A sheet and strip shape detection device and detection method

Abstract

A sheet / strip shape detection device and method are disclosed. The device includes a laser, a monocular high-speed area array camera, a laser velocimeter, a cooling device, a display, a data processing server, a PLC controller, a laser bracket, a camera bracket, and a through-beam metal detector. The device projects four sets of lasers onto the center of the upper surface of the sheet / strip, and acquires laser images of the upper surface of the sheet / strip using the four sets of monocular area array cameras to complete the optical plane calibration. Continuous imaging is performed while the sheet / strip is in motion, and all heights along the laser lines are extracted from each frame. Subsequently, the extracted lateral height values ​​of each frame are compensated to obtain the actual height of the entire sheet / strip surface. Finally, the flatness and maximum height of the sheet / strip are calculated based on the actual height values ​​of the sheet / strip surface and displayed. This device can effectively eliminate the effects of runout and rotation in sheet / strip flatness detection, has high measurement accuracy, is easy to install and maintain, and is beneficial for quality inspection and judgment of sheet / strip materials and subsequent model feedback control.

Description

Technical Field

[0001] This invention relates to the field of metallurgical testing technology, and in particular to a plate and strip shape testing device and testing method. Background Technology

[0002] During the rolling process of sheet and strip steel, the inconsistency between the incoming strip cross-sectional profile and the actual bearing roll gap shape leads to uneven reduction in the width direction of the strip during rolling. This results in uneven internal stress between the longitudinal fibers of the strip. When the internal stress exceeds the critical yield value, the strip undergoes buckling deformation, affecting the slab quality and potentially causing strip breakage, reduced operating speed, and equipment damage. Therefore, to evaluate, control, and improve the shape of sheet and strip steel, it is essential to establish a corresponding sheet and strip shape detection system.

[0003] In recent years, with the development of machine vision technology, it has been gradually explored and applied in industrial scenarios, showing outstanding advantages such as fast response speed in industrial measurement. Applying machine vision theory to plate shape detection has become an inevitable trend.

[0004] Currently, in the field of machine vision-based strip and sheet material inspection, the proposed multi-line laser-based strip shape detection technology is an improved solution, but this method still has some problems. First, the position of the multi-line laser is fixed, and the resolution in the length direction cannot be dynamically adjusted, which cannot meet the quality inspection and discrimination requirements for steel grades with different rolling quality requirements. In addition, although the multi-line laser-based strip shape detection method can suppress the influence of strip runout on flatness measurement to a certain extent, it cannot cope with the influence of strip deflection. Finally, from the perspective of image processing, the extraction of bright lines per frame of multi-line laser images is large and the extraction time is long, which also has a significant impact on the real-time performance of industrial inspection. Summary of the Invention

[0005] This invention provides a strip and sheet material shape detection device and method to solve the technical problems of existing technologies, such as difficulty in eliminating the influence of strip deflection on flatness detection, large amount of single-frame laser image extraction, and long extraction time.

[0006] To solve the above-mentioned technical problems, the present invention provides a sheet / strip shape detection device and method, comprising:

[0007] The device includes a laser, a monocular high-speed area array camera, a laser velocimeter, cooling equipment, a display, a data processing server, a PLC controller, a laser bracket, a camera bracket, and a through-beam metal detector. The camera bracket is installed above the roller conveyor at the mill exit. The monocular high-speed area array camera is installed inside the camera bracket, tilting downwards to capture images of the roller conveyor. The acquired image data is transmitted to the data processing server via gigabit Ethernet. The laser bracket is installed inside the camera bracket, above the roller conveyor at the mill exit. The laser is mounted on the laser bracket and projects four sets of laser beams vertically downwards onto the roller conveyor plane. The display is connected to the data processing server, which is connected to the PLC via gigabit Ethernet. The PLC controller is connected to the laser, monocular high-speed area array camera, and through-beam metal detector via signal lines. It controls the switching on / off state of the laser, the trigger switch of the monocular high-speed area array camera, and the on / off state of the through-beam metal detector. The laser consists of 4 groups: laser A, laser B, laser C, and laser D. The monocular high-speed area array camera consists of 4 groups: area array camera E, area array camera F, area array camera G, and area array camera H. The through-beam metal detector consists of 2 groups: through-beam metal detector P and through-beam metal detector Q. The monocular area array camera is connected to cooling device I, and lasers A, B, C, and D are connected to cooling device J.

[0008] The camera support includes a base, housing, lifting platform, camera pose adjustment device, camera cooling device, and dust removal device. The base is installed on the ground with anchor bolts, the housing is installed on the base with bolts, the lifting platform is installed on top of the base with bolts, the camera pose adjustment device is installed on top of the lifting platform with bolts, the camera cooling device is installed on the camera pose adjustment device, and the remaining parts of the area scan camera, except for the lens, are installed inside the camera cooling device. The dust removal part of the dust removal device is installed on the housing window and is responsible for cleaning dust and moisture on the housing window. The remaining parts are installed inside the housing.

[0009] The laser support includes a rectangular frame, a laser pose adjustment device, and a laser cooling device. The rectangular frame is installed on the ground with anchor bolts. The laser pose adjustment device is installed on top of the rectangular frame. The laser is installed on the laser pose adjustment device. The laser cooling device is installed on the outside of the laser.

[0010] A method for detecting the shape of a strip using the device described above includes the following steps:

[0011] S0: After the detection device is installed, the four sets of stereo imaging systems are jointly calibrated and the world coordinate system of the four systems is unified, with X being the width direction of the strip, Y being the movement direction of the strip, and Z being the height direction.

[0012] S1: When the strip reaches position P1, the PLC controller activates the corresponding four lasers. Laser A emits a single laser line along the width of the strip, covering the entire width of the strip, denoted as L1. Lasers B, C, and D emit three parallel laser lines along the length of the strip, denoted as L1, L2, L3, L4, L5, L6, L7, L8, L9, L10, L11, L12, L13, L14, L15, L16, L17, L18, L19 ... l L c and L r The points are located at the center of the laser line L1 and its symmetrical points to the left and right, respectively, and the intersection points are denoted as CRO. l (X l ,Y l Z l CRO c (X c ,Y c Z c ) and CRO r (X r ,Y r Z r );

[0013] S2: When the strip reaches position P2, the PLC controller triggers the corresponding monocular high-speed area array camera to continuously and synchronously capture 4 sets of laser images L1, L2, L3, L4 through a pulse signal. l L c and L r Images on sheet and strip materials;

[0014] S3: Calculate the translation distance s using the strip speed v and the shooting frame rate f;

[0015] S4: For the acquired image, extract the world coordinates of all points along the horizontal laser line L1, and then extract the world coordinates of the points along the three vertical laser lines L1. l L c and L r Extract the world coordinates of points along the entire line, resulting in a total of 4 sets of world coordinates;

[0016] S5: According to L l L c and L r Extract the world coordinates of the obtained points and calculate the vibration generated by the strip. and rotation θ final Compensate for the height value of the entire transverse laser measurement, and perform iterative calculations on each acquired image until the entire slab leaves position P2, thereby obtaining the actual height of the entire slab's upper surface.

[0017] S6: Divide the strip into N sampling fibers along the width direction. Calculate the straightness, number of waves, minimum and maximum wave values ​​and positions based on the actual height of each fiber. Display the real-time strip 3D topography, real-time 2D contour map, real-time 2D defect mapping, dynamic cross-sectional elongation distribution bar chart, dynamic cross-sectional profile, dynamic symmetrical and asymmetrical straightness parameter bar chart, and the current steel plate topographic profile along its length.

[0018] Furthermore, the formula for calculating the translation distance in S3 using the strip speed and the shooting frame rate is as follows:

[0019]

[0020] Where s represents the single translation distance; v represents the forward movement speed of the strip; and f represents the camera frame rate.

[0021] Furthermore, in S5, according to L l L c and L r The process of extracting the world coordinates of the points and compensating for the height value of the entire transverse laser measurement includes:

[0022] According to L l L c and L r Extract the world coordinates of the points, and select the point cloud data P of the overlapping measurement segment in the laser images at time (t-1) and time t. (t-1)测 and P t测 ; Calculate point cloud data P (t-1)测 Partial point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer ; Traverse rotation angle θ t ∈(-90°, 90°), the point cloud data P t测 Partial point cloud data Around midpoint Rotation θ t ,get Rotated point cloud data With ideal point cloud data P refer The height difference between corresponding points is defined as a set ΔZ(θ) t ), to find the optimal Make the set ΔZ(θ) t The variance is minimized based on the rotated point cloud data. The Z-coordinate and point cloud data P refer The Z-coordinate is used to calculate the single vibration amount; based on the rotation amount and cumulative vibration amount, coupling compensation is performed on the points extracted by L1 in each frame of the image to obtain... in,

[0023] Calculate L in the laser images at time (t-1) and time t. l Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and The formula is:

[0024]

[0025] in, Point cloud data within a specific y-axis range is selected for time (t-1); Y is the y-axis coordinate of the point cloud data at time (t-1); l For L l CRO, the intersection point with the left side of L1 l The y-coordinate; d is the half-length of the selected overlapping measurement segment in the y-direction; Select point cloud data within a specific y-axis range at time t; t represents the y-axis coordinate of the point cloud data at time t; s represents the single translation distance.

[0026] according to and The calculation method calculates L in the laser images at time (t-1) and time t. r Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and and L c Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and

[0027] Computing point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer The formula is:

[0028]

[0029] in, For laser line L l Partial point cloud data of the overlapping measurement segment at time (t-1); P refer For ideal point cloud data; s is the single translation distance of the steel plate along the Y-axis;

[0030] Calculate time t The formula for the coordinates of the center point is:

[0031]

[0032] Where m is The center point index; num is... The number of points in; for The center point, The center point x, y, and z coordinates;

[0033] Computational point cloud Around the center point Rotated point cloud data The z-coordinate of the i-th point The formula is:

[0034]

[0035] Where, θ lt for Around each point The angle of rotation ranges from -90° to 90°. and Center of rotation The y and z coordinates; and For point cloud data The x, y, and z coordinates of the i-th point in the array; and For point cloud data The x, y, and z coordinates of the i-th point before rotation;

[0036] Define the rotated point cloud data With ideal point cloud data P refer The set of height differences between corresponding points ΔZ(θ) t )for:

[0037]

[0038] in, for The z-coordinate of the i-th point in the equation; Z(P) refer ) i For P refer The z-coordinate of the i-th point in the array;

[0039] The formula for calculating the difference in Z coordinates for each point is:

[0040]

[0041] Calculate the variance Var(ΔZ(θ) tThe formula for )) is:

[0042]

[0043] Where num represents the point cloud. The number of points in; ΔZ(θ) t ) i The height difference is the height difference of the i-th point; Set of height differences The mean;

[0044] calculate Optimal rotation angle of all points The formula is:

[0045]

[0046] ΔZ(θ t (This refers to the rotated point cloud data) With ideal point cloud data P refer The set of height differences between corresponding points;

[0047] according to The calculation method yields... and rotation angle and

[0048] Calculate the rotation θ of the steel plate from time (t-1) to time t. final The formula is:

[0049]

[0050] Where, θ final Let be the amount of rotation of the steel plate from time (t-1) to time t; At time t The rotation angle of all points; For time t The rotation angle of all points; At time t The rotation angle of all points;

[0051] Based on the rotated point cloud data The corresponding rotation angle Calculate the translation Δz l The formula is:

[0052]

[0053] Where num represents the point cloud. The number of midpoints; For rotation Post-point cloud data The height difference of the i-th point;

[0054] According to Δz l The calculation method yields... and Translation amount Δz r and Δz c ;

[0055] Calculate the vibration compensation amount of laser line L1 at time t. The formula is:

[0056]

[0057] in, This is the compensation amount of the steel plate along the z-axis; Δz l At time t The translation of all points; Δz r At time t The translation of all points; Δz c At time t The translation of all points;

[0058] The formula for calculating the coordinates of the extracted points on L1 after calculating the rotational compensation and coupled vibration compensation is as follows:

[0059]

[0060] Where, θ final The rotation angle; and Center of rotation The y and z coordinates; For high compensation amount; and The x, y, and z coordinates of each point extracted on laser line L1 before rotation; and The x, y, and z coordinates of the i-th point extracted on laser line L1 after rotational coupling vibration compensation.

[0061] For the coordinates at time (t+1) and thereafter, compensation is performed using the same method for calculating vibration and rotational coupling compensation, and the calculation is repeated until the coordinate data of the last time step is obtained.

[0062] Furthermore, the number of sampling fibers N divided along the width direction of the strip in S6 ranges from 3 to 50.

[0063] Furthermore, in step S6, the formulas for calculating the flatness and maximum height of the strip based on the actual height value of the strip surface are as follows:

[0064]

[0065] L ref =min(L n )

[0066] ΔL n =L n -L ref

[0067]

[0068] in, This represents the length of the segment on the strip fiber from time (t-1) to time t; This represents the actual height of the nth measurement point in the transverse direction of the strip at time t; The value represents the actual height of the nth measurement point in the transverse direction at time (t-1); s represents the single translation distance; L n L represents the cumulative length of the nth fiber; ref ΔL represents the minimum cumulative length of n fibers. n Indicates the elongation difference of the nth fiber; F represents the flatness of the sheet / strip; height max Maximum height; This represents the actual height of the point on the strip.

[0069] Furthermore, the display of the real-time three-dimensional topography of the strip, the real-time two-dimensional contour map, the real-time two-dimensional defect mapping map, the dynamic cross-sectional profile, and the terrain profile of the current steel plate along its length in S6 depends on the actual height of the strip and the deflection θ of each frame, including:

[0070] The positions of points in each display image are reconstructed using each frame. To calculate.

[0071] The beneficial effects of the technical solution provided by this invention include at least the following:

[0072] The technical solution of this invention is based on a 4-group laser-monocular array camera imaging system. It acquires laser images of the upper surface of the strip to complete optical plane calibration. Then, it continuously captures images while the strip is moving, extracting all heights along the laser line from each frame. The extracted lateral height values ​​are compensated to obtain the actual height of the entire strip surface. Finally, based on the actual height values ​​of the strip surface, it calculates and displays the flatness and maximum height of the strip. This invention has the advantage of eliminating the influence of vibration and deflection on visual measurement during strip movement, achieving real-time online high-precision detection of the strip shape. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] Figure 1 This is a schematic diagram of the detection device provided in an embodiment of the present invention;

[0075] Figure 2 This is a schematic diagram of the camera bracket provided in an embodiment of the present invention;

[0076] Figure 3 This is a schematic diagram of the structure of the laser bracket provided in an embodiment of the present invention;

[0077] Among them, 1-laser, 1.1-laser A, 1.2-laser B, 1.3-laser C, 1.4-laser D, 2-monocular high-speed area array camera, 2.1-area array camera E, 2.2-area array camera F, 2.3-area array camera G, 2.4-area array camera H, 3-laser velocimeter, 4-cooling equipment, 5-display, 6-data processing server, 7-PLC controller, 8-laser bracket, 8.1-base, 8.2-shell, 8.3-lifting platform, 8.4-camera pose adjustment device, 8.5-camera cooling device, 8.6-dust removal device, 9-camera bracket, 9.1-rectangular frame, 9.2-laser pose adjustment device, 9.3-laser cooling device, 10-through-beam metal detector, 10.1-through-beam metal detector P, 10.2-through-beam metal detector Q, 4.1-cooling equipment I, 4.2-cooling equipment J. Detailed Implementation

[0078] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0079] First, it should be noted that in the embodiments of the present invention, the words "exemplarily," "for example," etc., are used to indicate that they are examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the term "exemplarily" is intended to present the concept in a specific manner. Furthermore, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either one or the other.

[0080] This invention provides a sheet and strip shape detection device and method.

[0081] like Figure 1As shown, the device includes a laser 1, a monocular high-speed area array camera 2, a laser velocimeter 3, a cooling device 4, a display 5, a data processing server 6, a PLC controller 7, a laser bracket, a camera bracket, and a through-beam metal detector. The camera bracket is installed above the roller conveyor at the mill exit. The monocular high-speed area array camera is installed inside the camera bracket, tilting downwards to capture images of the roller conveyor. The acquired image data is transmitted to the data processing server 6 via gigabit Ethernet. The laser bracket is installed inside the camera bracket, above the roller conveyor at the mill exit. The laser 1 is mounted on the laser bracket, projecting four sets of lasers vertically downwards onto the roller conveyor plane. The display 5 is connected to the data processing server 6, which is connected to the PLC controller 7 via gigabit Ethernet. The PLC controller 7 is connected to the laser velocimeter via signal lines. The laser 1, monocular high-speed area array camera 2, and through-beam metal detector are connected to control the switching on / off state of the laser 1, the trigger switch of the monocular high-speed area array camera 2, and the on / off state of the through-beam metal detector, respectively. The laser comprises four groups: laser A1.1, laser B1.2, laser C1.3, and laser D1.4; the monocular high-speed area array camera comprises four groups: area array camera E2.1, area array camera F2.2, area array camera G2.3, and area array camera H2.4; and the through-beam metal detector comprises two groups: through-beam metal detector P10.1 and through-beam metal detector Q10.2. The monocular area array camera is connected to cooling device I4.1, and lasers A1.1, B1.2, C1.3, and D1.4 are connected to cooling device J4.2.

[0082] like Figure 2 As shown, the camera support includes a base 8.1, a housing 8.2, a lifting platform 8.3, a camera pose adjustment device 8.4, a camera cooling device 8.5, and a dust removal device 8.6. The base 8.1 is mounted on the ground with anchor bolts, the housing 8.2 is mounted on the base 8.1 with bolts, the lifting platform 8.3 is mounted on top of the base 8.1 with bolts, the camera pose adjustment device 8.4 is mounted on top of the lifting platform 8.3 with bolts, and the camera cooling device 8.5 is mounted on the camera pose adjustment device 8.4. Except for the lens, the rest of the area scan camera is installed inside the camera cooling device 8.5. The dust removal part of the dust removal device 8.6 is installed on the housing window and is responsible for cleaning dust and moisture on the housing window. The remaining parts are installed inside the housing 8.2.

[0083] like Figure 3 As shown, the laser support includes a rectangular frame 9.1, a laser pose adjustment device 9.2, and a laser cooling device 9.3. The rectangular frame 9.1 is installed on the ground by anchor bolts. The laser pose adjustment device 9.2 is installed above the rectangular frame 9.1. The laser 1 is installed on the laser pose adjustment device 9.2. The laser cooling device 9.3 is installed on the outside of the laser 1.

[0084] When the slab reaches position P1, the PLC controller turns on the corresponding laser and emits four laser beams along the length of the strip. When the slab reaches position P2, the PLC controller triggers the corresponding area array camera to continuously capture images of the laser beams on the slab via pulse signals. The capturing ends when the entire slab leaves P2, and the corresponding laser is turned off. The acquired images are transmitted to the data processing server, where the center of the light beams is extracted and coordinates are transformed to obtain the measured height of the upper surface of the strip.

[0085] The states of the through-beam metal detectors corresponding to positions P1 and P2 of the area array camera are as follows:

[0086] Position P1: Q of the through-beam metal detector is low, and P of the through-beam metal detector is rising.

[0087] Position P2: For through-beam metal detectors, P is at a high level, and Q is at a rising edge;

[0088] Among them, the through-beam metal detector P and through-beam metal detector Q are at a low level when no object passes through, and at a high level when an object blocks their path.

[0089] The value of the number of sampling fibers N is in the range of 3-50. In this embodiment, the value of the number of sampling fibers N is set to 21.

[0090] The formula for calculating the translation distance using the strip speed and the shooting frame rate is:

[0091]

[0092] Where s represents the single translation distance, in mm;

[0093] v represents the forward movement speed of the strip, in mm / s. In this embodiment, v is taken as 5000 mm / s.

[0094] f represents the camera's frame rate, measured in fps. In this embodiment, f is set to 1000fps.

[0095] According to L l L c and L r The process of extracting the world coordinates of the points and compensating for the height value of the entire transverse laser measurement includes:

[0096] According to L l L c and L r Extract the world coordinates of the points, and select the point cloud data P of the overlapping measurement segment in the laser images at time (t-1) and time t. (t-1)测 and P t测; Calculate point cloud data P (t-1)测 Partial point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer ; Traverse rotation angle θ t ∈(-90°, 90°), the point cloud data P t测 Partial point cloud data Around midpoint Rotation θ t ,get Rotated point cloud data With ideal point cloud data P refer The height difference between corresponding points is defined as a set ΔZ(θ) t ), to find the optimal Make the set ΔZ(θ) t The variance is minimized based on the rotated point cloud data. The Z-coordinate and point cloud data P refer The Z-coordinate is used to calculate the single vibration amount; based on the rotation amount and cumulative vibration amount, coupling compensation is performed on the points extracted by L1 in each frame of the image to obtain... in,

[0097] Calculate L in the laser images at time (t-1) and time t. l Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and The formula is:

[0098]

[0099] in, Point cloud data within a specific y-axis range selected at time (t-1), dimensionless;

[0100] The y-axis coordinate of the point cloud data at time (t-1) is in mm;

[0101] Y l For L l CRO, the intersection point with the left side of L1 l The y-coordinate, in mm;

[0102] d is the half length of the selected overlapping measurement segment in the y direction, in mm. In this embodiment, d is taken as 4 mm.

[0103] Point cloud data within a specific y-axis range selected at time t, dimensionless;

[0104] y-coordinate of the point cloud data at time t, in mm;

[0105] s represents the distance of a single translation, in mm.

[0106] according to and The calculation method calculates L in the laser images at time (t-1) and time t. r Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and and L c Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and

[0107] Computing point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer The formula is:

[0108]

[0109] in, For laser line L l The point cloud data of the overlapping measurement segment at time (t-1) is dimensionless;

[0110] P refer Ideal point cloud data, dimensionless;

[0111] s represents the single translation distance of the steel plate along the Y-axis, in mm.

[0112] Calculate time t The formula for the coordinates of the center point is:

[0113]

[0114] Where m is The center point index is dimensionless.

[0115] num is The number of points in the data, expressed in units of individual points;

[0116] for The center point, dimensionless;

[0117] The center point The x, y, and z coordinates are in mm.

[0118] Computational point cloud Around the center point Rotated point cloud data The z-coordinate of the i-th point The formula is:

[0119]

[0120] in, for Around each point The angle of rotation ranges from -90° to 90°.

[0121] and Center of rotation The y and z coordinates, in mm;

[0122] and For point cloud data The x, y, and z coordinates of the i-th point in the array are in mm.

[0123] and For point cloud data The coordinates of the i-th point before rotation are x, y, and z, in mm.

[0124] Define the rotated point cloud data With ideal point cloud data P refer The set of height differences between corresponding points ΔZ(θ) t )for:

[0125]

[0126] in, for The z-coordinate of the i-th point in the array, in mm;

[0127] Z(P refer ) i For P refer The z-coordinate of the i-th point in the array is in mm.

[0128] The formula for calculating the difference in Z coordinates for each point is:

[0129]

[0130] Calculate the variance Var(ΔZ(θ) t The formula for )) is:

[0131]

[0132] Where num represents the point cloud. The number of points in the data, expressed in units of individual points;

[0133] ΔZ(θ t ) i The height difference of the i-th point is expressed in mm.

[0134] Set of height differences The mean value is in mm.

[0135] calculate Optimal rotation angle of all points The formula is:

[0136]

[0137] ΔZ(θ t (This refers to the rotated point cloud data) With ideal point cloud data P refer The set of height differences between corresponding points, in mm;

[0138] according to The calculation method yields... and rotation angle and

[0139] Calculate the rotation θ of the steel plate from time (t-1) to time t. final The formula is:

[0140]

[0141] Where, θ final The rotation of the steel plate from time (t-1) to time t is expressed in rad.

[0142] At time t The rotation angle of all points, in rad;

[0143] At time t The rotation angle of all points, in rad;

[0144] At time t The rotation angle of all points, in rad.

[0145] Based on the rotated point cloud data The corresponding rotation angle Calculate the translation Δz l The formula is:

[0146]

[0147] Where num represents the point cloud. The number of midpoints, in units of individual points;

[0148] For rotation Post-point cloud data The height difference of the i-th point, in mm;

[0149] According to Δz l The calculation method yields... and Translation amount Δz r and Δz c .

[0150] Calculate the vibration compensation amount of laser line L1 at time t. The formula is:

[0151]

[0152] in, This represents the compensation amount of the steel plate along the z-axis, in mm.

[0153] Δz l At time t The translation amount of all points, in mm;

[0154] Δz r At time t The translation amount of all points, in mm;

[0155] Δz c At time t The translation of all points is expressed in mm.

[0156] The formula for calculating the coordinates of the extracted points on L1 after calculating the rotational compensation and coupled vibration compensation is as follows:

[0157]

[0158] Where, θ final The rotation angle is expressed in rad.

[0159] and Center of rotation The y and z coordinates, in mm;

[0160] This is the height compensation amount, in mm;

[0161] and The coordinates of each point extracted on laser line L1 are the x, y, and z coordinates before rotation, in mm;

[0162] and The coordinates of the i-th point extracted on laser line L1 are the x, y, and z coordinates after rotational coupling vibration compensation, in mm.

[0163] For the coordinates at time (t+1) and thereafter, compensation is performed using the same method for calculating vibration and rotational coupling compensation, and the calculation is repeated until the coordinate data of the last time step is obtained.

[0164] The formulas for calculating the straightness and maximum height of the strip based on its actual surface height are as follows:

[0165]

[0166] L ref =min(L n )

[0167] ΔL n =L n -L ref

[0168]

[0169] in, This represents the length of the section on the fiber of the strip from time (t-1) to time t, in mm;

[0170] This represents the actual height of the nth measurement point in the transverse direction of the strip at time t, in mm;

[0171] This represents the actual height of the nth measurement point in the transverse direction of the strip at time (t-1), in mm;

[0172] 's' represents the distance of a single translation, in mm.

[0173] L n This represents the cumulative length of the nth fiber, in mm.

[0174] L ref This represents the minimum cumulative length of n fibers, in mm.

[0175] ΔL n This represents the difference in elongation of the nth fiber, in mm.

[0176] F represents the flatness of the sheet / strip, measured in I-Unit;

[0177] height max This is the maximum height, in mm.

[0178] This represents the actual height of a point on the strip, in mm.

[0179] The display of real-time 3D topographic images of the strip, real-time 2D contour maps, real-time 2D defect mapping maps, dynamic cross-sectional profiles, and the current topographic profile of the steel plate along its length depends on the actual height of the strip and the deflection θ in each frame, including:

[0180] The positions of points in each display image are reconstructed using each frame. To calculate.

[0181] In summary, this embodiment provides a strip and sheet material shape detection device and method. After the measuring device is installed and its posture is adjusted, the laser image on the upper surface of the strip and sheet is acquired based on a 4-group laser-monocular array camera imaging system to complete the optical plane calibration. Then, continuous shooting is performed when the strip and sheet moves, and all heights on the laser line are extracted from each frame of the image. The extracted lateral height value of each frame is compensated to obtain the actual height of the entire strip and sheet surface. Finally, the flatness and maximum height of the strip and sheet are calculated based on the actual height value of the strip and sheet surface and displayed. This invention has the advantage of eliminating the influence of vibration and deflection on visual measurement during the movement of the strip and sheet material, and realizes real-time online high-precision detection of the strip and sheet material shape.

[0182] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention. It should be pointed out that although preferred embodiments of the present invention have been described, those skilled in the art, once they understand the basic inventive concept of the present invention, can make several improvements and modifications without departing from the principles described herein. These improvements and modifications should also be considered within the scope of protection of the present invention. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. A method for detecting the shape of sheet / strip using a sheet / strip shape detection device, characterized in that: The device includes a laser, a monocular high-speed area array camera, a laser velocimeter, cooling equipment, a display, a data processing server, a PLC controller, a laser bracket, a camera bracket, and a through-beam metal detector. The camera bracket is installed above the roller conveyor at the mill exit. The monocular high-speed area array camera is installed inside the camera bracket, tilting downwards to capture images of the roller conveyor. The acquired image data is transmitted to the data processing server via gigabit Ethernet. The laser bracket is installed inside the camera bracket, above the roller conveyor at the mill exit. The laser is mounted on the laser bracket and projects four sets of laser beams vertically downwards onto the roller conveyor plane. The display is connected to the data processing server, which is connected to the PLC via gigabit Ethernet. The PLC controller is connected to the laser, monocular high-speed area array camera, and through-beam metal detector via signal lines. It controls the switching on / off state of the laser, the trigger switch of the monocular high-speed area array camera, and the on / off state of the through-beam metal detector, respectively. The laser consists of 4 groups: laser A, laser B, laser C, and laser D. The monocular high-speed area array camera consists of 4 groups: area array camera E, area array camera F, area array camera G, and area array camera H. The through-beam metal detector consists of 2 groups: through-beam metal detector P and through-beam metal detector Q. The monocular area array camera is connected to cooling device I, and lasers A, B, C, and D are connected to cooling device J. The method includes the following steps: S0: After the detection device is installed, the four sets of stereo imaging systems are jointly calibrated. The world coordinate system of the four stereo imaging systems is set together, with X being the width direction of the strip, Y being the movement direction of the strip, and Z being the height direction. S1: When the strip reaches position P1, the PLC controller activates the corresponding four lasers. Laser A emits a single laser line along the width of the strip, covering the entire width of the strip, denoted as L1. Lasers B, C, and D emit three parallel laser lines along the length of the strip, denoted as L1, L2, L3, L4, L5, L6, L7, L8, L9, L10, L11, L12, L13, L14, L15, L16, L17, L18, L19 ... l L c and L r The points are located at the center of the laser line L1 and its symmetrical points to the left and right, respectively, and the intersection points are denoted as CRO. l (X l ,Y l Z l CRO c (X c ,Y c Z c ) and CRO r (X r ,Y r Z r ); S2: When the strip reaches position P2, the PLC controller triggers the corresponding monocular high-speed area array camera to continuously and synchronously capture 4 sets of laser images L1, L2, L3, L4 through a pulse signal. l L c and L r Images on sheet and strip materials; S3: Calculate the translation distance s using the strip speed v and the shooting frame rate f; S4: For the acquired image, extract the world coordinates of all points along the horizontal laser line L1, and then extract the world coordinates of the points along the three vertical laser lines L1. l L c and L r Extract the world coordinates of points along the entire line, resulting in a total of 4 sets of world coordinates; S5: According to L l L c and L r Extract the world coordinates of the obtained points and calculate the vibration generated by the strip. and rotation θ final Compensate for the height value of the entire transverse laser measurement, and perform iterative calculations on each acquired image until the entire slab leaves position P2, thereby obtaining the actual height of the entire upper surface of the slab. S6: Divide the strip into N sampling fibers along the width direction. Calculate the straightness, number of waves, minimum and maximum wave values ​​and positions based on the actual height of each fiber. Display the real-time strip 3D topography, real-time 2D contour map, real-time 2D defect mapping, dynamic cross-sectional elongation distribution bar chart, dynamic cross-sectional profile, dynamic symmetrical and asymmetrical straightness parameter bar chart, and the current steel plate topographic profile along its length. The states of the through-beam metal detectors corresponding to positions P1 and P2 of the area array camera are as follows: Position P1: Q of the through-beam metal detector is low, and P of the through-beam metal detector is rising. Position P2: For through-beam metal detectors, P is at a high level, and Q is at a rising edge; Among them, the through-beam metal detector P and through-beam metal detector Q are at a low level when no object passes through, and at a high level when an object blocks their path.

2. The method for detecting the shape of a strip as described in claim 1, characterized in that, The camera bracket and laser bracket are installed in the following manner: The camera bracket includes a base, a housing, a lifting platform, a camera posture adjustment device, a camera cooling device, and a dust removal device. The base is installed on the ground with anchor bolts, the housing is installed on the base with bolts, the lifting platform is installed above the base with bolts, the camera posture adjustment device is installed above the lifting platform with bolts, the camera cooling device is installed on the camera posture adjustment device, the area scan camera (excluding the lens) is installed inside the camera cooling device, and the dust removal part of the dust removal device is installed on the window of the housing, responsible for cleaning dust and moisture on the window of the housing. The remaining part of the dust removal device is installed inside the housing. The laser support includes a rectangular frame, a laser position adjustment device, and a laser cooling device. The rectangular frame is installed on the ground by anchor bolts, and the laser position adjustment device is installed above the rectangular frame. The laser is installed on the laser position adjustment device, and the laser cooling device is installed on the outside of the laser.

3. The method for detecting the shape of strip as described in claim 1, characterized in that, The formula for calculating the translation distance in S3 using the strip speed and shooting frame rate is as follows: Where s represents the single translation distance of the strip; v represents the forward movement speed of the strip; and f represents the camera frame rate.

4. The method for detecting the shape of strip material as described in claim 1, characterized in that, According to L in S5 l L c and L r The process of extracting the world coordinates of the points and compensating for the height value of the entire transverse laser measurement includes: According to L l L c and L r Extract the world coordinates of the points, and select the point cloud data P of the overlapping measurement segment in the laser images at time (t-1) and time t. (t-1)测 and P t测 ; Calculate point cloud data P (t-1)测 Partial point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer ; Traverse rotation angle θ t ∈(-90°, 90°), the point cloud data P t测 Partial point cloud data Around midpoint Rotation θ t ,get Rotated point cloud data With ideal point cloud data P refer The height difference between corresponding points is defined as a set ΔZ(θ) t ), to find the optimal Make the set ΔZ(θ) t The variance is minimized based on the rotated point cloud data. The Z-coordinate and point cloud data P refer The Z-coordinate is used to calculate the single vibration amount; based on the rotation amount and cumulative vibration amount, coupling compensation is performed on the points extracted by L1 in each frame of the image to obtain... in, Calculate L in the laser images at time (t-1) and time t. l Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and The formula is: in, Point cloud data within a specific y-axis range is selected for time (t-1); Y is the y-axis coordinate of the point cloud data at time (t-1); l For L l CRO, the intersection point with the left side of L1 l The y-coordinate; d is the half length of the selected overlapping measurement segment in the y-direction; Select point cloud data within a specific y-axis range at time t; t represents the y-axis coordinate of the point cloud data at time t; s represents the single translation distance. according to and The calculation method calculates L in the laser images at time (t-1) and time t. r Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and and L c Extracted point cloud data and Partial point cloud data of the middle overlap measurement segment and Computing point cloud data Ideal point cloud data P after moving along the Y-axis for s without vibration or deflection. refer The formula is: in, For laser line L l Partial point cloud data of the overlapping measurement segment at time (t-1); P refer s represents ideal point cloud data; s is the single translation distance of the steel plate along the Y-axis; Calculate time t The formula for the coordinates of the center point is: Where m is The center point index; num is... The number of points in; for The center point, The center point x, y, and z coordinates; Computational point cloud Around the center point Rotated point cloud data The z-coordinate of the i-th point The formula is: in, for Around each point The angle of rotation ranges from -90° to 90°. and Center of rotation The y and z coordinates; and For point cloud data The x, y, and z coordinates of the i-th point in the array; and For point cloud data The x, y, and z coordinates of the i-th point before rotation; Define the rotated point cloud data With ideal point cloud data P refer The set of height differences between corresponding points ΔZ(θ) t )for: in, for The z-coordinate of the i-th point in the equation; Z(P) refer ) i For P refer The z-coordinate of the i-th point in the array; The formula for calculating the difference in Z coordinates for each point is: Calculate the variance Var(ΔZ(θ) t The formula for )) is: Where num represents the point cloud. The number of points in; ΔZ(θ) t ) i The height difference is the height difference of the i-th point; Set of height differences The mean; calculate Rotation angle of all points The formula is: ΔZ(θ t (This refers to the rotated point cloud data) With ideal point cloud data P refer The set of height differences between corresponding points; according to The calculation method yields... and rotation angle and Calculate the rotation θ of the steel plate from time (t-1) to time t. final The formula is: Where, θ final Let be the amount of rotation of the steel plate from time (t-1) to time t; At time t The rotation angle of all points; For time t The rotation angle of all points; For time t The rotation angle of all points; Based on the rotated point cloud data The corresponding rotation angle Calculate the translation Δz l The formula is: Where num represents the point cloud. The number of midpoints; For rotation Post-point cloud data The height difference of the i-th point; According to Δz l The calculation method yields... and Translation Δz r and Δz c ; Calculate the vibration compensation amount of laser line L1 at time t. The formula is: in, This is the compensation amount of the steel plate along the z-axis; Δz l For time t The translation of all points; Δz r For time t The translation of all points; Δz c For time t The translation of all points; The formula for calculating the coordinates of the extracted points on L1 after calculating the rotational compensation and coupled vibration compensation is as follows: Where, θ final The rotation angle; and Center of rotation The y and z coordinates; For high compensation amount; and The x, y, and z coordinates of each point extracted on laser line L1 before rotation; and The x, y, and z coordinates of the i-th point extracted on laser line L1 after rotational coupling vibration compensation. For the coordinates at time (t+1) and thereafter, compensation is performed using the same method for calculating vibration and rotational coupling compensation, and the calculation is repeated until the coordinate data of the last time step is obtained.

5. The method for detecting the shape of strip as described in claim 1, characterized in that, The number of sampling fibers N divided along the width direction of the strip in S6 ranges from 3 to 50.

6. The method for detecting the shape of a strip as described in claim 1, characterized in that, The formula for calculating the straightness and maximum height of the strip based on the actual height value of the strip surface in step S6 is as follows: L ref <min(L n ) ΔL n =L n -L ref in, This represents the length of the segment on the strip fiber from time (t-1) to time t; This represents the actual height of the nth measurement point in the transverse direction of the strip at time t; The value represents the actual height of the nth measurement point in the transverse direction at time (t-1); s represents the single translation distance; L n L represents the cumulative length of the nth fiber; ref ΔL represents the minimum cumulative length of n fibers. n Indicates the elongation difference of the nth fiber; F represents the flatness of the sheet / strip; height max Maximum height; This represents the actual height of the point on the strip.

7. The method for detecting the shape of a strip as described in claim 1, characterized in that, The display of the real-time three-dimensional topography of the strip, the real-time two-dimensional contour map, the real-time two-dimensional defect mapping map, the dynamic cross-sectional profile, and the terrain profile of the current steel plate along its length in S6 depends on the actual height of the strip and the deflection θ of each frame, including: The positions of points in each display image are reconstructed using each frame. To calculate.

Images