Alloy steel thickness detecting and marking system

Abstract

The invention discloses an alloy steel thickness detection marking system, belongs to the technical field of steel, and solves the problems that the detection mode can only detect local parts of the surface of the alloy steel, the detection range is limited, the detection mode cannot be applied to detection of the surface of the whole alloy steel, and the detection efficiency is low. And the thickness of the batch of alloy steel cannot be detected and marked. Comprising an unwinding group, a detection mark group and a transmission group, and steel is unwound by the unwinding group and passes through the detection mark group to the transmission group. When the device works, steel is unfolded from the unwinding set through the unwinding shaft and enters the detection mark set through the guide frame. The grid laser projects grid laser on the top face and the bottom face of the steel, the image collecting module shoots and transmits data to the processing module, and the area with the uneven thickness is detected and then fed back to the control module. The control module then controls the servo motor and the air pump, adjusts the position of the mark pen and marks the position with uneven thickness. And the steel continues to flow, and continuous unwinding, detection and marking are achieved.

Description

Technical Field

[0001] The invention relates to the technical field of steel materials, and in particular to a thickness detection marking system for alloy steel materials. Background Art

[0002] Alloy steel is a type of steel that improves its performance by adding one or more alloying elements to ordinary carbon steel. These alloying elements can significantly improve the strength, toughness, wear resistance, corrosion resistance and high temperature performance of steel, making it widely used in many industrial fields. For example, chromium-containing stainless steel has excellent corrosion resistance and can be used in chemical equipment and medical devices; while nickel-containing alloy steel has good low-temperature toughness and is suitable for aerospace and marine engineering. The production process of alloy steel is complex, and usually requires high-temperature smelting, precise composition control, and special heat treatment processes to ensure that its performance meets the design requirements. Despite its relatively high cost, the high performance of alloy steel makes it indispensable in modern industry. From automobile manufacturing to high-end mechanical processing, from building structures to energy equipment, alloy steel plays a key role. With the continuous advancement of science and technology, the performance of alloy steel is also continuously optimized, providing important support for promoting industrial development and technological innovation.

[0003] Thickness uniformity testing of steel is a key link to ensure its quality and performance. In practical applications, the thickness uniformity of steel directly affects its bearing capacity, processing accuracy, corrosion resistance and service life. For example, in building structures, if the thickness of steel is uneven, it may lead to insufficient structural strength or local stress concentration, thus causing safety hazards. In the field of mechanical manufacturing, steel with uneven thickness may not meet the processing accuracy requirements, resulting in assembly difficulties or equipment operation failures. In addition, for some special-purpose steels, such as aerospace materials and pressure vessels, thickness uniformity is an even more critical indicator because it is directly related to the reliability and safety of the product. Therefore, thickness uniformity testing is an indispensable part of steel quality control. By adopting modern technical means such as ultrasonic testing, electromagnetic induction testing, and laser thickness measurement, the thickness uniformity of steel can be evaluated efficiently and accurately, thereby providing reliable quality assurance for the production, processing and use of steel.

[0004] In the existing technology, the thickness uniformity detection of steel mainly adopts ultrasonic detection, electromagnetic induction detection, laser thickness measurement and mechanical measurement. Ultrasonic detection calculates the thickness by measuring the propagation time of sound waves in steel, which is suitable for a variety of steels; electromagnetic induction detection uses the eddy current principle and is suitable for thin plates and rapid detection; laser thickness measurement uses laser reflection or scattering signals, which is highly accurate and non-contact. These technologies can meet the detection needs of different scenarios and ensure the uniformity of steel thickness.

[0005] The above detection method can only detect a part of the surface of the alloy steel, and its detection range is limited. It cannot be applied to the detection of the surface of the entire alloy steel, and it is even more impossible to detect and mark the thickness of this batch of alloy steel.

[0006] Therefore, an alloy steel thickness detection marking system is proposed to solve or alleviate the above problems. Summary of the invention

[0007] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a thickness detection marking system for alloy steel.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions: A thickness detection and marking system for alloy steel comprises an unwinding group, a detection and marking group, and a transmission group. The unwinding group unwinds the steel through the detection and marking group to the transmission group, and the transmission group pulls the steel to unwind. The detection and marking group comprises a detection part and a marking part. The detection part releases a grid-shaped laser to the top and bottom surfaces of the steel and collects images to the marking part. The marking part recognizes the image and marks the contour at the uneven thickness of the top and bottom surfaces of the steel.

[0009] Preferably, the unwinding group includes an unwinding frame, an unwinding roller rotatably connected to the unwinding frame, a guide frame fixedly connected to the unwinding frame and having a horizontal height greater than a horizontal height of the unwinding roller, and a guide roller rotatably connected to the guide frame.

[0010] Preferably, the detection mark group further includes an intermediate frame, and the detection part and the marking part are arranged on the intermediate frame.

[0011] Preferably, the detection unit includes a vertical frame, an upper frame and a lower frame fixedly connected to the vertical frame, and two groups of grid lasers and image acquisition modules respectively arranged on the upper frame and the lower frame, a channel for steel to pass through is formed between the upper frame and the lower frame, the two groups of grid lasers and image acquisition modules are arranged on the side where the upper frame and the lower frame face each other, and the image acquisition module is used to capture images of grid lasers attached to the steel.

[0012] Preferably, the detection unit also includes a data processing module and a control module. The output end of the image acquisition module is coupled to the input end of the data processing module, and the output end of the data processing module is coupled to the control module. The data processing module identifies the position of the pits on the surface of the steel material according to the image and outputs its contour and then transmits it to the control module.

[0013] Preferably, the marking part includes a dwarf gantry and a high gantry, and the dwarf gantry and the high gantry are both provided with two groups of horizontally displaceable drawing components, the drawing components located on the dwarf gantry are vertically arranged upward, and the drawing components located on the high gantry are vertically arranged downward, and the dwarf gantry and the high gantry are both fixedly connected with guide rails and racks, and the drawing components include a mounting frame sleeved on the outside of the dwarf gantry and the high gantry, a servo motor fixedly connected to the mounting frame, a cylinder and a mounting seat fixedly connected to the mounting frame, a slide seat slidably connected in the mounting seat and fixedly connected to the movable end of the cylinder, and a marker pen fixedly connected to the slide seat through a connecting seat, a transmission gear meshing with the rack is fixedly connected to the rotor shaft of the servo motor, the cylinder is connected to an air pump, and the air pump and the servo motor are both coupled to a control module.

[0014] Preferably, the transmission group includes a main frame, a flattening part fixedly connected to the main frame, and a roller part fixedly connected to the main frame, the pressing part includes two left and right groups of pressing plates, and a through path for steel to pass through is formed between the two groups of pressing plates, and the roller part includes several groups of pressing rollers located at the same horizontal height, and each group of pressing rollers includes two pressing rollers that are parallel and rotatably arranged up and down.

[0015] Preferably, the pressing plate portion includes two groups of front slides fixedly connected to the main frame, and a front slider slidably connected to the front slide, the upper end of the front slide is provided with a front threaded hole, and a front guide bolt rotatably connected to the front slider is threadedly connected in the front threaded hole, the pressing plate is fixedly connected to the front slider, the roller portion includes several groups of rear slides, and a rear slider slidably connected to the rear slide, the upper end of the rear slide is provided with a rear threaded hole, and a rear guide bolt rotatably connected to the rear slider is threadedly connected in the rear threaded hole, and the pressure roller is rotatably connected between the same group of rear sliders.

[0016] Preferably, the data processing module identifies the location of the pits on the steel surface according to the image and outputs its contour and transmits it to the control module, comprising the following steps: Generate a laser projection pattern that is a superposition of a main grid and an adaptively activated auxiliary grid based on the historical defect probability distribution and real-time image gradient; Through anisotropic diffusion filtering and non-uniform illumination correction; The phase correlation method is used to extract the sub-pixel coordinates of grid intersections, and sub-pixel positioning is achieved through topological constraint optimization; Based on the elastic registration model, the displacement fields of the top and bottom images are fused and the three-dimensional deformation field is calculated by combining the stereo vision back projection. The thickness field is constructed by coupling the three-dimensional deformation field with the material mechanical parameters, and the abnormal area is detected by wavelet multiscale analysis; Level set evolution is used to extract continuous closed contours, and pseudo-defect areas are eliminated through topological invariance verification.

[0017] The present invention has the following beneficial effects: When the present invention is working, the steel is unwound from the unwinding group through the unwinding shaft and enters the detection and marking group through the guide frame. The grid laser projects the grid laser on the top and bottom surfaces of the steel, and the image acquisition module captures and transmits data to the processing module, and then feeds back to the control module after detecting the uneven thickness area. The control module then controls the servo motor and air pump to adjust the position of the marker pen and mark the uneven thickness position. The steel continues to flow, achieving continuous unwinding, detection and marking. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 It is a structural schematic diagram of the present invention; Figure 2 for Figure 1 The enlarged view of point A in the middle; Figure 3 for Figure 1 The enlarged view of point B in the middle; Figure 4 It is a structural block diagram of the detection system in the present invention.

[0020] 1. Unwinding frame; 2. Unwinding roller; 3. Guide frame; 4. Guide roller; 5. Intermediate frame; 6. Flat frame; 7. Clamping rod; 8. Vertical frame; 9. Upper frame; 10. Lower frame; 11. Low gantry; 12. High gantry; 13. Guide rail; 14. Rack; 15. Mounting frame; 16. Servo motor; 17. Mounting seat; 18. Cylinder; 19. Slide seat; 20. Connecting seat; 21. Marker pen; 22. Main frame; 23. Front slide; 24. Front slider; 25. Front guide bolt; 26. Pressing sheet; 27. Rear slide; 28. Rear slider; 29. ​​Rear guide bolt; 30. Pressing roller; 31. Grid laser; 32. Image acquisition module; 33. Data processing module; 34. Control module; 35. Air pump. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0024] In the description of the present invention, it should be understood that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, or are the orientations or positional relationships in which the product of the invention is conventionally placed when in use, or are the orientations or positional relationships conventionally understood by those skilled in the art. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0025] Furthermore, the terms “first”, “second”, “third”, etc. are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0026] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0027] A thickness detection and marking system for alloy steel. like Figure 1As shown, it includes an unwinding group, a detection and marking group, and a transmission group. The unwinding group unwinds the steel through the detection and marking group to the transmission group. The transmission group pulls the steel to unwind. The detection and marking group includes an intermediate frame 5, a detection part and a marking part. The detection part and the marking part are arranged on the intermediate frame 5. The detection part releases a grid-shaped laser to the top and bottom surfaces of the steel and collects images to the marking part. The marking part recognizes the image and marks the contour at the uneven thickness of the top and bottom surfaces of the steel.

[0028] like Figure 1 As shown, the unwinding group includes an unwinding frame 1, an unwinding roller 2 rotatably connected to the unwinding frame 1, a guide frame 3 fixedly connected to the unwinding frame 1 and having a horizontal height greater than that of the unwinding roller 2, and a guide roller 4 rotatably connected to the guide frame 3.

[0029] like Figure 2 As shown, the detection unit includes a vertical frame 8, an upper frame 9 and a lower frame 10 fixedly connected to the vertical frame 8, two groups of grid lasers 31 and image acquisition modules 32, a data processing module 33, and a control module 34 respectively arranged on the upper frame 9 and the lower frame 10, and a channel for steel to pass through is formed between the upper frame 9 and the lower frame 10, as shown in FIG. Figure 4 As shown, two groups of grid lasers 31 and image acquisition modules 32 are arranged on the side where the upper frame 9 and the lower frame 10 face each other. The image acquisition module 32 is used to collect images of grid lasers attached to steel. The output end of the image acquisition module 32 is coupled to the input end of the data processing module 33. The output end of the data processing module 33 is coupled to the control module 34. The data processing module 33 identifies the position of the pits on the surface of the steel according to the image and outputs its contour and then transmits it to the control module 34.

[0030] like Figure 2 As shown, the marking part includes a short gantry 11 and a high gantry 12. The short gantry 11 and the high gantry 12 are both provided with two sets of horizontally displaceable picture components. The picture components on the short gantry 11 are vertically arranged upwards, and the picture components on the high gantry 12 are vertically arranged downwards. The short gantry 11 and the high gantry 12 are both fixedly connected with guide rails 13 and racks 14. The picture components include a mounting frame 1 sleeved on the outside of the short gantry 11 and the high gantry 12. 5. A servo motor 16 fixedly connected to a mounting frame 15, a cylinder 18 and a mounting seat 17 fixedly connected to the mounting frame 15, a slide seat 19 slidably connected in the mounting seat 17 and fixedly connected to a movable end of the cylinder 18, and a marker pen 21 fixedly connected to the slide seat 19 via a connecting seat 20. A transmission gear meshing with the rack 14 is fixedly connected to the rotor shaft of the servo motor 16. The cylinder 18 is connected to an air pump 35. The air pump 35 and the servo motor 16 are both coupled to a control module 34.

[0031] like Figure 3As shown, the transmission group includes a main frame 22, a flattening part fixedly connected to the main frame 22, and a roller part fixedly connected to the main frame 22. The pressing plate 26 part includes two left and right groups of pressing plates 26, and a through path for steel to pass through is formed between the two groups of pressing plates 26. The roller part includes several groups of pressing rollers 30 located at the same horizontal height, and each group of pressing rollers 30 includes two pressing rollers 30 that are parallel and rotatably arranged up and down.

[0032] The pressing plate 26 includes two groups of front slides 23 fixedly connected to the main frame 22, and a front slider 24 slidably connected to the front slide 23. A front threaded hole is provided at the upper end of the front slide 23, and a front guide bolt 25 rotatably connected to the front slider 24 is threadedly connected in the front threaded hole. The pressing plate 26 is fixedly connected to the front slider 24. The roller part includes several groups of rear slides 27 and a rear slider 28 slidably connected in the rear slide 27. A rear threaded hole is provided at the upper end of the rear slide 27, and a rear guide bolt 29 rotatably connected to the rear slider 28 is threadedly connected in the rear threaded hole. The pressure roller 30 is rotatably connected between the same group of rear sliders 28.

[0033] In the actual working process of the present invention, the steel on the unwinding group can be unwound by the unwinding shaft. After being guided by the guide roller 4 on the guide frame 3, the steel enters the detection mark group. The steel passes between the upper frame 9 and the lower frame 10. The grid laser 31 projects a grid-shaped laser on the top and bottom surfaces of the steel, and then is photographed by the image acquisition module 32 and transmitted to the data processing module 33. The data processing module 33 detects the area with uneven thickness on the surface of the steel and transmits the result to the control module 34. The control module 34 can subsequently control the servo motor 16 and the air pump 35.

[0034] When the steel flows to the position of the short gantry 11 and the high gantry 12, the servo motor 16 can be used to drive the transmission gear to rotate, and the transmission gear is meshed with the rack 14, so that the left and right position of the marker 21 can be adjusted. At the same time, the air pump 35 works to add air to the cylinder 18, and the cylinder 18 moves downward through the slide 19 and the marker 21, so that the marker 21 can draw the contour of the uneven thickness position on the top and bottom surfaces of the steel to mark it. After the steel enters the transmission again, the steel is continuously pulled so that the steel can be continuously unwound and tested. In this way, the thickness of the final steel surface can be detected and marked.

[0035] Preferably, the data processing module 33 identifies the location of the pit on the steel surface according to the image and outputs its contour and transmits it to the control module 34, including the following steps: Generate a laser projection pattern that is a superposition of a main grid and an adaptively activated auxiliary grid based on the historical defect probability distribution and real-time image gradient; Through anisotropic diffusion filtering and non-uniform illumination correction; The phase correlation method is used to extract the sub-pixel coordinates of grid intersections, and sub-pixel positioning is achieved through topological constraint optimization; Based on the elastic registration model, the displacement fields of the top and bottom images are fused and the three-dimensional deformation field is calculated by combining the stereo vision back projection. The thickness field is constructed by coupling the three-dimensional deformation field with the material mechanical parameters, and the abnormal area is detected by wavelet multiscale analysis; Level set evolution is used to extract continuous closed contours, and pseudo-defect areas are eliminated through topological invariance verification.

[0036] Through the above method steps, dynamic grid technology can be combined with multi-physics field coupling modeling to achieve high-precision detection of uneven steel thickness.

[0037] By generating a laser pattern with superimposed main and auxiliary grids based on the historical defect probability distribution and real-time image gradient, the resolution of the detection area is adaptively adjusted, which not only ensures the fine detection of high-frequency defect areas, but also takes into account the overall efficiency.

[0038] Image preprocessing uses anisotropic diffusion filtering and non-uniform illumination correction to effectively enhance grid features and eliminate environmental interference, thereby improving the accuracy of subsequent feature extraction. Sub-pixel grid positioning uses the phase correlation method to extract intersection coordinates, and achieves high-precision positioning through topological constraint optimization to ensure the reliability of displacement field calculation.

[0039] The multimodal displacement field calculation is based on the elastic registration model to fuse the displacement fields of the top and bottom images, and combines the stereoscopic vision back projection to calculate the three-dimensional deformation field, which overcomes the limitations of rigid body matching and can handle the nonlinear deformation of materials. The thickness field modeling couples the three-dimensional deformation field with the material mechanical parameters, uses wavelet multi-scale analysis to detect abnormal areas, and captures micro and macro defects at the same time.

[0040] The contour extraction adopts the level set evolution method to generate smooth closed contours, and removes pseudo defects through topological verification, avoiding the fragmentation problem of traditional edge detection.

[0041] Preferably, generating a laser projection pattern of a main grid superimposed with an adaptively activated auxiliary grid according to the historical defect probability distribution and the real-time image gradient comprises the following steps: Introduce probability density function to control grid distribution, and encrypt grid in high-frequency defect areas ,in, is the defect probability distribution of historical data statistics, ϵ is a small constant to prevent division by zero, is the basic grid density, which is the default density value of the entire grid system; Superposition of main grid and auxiliary grid is adopted to improve sensitivity through frequency domain fusion , where the auxiliary grid is at the gradient amplitude The regional activation, is the final multi-resolution grid, The main mesh for the base resolution, is a high-resolution auxiliary grid. For frequency domain fusion operation, the two grids are combined in the frequency domain. is the image gradient amplitude, which is used to determine the edge strength of the region. Gradient magnitude threshold. When the gradient magnitude exceeds this threshold, the auxiliary grid is activated.

[0042] Through the above method steps, a probability distribution model is constructed based on historical defect statistics, and the image gradient amplitude is calculated in real time to identify potential deformation areas. The main grid is projected according to the reference density, and the auxiliary grid is activated in the satisfying area. The dual-grid frequency domain fusion is realized through Fourier transform. The main grid frequency band covers 0-50Hz, and the auxiliary grid is extended to 200Hz to ensure that high-frequency defect features are sampled without aliasing. By adaptively adjusting the grid density, fine detection is achieved in high-incidence areas of defects, significantly reducing the missed detection rate of small defects (<5%), while avoiding the waste of computing resources caused by global high-density grids. The main-auxiliary grid superposition structure increases the detection sensitivity by 2-3 times in key areas, and the overall system response speed remains at the millisecond level.

[0043] Preferably, anisotropic diffusion filtering and non-uniform illumination correction are performed, including the following steps: Preserve edges while suppressing noise, based on partial differential equations , where K is the edge-sensitive parameter, which is solved iteratively until convergence, I is the image grayscale value, which is the object of filtering, and t is the iteration time variable, which is used to control the degree of filtering process. is a gradient operator, which is used to calculate the gradient information of the image. is the diffusion function, which controls the intensity of diffusion and depends on the gradient amplitude; Estimating the background illumination field And normalized ,in, is the Gaussian kernel, Much larger than the grid spacing, is the gray value of the original image, is the local mean, which is used to estimate the changing trend of background illumination. It is the background illumination field, which reflects the overall illumination distribution.

[0044] In the above method steps, anisotropic diffusion filtering is based on partial differential equations, which can retain strong edges (grid lines) and smooth weak noise by adjusting adaptively. Non-uniform illumination correction uses Gaussian kernel to estimate the background field and normalizes to eliminate low-frequency illumination components. Finally, the grid line contrast is enhanced by local histogram equalization (window size), eliminating interference such as uneven illumination and surface reflection, and the signal-to-noise ratio (SNR) is increased to more than 40dB. The grid line contrast is enhanced by 3-5 times, providing high-quality input for sub-pixel positioning.

[0045] Preferably, the phase correlation method is used to extract the sub-pixel coordinates of the grid intersections, and the sub-pixel positioning is achieved through topological constraint optimization, including the following steps: Perform Fourier phase analysis on each grid intersection neighborhood Ω: ,in, and is the frequency domain phase gradient, and is the neighborhood size, and is the sub-pixel offset in the x and y directions, used to accurately locate grid intersections; Construct a grid topology energy function to force the intersection points to conform to a regular arrangement , the intersection coordinates are optimized by nonlinear least squares method .

[0046] Through the above method steps, the phase correlation method performs FFT transformation on each intersection neighborhood, the sub-pixel offset is calculated by the frequency domain phase gradient, the topological constraint energy function forces the orthogonality of the grid rows and columns, and the Levenberg-Marquardt algorithm is minimized to improve the positioning accuracy of the grid intersections. Topological optimization reduces the coordinate error of the grid distortion area by 60%, supporting micron-level thickness resolution.

[0047] Preferably, the top and bottom surface image displacement fields are fused based on the elastic registration model, and the three-dimensional deformation field is calculated by combining stereoscopic vision back projection, which includes the following steps: Defining the displacement field , minimize the energy function The variational method is used to solve the Euler-Lagrange equation to obtain a smooth and accurate displacement field. and are the image grayscale values ​​of the top and bottom perspectives, respectively, x is the position coordinate in the image, Ω is the integration area, covering the entire image or the region of interest, is a regularization parameter used to balance the weights of similarity terms and smoothness terms to prevent overfitting; Combine the top and bottom camera parameters to calculate the 3D deformation through stereo matching ,in, is the back-projection operator, and is the internal parameter, which includes information such as focal length and principal point coordinates. R and t are external parameters, which are the rotation matrix and translation vector of the top camera relative to the bottom camera, describing the relative position and posture of the two, and outputting the three-dimensional coordinates. , and are the corresponding point coordinates in the images taken by the top and bottom cameras.

[0048] In the above method steps, the displacement field calculation accuracy is high, the elastic registration model reduces the nonlinear deformation matching error by 45%, and the three-dimensional reconstruction error is <0.1%.

[0049] Preferably, the thickness field is constructed by coupling the three-dimensional deformation field with the material mechanical parameters, and the abnormal area is detected by wavelet multi-scale analysis, including the following steps: thickness Joint modeling of three-dimensional deformation field and material elastic modulus E ,in, is the stress tensor, v is the Poisson's ratio, which needs to be solved iteratively through finite element analysis. is the initial thickness, the thickness of the material when it is not deformed, is the normal stress in the vertical direction, which is directly related to the thickness change. E is the elastic modulus of the material, which reflects the material's ability to resist deformation. and is the normal stress component in the horizontal direction; For thickness field Perform a two-dimensional wavelet decomposition: , detect abnormal energy concentration areas at multiple scales j, and the threshold function is: ,in, and is the mean and standard deviation of the wavelet coefficients of scale j, is the wavelet coefficient, the decomposition result at scale j, position k, and direction i, is the wavelet basis function, the basis function at scale j, position k, and direction i, is the inner product operation, which is used to calculate the similarity between the signal and the basis function. i is the direction of wavelet decomposition, including horizontal-vertical, vertical-horizontal, and diagonal directions. j is the scale of decomposition, from 1 to the maximum scale. Different scales correspond to different resolutions. is the anomaly detection threshold at scale j, which is used to determine whether the wavelet coefficients are abnormal. is the mean value of the wavelet coefficients at scale j, reflecting the normal level at this scale, is the standard deviation of the wavelet coefficients at scale j, which measures the degree of dispersion of the coefficients. k is the threshold coefficient, which determines the degree to which the threshold deviates from the mean.

[0050] In the above method steps, the thickness-stress relationship is established based on Hooke's law, the high-frequency sub-band (HL / LH / HH) detects local mutations, the low-frequency sub-band (LL) identifies slowly varying defects, and the multi-scale results are fused to generate a thermal map of the abnormal area.

[0051] Preferably, the continuous closed contour is extracted by level set evolution, and the pseudo defect area is eliminated by topological invariance verification, which includes the following steps: Define the signed distance function , extracting contours through curvature-driven evolution: ,in, Controls the smoothness of the contour, is the thickness deviation weight, t is the evolution time variable, which controls the progress of contour extraction. is the Dirac delta function, used to concentrate calculations near the contour, is the divergence operator, is the thickness value, is the initial thickness; Analyzing the topological rationality of contours using Betti numbers , eliminating the false detection areas with holes or non-simply connected areas.

[0052] Through the above method steps, the narrow band method is used to accelerate the calculation, and the zero level set is extracted after 50 iterations to generate a smooth closed contour and improve the pseudo defect removal rate.

[0053] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An alloy steel thickness detection and marking system, characterized in that: It includes an unwinding group, a detection and marking group, and a transmission group. The unwinding group unwinds the steel through the detection and marking group to the transmission group. The transmission group pulls the steel to unwind. The detection and marking group includes a detection part and a marking part. The detection part releases grid-shaped laser to the top and bottom surfaces of the steel and collects images to the marking part. The marking part recognizes the image and marks the contour at the uneven thickness of the top and bottom surfaces of the steel.

2. The alloy steel thickness detection marking system according to claim 1, characterized in that: The unwinding group comprises an unwinding frame (1), an unwinding roller (2) rotatably connected to the unwinding frame (1), a guide frame (3) fixedly connected to the unwinding frame (1) and having a horizontal height greater than that of the unwinding roller (2), and a guide roller (4) rotatably connected to the guide frame (3).

3. The alloy steel thickness detection marking system according to claim 1, characterized in that: The detection marking set further comprises an intermediate frame (5), and the detection part and the marking part are arranged on the intermediate frame (5).

4. The alloy steel thickness detection marking system according to claim 3 is characterized in that: The detection unit comprises a vertical frame (8), an upper frame (9) and a lower frame (10) fixedly connected to the vertical frame (8), and two groups of grid lasers (31) and image acquisition modules (32) respectively arranged on the upper frame (9) and the lower frame (10), wherein a channel through which steel can pass is formed between the upper frame (9) and the lower frame (10), and the two groups of grid lasers (31) and image acquisition modules (32) are both arranged on the side of the upper frame (9) and the lower frame (10) facing each other, and the image acquisition module (32) is used to acquire an image of the grid laser attached to the steel.

5. The alloy steel thickness detection marking system according to claim 4, characterized in that: The detection unit further comprises a data processing module (33) and a control module (34); the output end of the image acquisition module (32) is coupled to the input end of the data processing module (33); the output end of the data processing module (33) is coupled to the control module (34); the data processing module (33) identifies the position of the pit on the surface of the steel material according to the image and outputs its contour and transmits it to the control module (34).

6. The alloy steel thickness detection marking system according to claim 5, characterized in that: The marking part comprises a short gantry (11) and a high gantry (12), and the short gantry (11) and the high gantry (12) are both provided with two sets of picture components that can be horizontally displaced, the picture components located on the short gantry (11) are vertically arranged upwards, and the picture components located on the high gantry (12) are vertically arranged downwards, and the short gantry (11) and the high gantry (12) are both fixedly connected with guide rails (13) and racks (14), and the picture components comprise a mounting frame (15) sleeved on the outside of the short gantry (11) and the high gantry (12), a fixing frame (15) and a fixing frame (16) arranged on the outside of the short gantry (11) and the high gantry (12). A servo motor (16) connected to a mounting frame (15), a cylinder (18) and a mounting seat (17) fixedly connected to the mounting frame (15), a slide seat (19) slidably connected in the mounting seat (17) and fixedly connected to the movable end of the cylinder (18), and a marker pen (21) fixedly connected to the slide seat (19) via a connecting seat (20), wherein a transmission gear meshing with a rack (14) is fixedly connected to the rotor shaft of the servo motor (16), the cylinder (18) is connected to an air pump (35), and the air pump (35) and the servo motor (16) are both coupled to a control module (34).

7. The alloy steel thickness detection marking system according to claim 1, characterized in that: The transmission group comprises a main frame (22), a flattening portion fixedly connected to the main frame (22), and a roller portion fixedly connected to the main frame (22); the pressing plate (26) portion comprises two left and right groups of pressing plates (26); a through path for steel to pass through is formed between the two groups of pressing plates (26); the roller portion comprises a plurality of groups of pressing rollers (30) located at the same horizontal height; each group of pressing rollers (30) comprises two pressing rollers (30) which are parallel and rotatably arranged up and down.

8. The alloy steel thickness detection marking system according to claim 7, characterized in that: The pressing plate (26) portion comprises two groups of front slides (23) fixedly connected to the main frame (22), and a front slider (24) slidably connected to the front slide (23); a front threaded hole is formed at the upper end of the front slide (23), and a front guide bolt (25) rotatably connected to the front slider (24) is threadedly connected in the front threaded hole; the pressing plate (26) is fixedly connected to the front slider (24); the roller portion comprises a plurality of groups of rear slides (27), and a rear slider (28) slidably connected to the rear slide (27); a rear threaded hole is formed at the upper end of the rear slide (27), and a rear guide bolt (29) rotatably connected to the rear slider (28) is threadedly connected in the rear threaded hole; the pressing roller (30) is rotatably connected between the same group of rear sliders (28).

9. The alloy steel thickness detection marking system according to claim 5, characterized in that: The data processing module (33) identifies the location of the pit on the surface of the steel material according to the image and outputs its contour and transmits it to the control module (34), comprising the following steps: Generate a laser projection pattern that is a superposition of a main grid and an adaptively activated auxiliary grid based on the historical defect probability distribution and real-time image gradient; Through anisotropic diffusion filtering and non-uniform illumination correction; The phase correlation method is used to extract the sub-pixel coordinates of grid intersections, and sub-pixel positioning is achieved through topological constraint optimization; Based on the elastic registration model, the displacement fields of the top and bottom images are fused and the three-dimensional deformation field is calculated by combining stereo vision back projection. The thickness field is constructed by coupling the three-dimensional deformation field with the material mechanical parameters, and the abnormal area is detected by wavelet multiscale analysis; Level set evolution is used to extract continuous closed contours, and pseudo-defect areas are eliminated through topological invariance verification.

Images