Dynamic height compensation-based hot-rolled strip steel deviation visual detection method and dynamic height compensation-based hot-rolled strip steel deviation visual detection system

Abstract

The invention discloses a dynamic height compensation-based hot-rolled strip steel deviation visual detection method and system, and the method comprises the steps: installing an industrial camera between two groups of racks, installing a plurality of distance measuring sensors at the same height, calibrating an internal parameter matrix and an external parameter matrix of the camera, and recording the vertical distance from each distance measuring sensor to a calibration plane; in the detection stage, the vertical distance between the distance measuring sensor and the strip steel surface is obtained in real time, and the height offset and the mean value thereof are calculated; in the detection stage, an industrial camera is used for collecting strip steel images in real time, and left and right edges of the strip steel images are detected; calculating coordinates of points on the left and right edges of the strip steel in a world coordinate system through an internal parameter matrix and an external parameter matrix of a camera based on pixel coordinates and a height offset mean value on the left and right edges of the strip steel; outlier filtering and mean value calculation are carried out on the left edge point and the right edge point of the strip steel under the world coordinate system, the actual position of the edge is determined, and the deviation amount is calculated.

Description

Technical Field

[0001] The present invention belongs to the technical field of hot strip finishing control, and relates to a visual detection method and system for hot rolled strip deviation based on dynamic height compensation. Background Art

[0002] In the production process of hot rolled strip, strip deviation is a key issue affecting rolling quality and equipment safety. Due to factors such as rolling tension fluctuation, temperature gradient change, and roll wear, the strip often undergoes lateral displacement during operation between rolling mills. In severe cases, it may lead to edge scraping, steel piling accidents, or even damage to roll bearings, resulting in unplanned shutdown of the production line. Therefore, real-time and high-precision detection of strip deviation is crucial for improving rolling stability and product quality.

[0003] The Chinese invention patent application "A Method for Detecting Deviation of Hot Rolled Strip Considering the Influence of Looper" (CN116274419A) relies on the looper angle and strip thickness to obtain the transformation relationship of the world coordinate system when the strip rotates, so as to eliminate the influence of the looper on strip deviation detection. However, in actual production, this method is limited by mechanical hysteresis and cannot respond to strip transient fluctuations in real time. Moreover, due to factors such as strip thermal expansion and vibration, it cannot accurately reflect the change in strip height, resulting in errors in the coordinate transformation model.

[0004] The Chinese invention patent "Laser Scanning Detection Method and Deviation Correction System for Strip Deviation and Floating Amount" (CN102319743B) uses the shape of the line structured light beam to determine the strip edge, and calculates the up and down floating amount of the strip according to the displacement of the laser linear beam in the strip length direction. However, this patent does not consider the influence of the floating amount on the measurement of the deviation amount. Summary of the Invention

[0005] To solve the above technical problems, the purpose of the present invention is to provide a visual detection method and system for hot rolled strip deviation based on dynamic height compensation.

[0006] The present invention provides a visual detection method for hot rolled strip deviation based on dynamic height compensation, including:

[0007] Step 1: Install an industrial camera between two groups of stands and install multiple ranging sensors at the same height. Calibrate the internal parameter matrix K and external parameter matrix T = [R|t] of the camera to establish the initial mapping relationship from the pixel coordinate system to the world coordinate system; at the same time, record the vertical distance a from each ranging sensor to the calibration plane g ;

[0008] Step 2: In the detection stage, real-time obtain the vertical distance b from the ranging sensor to the strip surface g , calculate the height offset Δh g = a g - b gand its mean value ΔH;

[0009] Step 3: During the detection stage, use an industrial camera to collect strip images in real time, and perform edge detection on the left and right edges of the strip images;

[0010] Step 4: Based on the pixel coordinates of the points on the left and right edges of the strip and ΔH, through the internal parameter matrix and external parameter matrix of the camera, calculate the coordinates of the points on the left and right edges of the strip in the world coordinate system;

[0011] Step 5: Filter out outliers and calculate the mean value of the left and right edge points of the strip in the world coordinate system, determine the actual position of the edge, and calculate the deviation amount.

[0012] Further, the specific content of Step 1 is as follows:

[0013] Step 1.1: Install an industrial camera between two groups of rolling mills. The edge of the strip needs to be within the field of view of the camera, and the strip moves vertically in the camera's field of view;

[0014] Step 1.2: Install multiple ranging sensors at the same height between two groups of rolling mills. The ranging sensors form a predetermined angle with the optical axis of the industrial camera to ensure coverage of the strip detection area;

[0015] Step 1.3: Use a checkerboard calibration board. When the mill is stopped, lower the loop to the horizontal position, place the checkerboard calibration board on the upper surface of the loop, make its plane parallel to the rolling center line and coincide with the plane of the initial position of the loop. Define this plane as the calibration plane, that is, the plane where the world coordinate system Z = 0, and the positive direction of the Z-axis is vertically downward. Record the vertical distance a from each ranging sensor to the calibration plane g ;

[0016] Step 1.4: Collect multiple groups of calibration board images through the industrial camera, and calculate the internal parameter matrix of the camera by the Zhang Zhengyou calibration method:

[0017]

[0018] and the external parameter matrix, including the rotation matrix R and the translation vector t:

[0019]

[0020] where, f x 、f y are the focal length dimensions measured in pixels in the x and y directions of the camera respectively, u0 and v0 are the principal point coordinates of the image in pixels, R is a 3×3 rotation matrix, and t is a 3×1 translation vector.

[0021] Further, the specific content of Step 2 is as follows:

[0022] Step 2.1: During the detection phase, the vertical distance b from the ranging sensor to the strip surface is obtained in real time g , and the height offset Δh is calculated g = a g - b g ;

[0023] Step 2.2: Calculate the average value of the height offset ΔH:

[0024]

[0025] where G is the total number of ranging sensors.

[0026] Furthermore, in step 3, sub-pixel edge extraction algorithm is used for edge detection, and then the pixel coordinates of the left and right edges of the strip are obtained.

[0027] Furthermore, step 4 is specifically as follows:

[0028] Step 4.1: Calculate the normalized imaging plane coordinates of the left edge points:

[0029]

[0030] where u 1i , v 1i are the coordinates of the i-th point on the left edge in the pixel coordinate system; x ni , y ni are the coordinates of the i-th point on the left edge converted to the normalized imaging plane coordinate system;

[0031] The points x ni , y ni in the normalized imaging plane coordinate system satisfy:

[0032]

[0033] where X c , Y c , Z c are the coordinates of the edge points converted to the camera coordinate system;

[0034] External parameter relationship from the world coordinate system to the camera coordinate system:

[0035]

[0036] Step 4.2: Substitute X c = x ni Z c and Y c = y ni Z c into the external parameter equation to construct a linear equation system:

[0037]

[0038] where Z w = -ΔH; X wi , Y wi are the world coordinates of the i-th point on the left edge;

[0039] Step 4.3: Solve the world coordinates of the left edge points:

[0040]

[0041] Step 4.4: Use the same method to calculate the normalized imaging plane coordinate system coordinates x nj , y nj of the j-th point on the right edge, construct a system of linear equations and solve the world coordinates X wj , Y wj of the j-th point on the right edge.

[0042] Furthermore, the specific steps of step 5 are as follows:

[0043] Step 5.1: Calculate the mean and standard deviation of the world coordinates of the left edge points and the right edge points:

[0044]

[0045]

[0046] where N is the total number of left edge points; M is the total number of right edge points; μ1 and σ1 are the mean and standard deviation of the world coordinates of the left edge points respectively; μ2 and σ2 are the mean and standard deviation of the world coordinates of the right edge points respectively;

[0047] Step 5.2: Retain the coordinate points on the left edge that satisfy |X wi - μ1| < 3σ1, and retain the coordinate points on the right edge that satisfy |X wj - μ2| < 3σ2, and regard the rest as outliers and eliminate them;

[0048] Step 5.3: Repeat steps 5.1 - 5.2 until either of the following conditions is met:

[0049] 1) The standard deviations of the remaining valid points on the left side and the remaining valid points on the right side are both less than or equal to the set corresponding thresholds;

[0050] 2) The number of iterations reaches 3 times;

[0051] Step 5.4: Calculate the mean of the world coordinates of the remaining valid points as the reference positions of the left and right edges, and the formula is:

[0052]

[0053] wherein, n and m are the effective number of points on the left and right edges respectively;

[0054] Step 5.5: Calculate the deviation amount according to the reference positions of the left and right edges. The formula is:

[0055]

[0056] Furthermore, the ranging sensor is a non-contact ranging sensor, including a laser ranging sensor or an ultrasonic ranging sensor.

[0057] The present invention also provides a visual detection system for the deviation of hot-rolled strip based on dynamic height compensation, which is used to implement the above-mentioned visual detection method for the deviation of hot-rolled strip based on dynamic height compensation. The system includes: an image acquisition module, a distance detection module, a data processing module, and a protection structure;

[0058] The image acquisition module is an industrial camera set between two groups of racks. The industrial camera covers the strip area between the racks to ensure that the edges of the strip are within the field of view of the camera. During the calibration stage, it is necessary to verify the field of view range to ensure that it only covers the strip area directly below the ranging sensor;

[0059] The distance detection module includes multiple ranging sensors, which are used to measure the vertical distance from the sensor to the calibration plane during the calibration stage and the vertical distance from the sensor to multiple measurement points on the strip surface during the detection stage;

[0060] The data processing module is respectively connected to the image acquisition module and the distance detection module; the data processing module is built-in with a calibration parameter storage unit, an edge extraction unit, a first calculation unit, an outlier filtering unit, and a second calculation unit;

[0061] The edge extraction unit extracts the pixel coordinates of the left and right edges of the strip according to the acquired strip image; the first calculation unit calculates the height offset and its average value according to the vertical distance from the sensor to the calibration plane and the vertical distance from the sensor to multiple measurement points on the strip surface, and resolves the actual coordinates of the left and right edge points in the world coordinate system according to the pixel coordinates of the left and right edge points and the average value of the height offset; the outlier filtering unit is used to filter the actual coordinates of the left and right edge points in the world coordinate system to eliminate invalid points; the second calculation unit is used to calculate the average value of the world coordinates of the remaining valid points as the reference positions of the left and right edges, and calculate the deviation amount according to the reference positions of the left and right edges;

[0062] The protection structure is used to wrap the image acquisition module and the distance detection module, and has the functions of cooling and dust prevention.

[0063] Furthermore, the protection structure includes a water-cooled sandwich layer and an air curtain isolation device to adapt to the high-temperature environment of the rolling site.

[0064] Further, the data processing module communicates with the mill level 1 system through the industrial Ethernet and outputs the deviation amount to the HMI interface and the control unit of the mill level 1 system.

[0065] A visual detection method for strip deviation in hot rolling based on dynamic height compensation has the following

[0066] Beneficial effects:

[0067] This method uses a laser ranging device or an ultrasonic sensor to collect the height change data of the upper surface of the strip between the rolling mills in real time. Combining the pre-calibrated internal and external parameters of the camera, it dynamically corrects the conversion relationship from the pixel coordinate system to the world coordinate system, eliminating the error of strip deviation detection caused by strip height fluctuations. The present invention solves the problem of inaccurate detection caused by strip height fluctuations in traditional methods and is applicable to the rolling industrial environment with high temperature and high vibration. Description of the Drawings

[0068] Figure 1 is a flowchart of a visual detection method for strip deviation in hot rolling based on dynamic height compensation according to the present invention;

[0069] Figure 2 is a deployment diagram of a visual detection system for strip deviation in hot rolling based on dynamic height compensation according to the present invention. Detailed Embodiments

[0070] As Figure 1 shown, a visual detection method for strip deviation in hot rolling based on dynamic height compensation according to the present invention includes:

[0071] Step 1: Install an industrial camera between two sets of stands and install multiple ranging sensors at the same height. Calibrate the internal camera parameter matrix K and the external parameter matrix T = [R|t] to establish the initial mapping relationship from the pixel coordinate system to the world coordinate system; at the same time, record the vertical distance a from each ranging sensor to the calibration plane g , specifically:

[0072] Step 1.1: Install an industrial camera between two sets of stands. The edge of the strip needs to be within the field of view of the camera, and the strip moves vertically in the camera's field of view.

[0073] Step 1.2: Install multiple ranging sensors at the same height between two sets of stands. The ranging sensors form a predetermined angle with the optical axis of the industrial camera to ensure coverage of the strip detection area. The ranging sensors are non-contact ranging sensors, including laser ranging sensors or ultrasonic ranging sensors.

[0074] Step 1.3: Use a checkerboard calibration plate. When the rolling mill is stopped, lower the loop to the horizontal position. Place the checkerboard calibration plate on the upper surface of the loop, making its plane parallel to the rolling center line and coinciding with the plane of the initial position of the loop. Define this plane as the calibration plane, which is the plane of the world coordinate system where Z = 0. The positive direction of the Z-axis is vertically downward. Record the vertical distance a from each distance measuring sensor to the calibration plane. g 。

[0075] In the world coordinate system, the X-axis: lies in the calibration plane and is perpendicular to the rolling center line, with the positive direction pointing to the right side of the strip (observed along the strip movement direction); the Y-axis: lies in the calibration plane and is parallel to the rolling center line, with the positive direction opposite to the strip forward direction; the Z-axis: is perpendicular to the calibration plane, with the positive direction vertically downward. Calibration plane: Z = 0 plane; the rolling center line corresponds to X = 0, Z = 0.

[0076] Step 1.4: Collect multiple groups of calibration plate images with different poses through an industrial camera, and calculate the internal parameter matrix of the camera by the Zhang Zhengyou calibration method:

[0077]

[0078] and the external parameter matrix, including the rotation matrix R and the translation vector t:

[0079]

[0080] Among them, f x 、f y are the focal length dimensions measured in pixels in the x and y directions of the camera respectively, u0, v0 are the principal point coordinates of the image in pixels, R is a 3×3 rotation matrix, and t is a 3×1 translation vector.

[0081] Step 2: In the detection stage, obtain the vertical distance b from the distance measuring sensor to the strip surface in real time g , calculate the height offset Δh g = a g - b g and its mean value ΔH, specifically:

[0082] Step 2.1: In the detection stage, obtain the vertical distance b from the distance measuring sensor to the strip surface in real time g , calculate the height offset Δh g = a g - b g 。

[0083] Step 2.2: Calculate the mean value ΔH of the height offset:

[0084]

[0085] Among them, G is the total number of distance measuring sensors.

[0086] Step 3: In the detection stage, an industrial camera is used to collect strip images in real time, and edge detection is performed on the left and right edges of the strip images;

[0087] Specifically, in implementation, sub-pixel edge extraction algorithm is adopted for edge detection, and then the pixel coordinates of the left and right edges of the strip are obtained.

[0088] Step 4: Based on the pixel coordinates of the points on the left and right edges of the strip and ΔH, the actual coordinates of the points on the left and right edges of the strip in the world coordinate system are calculated through the internal parameter matrix and external parameter matrix of the camera, specifically:

[0089] Step 4.1: Calculate the normalized imaging plane coordinates of the left edge points:

[0090]

[0091] where u 1i , v 1i are the coordinates of the i-th point on the left edge in the pixel coordinate system; x ni , y ni are the coordinates of the i-th point on the left edge transformed to the normalized imaging plane coordinate system.

[0092] The points x ni , y ni in the normalized imaging plane coordinate system satisfy:

[0093]

[0094] where X c , Y c , Z c are the coordinates of the edge points transformed to the camera coordinate system.

[0095] The external parameter relationship from the world coordinate system to the camera coordinate system:

[0096]

[0097] Step 4.2: Substitute X c = x ni Z c and Y c = y ni Z c into the external parameter equation to construct a linear equation system:

[0098]

[0099] where Z w = -ΔH; X wi , Y wi are the world coordinates of the i-th point on the left edge.

[0100] Step 4.3: Solve the world coordinates of the left edge points:

[0101]

[0102] Step 4.4: Calculate the normalized imaging plane coordinate x of the j-th point on the right edge using the same method nj , y nj , construct a linear equation system and solve the world coordinates X wj , Y wj .

[0103] Step 5: Filter outlier points and calculate the mean value for the left and right edge points of the strip in the world coordinate system, determine the actual edge position, and calculate the deviation amount, specifically:

[0104] Step 5.1: Calculate the mean value and standard deviation of the world coordinates of the left edge points and the right edge points:

[0105]

[0106] Among them, N is the total number of left edge points; M is the total number of right edge points; μ1 and σ1 are the mean value and standard deviation of the world coordinates of the left edge points respectively; μ2 and σ2 are the mean value and standard deviation of the world coordinates of the right edge points respectively.

[0107] Step 5.2: Retain the coordinate points on the left edge that satisfy |X wi - μ1| < 3σ1, and retain the coordinate points on the right edge that satisfy |X wj - μ2| < 3σ2, and regard the rest as outlier points and eliminate them.

[0108] Step 5.3: Repeat Steps 5.1 - 5.2 until either of the following conditions is met:

[0109] 1) The standard deviation of the remaining valid points on the left side and the standard deviation of the remaining valid points on the right side are both less than or equal to the set corresponding thresholds;

[0110] 2) The number of iterations reaches 3 times;

[0111] Step 5.4: Calculate the mean value of the world coordinates of the remaining valid points as the reference positions of the left and right edges, and the formula is:

[0112]

[0113]

[0114] Among them, n and m are the numbers of valid points on the left and right edges respectively.

[0115] Step 5.5: Calculate the deviation amount according to the reference positions of the left and right edges, and the formula is:

[0116]

[0117] Among them, D is the deviation amount of the strip steel.

[0118] As Figure 2 shown, it is a deployment diagram of a visual detection system for strip steel deviation in hot rolling based on dynamic height compensation. The detection system includes: an image acquisition module, a distance detection module, a data processing module, and a protection structure.

[0119] The image acquisition module is an industrial camera set between two groups of racks. The industrial camera covers the strip steel area between the racks to ensure that the edges of the strip steel are within the field of view of the camera. During the calibration phase, it is necessary to verify the field of view range to ensure that it only covers the strip steel area directly below the ranging sensor.

[0120] The distance detection module includes multiple ranging sensors. During the calibration phase, it is used to measure the vertical distance from the sensor to the calibration plane, and during the detection phase, it is used to measure the vertical distance from the sensor to multiple measurement points on the strip steel surface.

[0121] The data processing module is respectively connected to the image acquisition module and the distance detection module; the data processing module internally has a calibration parameter storage unit, an edge extraction unit, a first calculation unit, an outlier filtering unit, and a second calculation unit.

[0122] The edge extraction unit extracts the pixel coordinates of the left and right edges of the strip steel according to the collected strip steel image; the first calculation unit calculates the height offset and its average value based on the vertical distance from the sensor to the calibration plane and the vertical distance from the sensor to multiple measurement points on the strip steel surface, and resolves the actual coordinates of the left and right edge points in the world coordinate system according to the pixel coordinates of the left and right edge points and the average value of the height offset; the outlier filtering unit is used to filter the actual coordinates of the left and right edge points in the world coordinate system and eliminate invalid points; the second calculation unit is used to calculate the average value of the world coordinates of the remaining valid points as the left and right edge reference positions, and calculate the deviation amount based on the left and right edge reference positions;

[0123] The protection structure is used to wrap the image acquisition module and the distance detection module, and has the functions of cooling and dust prevention. During specific implementation, the protection structure includes a water-cooled sandwich layer and an air curtain isolation device to adapt to the high-temperature environment of the rolling site.

[0124] The data processing module communicates with the mill level 1 system through industrial Ethernet, and outputs the deviation amount to the HMI interface and control unit of the mill level 1 system.

[0125] The above is only a preferred embodiment of the present invention, and it is not intended to limit the idea of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A visual detection method for the deviation of hot-rolled strip based on dynamic height compensation, characterized in that Including: Step 1: Install an industrial camera between two sets of racks and install multiple ranging sensors at the same height. Calibrate the internal parameter matrix K and the external parameter matrix T = [R|t] of the camera to establish the initial mapping relationship from the pixel coordinate system to the world coordinate system; meanwhile, record the vertical distance a from each ranging sensor to the calibration plane g ; Step 2: During the detection phase, the vertical distance b from the ranging sensor to the strip surface is obtained in real time g , and the height offset Δh is calculated g = a g - b g and its mean value ΔH; Step 3: In the detection stage, an industrial camera is used to collect strip images in real time, and edge detection is performed on the left and right edges of the strip images. Step 4: Based on the pixel coordinates of the points on the left and right edges of the strip and ΔH, the coordinates of the points on the left and right edges of the strip in the world coordinate system are solved through the internal parameter matrix and external parameter matrix of the camera. Step 5: Filter out outliers and calculate the mean value of the left and right edge points of the strip in the world coordinate system to determine the actual position of the edge and calculate the deviation amount.

2. The visual detection method for deviation of hot-rolled strip based on dynamic height compensation according to claim 1, characterized in that The specific content of step 1 is as follows: Step 1.1: Install an industrial camera between two sets of racks. The edge of the strip needs to be within the field of view of the camera, and the strip moves vertically in the camera's field of view. Step 1.2: Install multiple ranging sensors at the same height between two sets of racks. The ranging sensors form a predetermined angle with the optical axis of the industrial camera to ensure coverage of the strip detection area. Step 1.3: Use a checkerboard calibration plate. When the rolling mill is stopped, lower the loop to the horizontal position. Place the checkerboard calibration plate on the upper surface of the loop, making its plane parallel to the rolling center line and coincident with the plane of the initial position of the loop. Define this plane as the calibration plane, that is, the plane where the world coordinate system Z = 0, and the positive direction of the Z-axis is vertically downward. Record the vertical distance a from each distance measuring sensor to the calibration plane g ; Step 1.4: Collect calibration board images with multiple different poses through the industrial camera, and calculate the internal parameter matrix of the camera by the Zhang Zhengyou calibration method: And the external parameter matrix, including the rotation matrix R and the translation vector t: where f x and f y are the focal lengths measured in pixels in the x and y directions of the camera, u0 and v0 are the principal point coordinates of the image in pixels, R is a 3×3 rotation matrix, and t is a 3×1 translation vector.

3. The visual detection method for strip deviation in hot rolling based on dynamic height compensation according to claim 1, characterized in that, The specific content of step 2 is as follows: Step 2.1: During the detection phase, the vertical distance b from the ranging sensor to the strip surface is obtained in real time g , and the height offset Δh is calculated g = a g - b g ; Step 2.2: Calculate the mean value of the height offset ΔH: where G is the total number of ranging sensors.

4. The visual detection method for the deviation of hot-rolled strip based on dynamic height compensation according to claim 1, wherein, In step 3, sub-pixel edge extraction algorithm is used for edge detection, and then the pixel coordinates of the left and right edges of the strip are obtained.

5. The visual detection method for the deviation of hot-rolled strip based on dynamic height compensation according to claim 1, wherein, The specific content of step 4 is as follows: Step 4.1: Calculate the normalized imaging plane coordinates of the left edge points: where u 1i , v 1i are the coordinates of the i-th point on the left edge in the pixel coordinate system; x ni , y ni are the coordinates of the i-th point on the left edge after being transformed to the normalized imaging plane coordinate system; Point x in the normalized imaging plane coordinate system ni , y ni satisfies: Among them, X c , Y c , Z c are the coordinates of the edge points in the camera coordinate system; The external parameter relationship from the world coordinate system to the camera coordinate: Step 4.2: Substitute X c = x ni Z c and Y c = y ni Z c into the external parameter equation to construct a system of linear equations: where Z w = -ΔH; X wi , Y wi are the world coordinates of the i-th point on the left edge; Step 4.3: Solve the world coordinates of the left edge points: Step 4.4: Calculate the normalized imaging plane coordinate x of the j-th point on the right edge using the same method nj , y nj , construct a system of linear equations and solve for the world coordinates X wj , Y wj .

6. The visual detection method for the deviation of hot-rolled strip based on dynamic height compensation according to claim 5, wherein The specific content of step 5 is as follows: Step 5.1: Calculate the mean value and standard deviation of the world coordinates of the left edge points and the right edge points: where N is the total number of left edge points; M is the total number of right edge points; μ1 and σ1 are the mean value and standard deviation of the world coordinates of the left edge points respectively; μ2 and σ2 are the mean value and standard deviation of the world coordinates of the right edge points respectively; Step 5.2: The left edge is retained to satisfy |X wi -μ1|<3σ1 coordinate point, the right edge retains the satisfying |X wj -μ2|<3σ2 coordinate points, and the rest are considered as outliers and removed; Step 5.3: Repeat steps 5.1 - 5.2 until any of the following conditions is met: 1) The standard deviations of the remaining valid points on the left side and the remaining valid points on the right side are both less than or equal to the set corresponding thresholds; 2) The number of iterations reaches 3 times; Step 5.4: Calculate the mean value of the world coordinates of the remaining valid points as the reference positions of the left and right edges, and the formula is: where n and m are the number of valid points on the left and right edges respectively; Step 5.5: Calculate the deviation amount according to the reference positions of the left and right edges, and the formula is:

7. The visual detection method for strip deviation in hot rolling based on dynamic height compensation according to claim 1, characterized in that The ranging sensor is a non-contact ranging sensor, including a laser ranging sensor or an ultrasonic ranging sensor.

8. A visual deviation detection system for hot-rolled strip steel based on dynamic height compensation, which is used to implement the visual deviation detection method for hot-rolled strip steel based on dynamic height compensation according to any one of claims 1-7, characterized in that, Including: An image acquisition module, a distance detection module, a data processing module and a protection structure; The image acquisition module is an industrial camera set between two sets of racks. The industrial camera covers the strip area between the racks to ensure that the edge of the strip is within the field of view of the camera. In the calibration stage, the field of view range needs to be verified to ensure that it only covers the strip area directly below the ranging sensor. The distance detection module includes multiple ranging sensors. In the calibration stage, it is used to measure the vertical distance from the sensor to the calibration plane, and in the detection stage, it is used to measure the vertical distance from the sensor to multiple measurement points on the strip surface. The data processing module is respectively connected to the image acquisition module and the distance detection module; the data processing module is internally provided with a calibration parameter storage unit, an edge extraction unit, a first calculation unit, an outlier filtering unit and a second calculation unit; The edge extraction unit extracts the pixel coordinates of the left and right edges of the strip steel according to the acquired strip steel image; the first calculation unit calculates the height offset and its average value according to the vertical distance from the sensor to the calibration plane and the vertical distances from the sensor to multiple measurement points on the strip steel surface, and resolves the actual coordinates of the left and right edge points in the world coordinate system according to the pixel coordinates of the left and right edge points and the average value of the height offset; the outlier filtering unit is used to filter the actual coordinates of the left and right edge points in the world coordinate system and eliminate invalid points; The second calculation unit is used to calculate the average value of the world coordinates of the remaining valid points as the left and right edge reference positions, and calculate the deviation amount according to the left and right edge reference positions; The protection structure is used to wrap the image acquisition module and the distance detection module, and has the functions of cooling and dust prevention.

9. The visual deviation detection system for hot-rolled strip based on dynamic height compensation according to claim 8, wherein, The protection structure includes a water-cooled interlayer and an air curtain isolation device, which is suitable for the high-temperature environment of the rolling site.

10. The visual deviation detection system for hot-rolled strip based on dynamic height compensation according to claim 8, characterized in that, The data processing module communicates with the mill level 1 system through industrial Ethernet, and outputs the deviation amount to the HMI interface and the control unit of the mill level 1 system.

Images