​​detection image processing module and the application server analysis module, and triggers the printer, labeling machine, and sound and light alarm to work.

2. The surface defect location system for a steel coil slitting line according to claim 1, characterized in that: The camera deployment image acquisition module includes a camera distance determination unit and an image information acquisition unit. The specific contents are as follows: The camera distance determination unit: before the camera captures the image, the camera position is fixed, the actual horizontal field of view distance of each camera is measured, the encoder trigger frequency is preset, and the fixed actual vertical field of view distance of the camera is obtained; the horizontal and vertical field of view distances of the camera and the distance value of the image overlap area between the cameras are measured and determined, the horizontal and vertical resolution information of the camera is obtained, and the camera configuration operation is completed; The image information acquisition unit: The camera that collects images performs image information acquisition, wherein each camera is connected to the corresponding AI server and transmits the collected image information to the AI ​​detection server for processing.

3. The surface defect location system for a steel coil slitting line according to claim 1, characterized in that: The specific contents of the image segmentation processing unit are as follows: Step S1: Segment the captured image, obtain defect recognition results, and save the corresponding batch and defect type; Step S2: Obtain the horizontal and vertical field of view of the camera being operated through the steel coil batch information and , and the camera's horizontal and vertical resolution information and ; Step S3: Combine the edge finding algorithm to obtain the relative coordinate information of the operating camera defect , and the pixel coordinate values ​​of the edges of the horizontal and vertical fields of view of the steel coil and ; Step S4: Calculate the horizontal and vertical per-pixel meter value based on the horizontal and vertical field of view distances of the camera and the horizontal and vertical resolutions of the camera. The calculation formula for the horizontal per-pixel meter value is: ,in The horizontal value per pixel in meters is calculated as follows: ,in Indicates the horizontal value of meters per pixel.

4. The surface defect location system for a steel coil slitting line according to claim 1, characterized in that: The application server analysis module calculates the horizontal and vertical positions of the defect and analyzes the absolute position information of the defect. The specific contents are as follows: Based on the recognition results of the horizontal and vertical blank areas of the steel coil and machine tool edge contours, the distance value of the overlapping area between the cameras is collected. The horizontal and vertical position coordinates of the cameras are calculated by combining the pixel coordinate values ​​of the horizontal and vertical field of view edges of the steel coil and the horizontal and vertical pixel per meter values: If there are no blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel meters, the vertical resolution, and the overlap area distance. If there is a blank area in the horizontal direction but no blank area in the vertical direction, the horizontal and vertical positions of the camera are calculated based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap area distance, and the horizontal edge blank area length. If there is no blank area in the horizontal direction but there is a blank area in the vertical direction, the horizontal and vertical positions of the camera are calculated based on the pixel coordinates of the defect, the horizontal and vertical pixel meters, the vertical resolution, the overlap area distance, and the vertical edge blank area length. If there are blank areas in both the horizontal and vertical directions, calculate the horizontal and vertical positions of the camera based on the pixel coordinates of the defect, the horizontal and vertical pixel values ​​in meters, the vertical resolution, the overlap distance, and the length of the blank areas at the horizontal and vertical edges. The horizontal and vertical position information of the corresponding frame camera is saved, an alarm instruction is triggered and transmitted to the sound and light alarm information sending module, and the horizontal and vertical position information represents the absolute position information of the defect.

5. The surface defect location system for a steel coil slitting line according to claim 1, characterized in that: The front-end information receiving module saves the information transmitted by the AI ​​detection image processing module and the application server analysis module, and displays the defect information on the front end. The defect information includes: defect identification results, corresponding batches, corresponding defect types, and absolute location information of the defects.

6. The surface defect location system for a steel coil slitting line according to claim 1, characterized in that: The sound and light alarm information sending module receives the trigger instructions issued by the edge-finding algorithm analysis unit in the AI ​​detection image processing module and the trigger instructions issued by the application server analysis module, triggering the printer, labeling machine, and sound and light alarm to work, wherein the printer is used to print labels with defect information, the labeling machine is used to stick the printed labels on the defective steel coils, and the sound and light alarm is used to send sound and light alarm signals.

7. A method for locating surface defects on a steel coil slitting line, using a system for locating surface defects on a steel coil slitting line according to any one of claims 1 to 6, characterized in that: The following steps are involved: Step S01: Measure and determine the horizontal and vertical field of view distances of the cameras and the distance values ​​of the image overlap areas between the cameras, and obtain the horizontal and vertical resolution information of the cameras; Step S02: Determine the origin of the defect position coordinates and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil, and segment the collected image to obtain the defect recognition result; Step S03: Calculate the horizontal and vertical positions of the defect based on the coordinate origin of the defect position and the pixel coordinate value information of the horizontal and vertical offsets of the steel coil and the defect recognition result, and analyze the absolute position information of the defect; Step S04: Save the defect identification information and display the defect information on the front end; Step S05: receiving a trigger instruction, triggering the printer, labeling machine, and sound and light alarm to work.

Images