JP4797887B2 - Flatness measuring method of plate material and flatness measuring device of plate material - Google Patents

Description

  The present invention relates to a plate material flatness measuring method, a plate material flatness measuring apparatus, and a plate material manufacturing method, in which a linear pattern projected on a plate material is photographed by a camera and the flatness is measured from an image.

The plate material is required to have flatness due to its quality. Also, flatness is required for stable production. Therefore, it is a conventional problem to properly manage the flatness in the plate material production process. In general, as the flatness, values such as a difference in elongation and a steepness are used. The elongation difference rate Δε is expressed by the following equation (1) using the elongation difference ΔL between the central portion of the plate material and the vicinity of the plate end in the constant section L.
Δε = ΔL / L (1)
Further, the steepness λ is expressed by the following equation (2) using the wave height δ of the plate material and its pitch P.
λ = δ / P (2)
The elongation difference rate Δε and the steepness λ have the relationship of the following equation (3).
λ = (2 / π) (| Δε |) ^1/2 × 100 (3)

  For example, in a hot rolling line for steel sheets, producing a steel sheet with good flatness not only ensures the quality of the rolled product, but also stabilizes the sheet passing to the finishing mill and coiling with a coil winder. It is also important to maintain high productivity. A hot rolling line is generally composed of a heating furnace, a rough rolling mill, a finish rolling mill row, a cooling zone, and a coil winding machine. The steel sheet overheated in the heating furnace is rolled by a roughing mill to form a steel slab. Thereafter, the steel slab is rolled by a finishing mill row composed of 6 to 7 finishing rolling mills, cooled in a cooling zone, and wound by a coil winder.

  In the hot rolling line, the flatness of the steel sheet is determined by the finish rolling mill and the cooling zone. Therefore, in the finishing mill row, a flatness measuring device is installed on the exit side of the finishing rolling mill row, and the work roll position of each finishing mill is feedback controlled from the measured flatness to ensure flatness. Has been. In the cooling zone, the steel plate is not cooled uniformly in the width direction (in the surface of the plate material, the direction orthogonal to the rolling direction of the plate material; the same applies hereinafter), and thus the steel plate is distorted by thermal stress. For this reason, a flatness measuring device is installed on the exit side of the cooling zone, and the flatness is ensured by feedback control of the cooling water amount of each cooling nozzle from the measured flatness. Thus, in order to produce a steel plate with good flatness, the flatness measuring device is provided between the finishing mills, at the exit side of the finishing mill row, before the coil winding machine, or at a plurality of these locations. It is preferable to install.

  Therefore, Patent Document 1 discloses a plate that scans light in the width direction of the plate, images diffused light under the scanning beam locus, and measures the surface shape from the coordinates of points on the scanning beam obtained from the imaging. A flatness measuring method is disclosed. According to such a technique, even if the plate is in a deformed state such as a wave or a floating state, the flatness can be measured, so that a measuring device excellent for flatness control is provided, and the shape control means can be effectively utilized. It is said that.

  Patent Document 2 discloses a method for measuring the flatness of a coil surface when coiling a metal strip or a strip, wherein a linear pattern is formed on a measurement surface by a projector and directly detected by a camera. ing. According to such a technique, the flatness can be detected and controlled effectively in real time, and the rolling parameters and / or the coil parameters can be finely adjusted quickly. The result is that high strip flatness is achieved at normal high finishing speeds up to 25 m / s used in hot strip mills.

On the other hand, the actual rolled material moves while meandering. In order to measure the flatness at a specific position in the width direction of the plate, or to prevent the measurement position from being removed from the plate by meandering, the flatness measuring device needs to change the measurement position according to the meandering. is there. Patent Document 3 discloses a method of installing a width meter capable of measuring a meandering amount in addition to a flatness measuring device and determining a width direction position of flatness measurement based on a measured value of the meandering amount. . According to this technique, it is said that the flatness control can be performed with high accuracy even when the rolled material is not in the central portion of the plate path.
JP 61-40503 A Japanese Patent Laid-Open No. 11-2511 JP 2000-61520 A

  However, in the invention described in Patent Document 1, the luminance is very strong when measured at a position of regular reflection with respect to the irradiation light source, particularly in a plate material having a strong regular reflection component, and the luminance rapidly decreases when the angle deviates from the angle. For this reason, in the case of a surface shape with a large inclination of the whole plate material or a large steepness, luminance unevenness may occur due to partial reduction in luminance of an image photographed by the camera. In addition, luminance unevenness may occur due to partial reduction in luminance due to a decrease in reflectance due to water on the surface of the plate material or a partial scale. Furthermore, when a large area is projected by a single projector, partial luminance unevenness occurs at the projection center and the periphery. Due to these uneven brightness, when the camera sensitivity is matched to the low brightness part, the camera sensitivity is saturated and the linear pattern is crushed in the high brightness part. In addition, when the camera sensitivity was adjusted to a portion with high luminance, the linear pattern could not be clearly photographed with the portion having low luminance. Therefore, there is a problem that it is difficult to capture a clear linear pattern with a camera on the entire surface of the steel plate.

  On the other hand, it is conceivable to use a powerful light source such as a laser in order to obtain sufficient brightness over the entire surface of the steel sheet. However, since the laser is very expensive, the number of linear patterns is limited and the pitch of the plate wave is short. In some cases, the measurement accuracy of the flatness deteriorates. In addition, the use of a powerful laser has a problem in that it requires strict management, which increases man-hours and is uneconomical.

  The invention described in Patent Document 2 has the same problem as that of Patent Document 1 because it has the same configuration as the invention described in Patent Document 1.

  In the invention described in Patent Document 3, there is a problem that the installation location of the width meter is limited due to its size. In addition, there is a problem that the equipment cost increases due to the installation.

  Therefore, in order to solve the above problem, the present invention does not require a width meter, can clearly shoot a linear pattern even for luminance unevenness, and is a simple method for measuring the flatness of a plate material. It aims at providing the manufacturing method of a flatness measuring apparatus, a finishing rolling mill row | line | column, and a board | plate material.

  The present invention will be described below. In order to facilitate understanding of the present invention, reference numerals in the accompanying drawings are appropriately added in parentheses, but the present invention is not limited to the illustrated embodiment.

According to the first aspect of the present invention, a projector (13) that projects a linear pattern (12) composed of a plurality of parallel lines (14, 14,...) Onto the surface of the plate (11), and the projection onto the plate. A plate material flatness measurement method for measuring the flatness of a plate material by analyzing the image of the linear pattern photographed by the camera (15a, 15b), About the direction, by making the size of the linear pattern larger than that of the plate material, the step of detecting the edges (17a, 17b) of the plate material and the surface of the plate material orthogonal to the line on the basis of the detected edge side of the plate material The step of setting the shape measurement lines (18a to 18e) and the interval between the adjacent lines (14, 14,...) In the shape measurement lines (18a to 18e) on the plate (hereinafter referred to as “linear pattern interval”). )) In advance. A step of calculating the surface angle distribution of the plate material by comparing the distance between adjacent lines in the shape measurement line in the flat reference plate, and the step of calculating the surface angle distribution is a shape measurement line. A step of obtaining a spatial frequency of the linear pattern by converting the luminance distribution (19a, 19b, 19a ′, 19b ′) by the linear pattern (12) at (18a to 18e) into a spatial frequency domain by Fourier transform; The problem is solved by providing a method for measuring the flatness of a plate material, comprising: calculating a reciprocal of a spatial frequency to obtain a line interval .

  As an interval between adjacent lines (interval between linear patterns), the size of at least one line and one dark part adjacent to this line in the width direction of the line, or the size of each line or dark part Can be used.

The invention according to claim 2 is the flatness measurement method for the plate material (11) according to claim 1 , wherein the flatness of the plate material is measured while rolling the plate material in a rolling line, and the linear pattern is parallelized. The longitudinal direction of the plurality of lines is projected so as to intersect the rolling direction of the plate material, and the shape measurement line is set in the rolling direction of the plate material .

According to a third aspect of the present invention, in the flatness measurement method for the plate material (11) according to the first or second aspect , the number of data of the luminance distribution (19a, 19b, 19a ′, 19b ′) is set to 2 by resampling. ^The number of data is ⁿ (n is a natural number), and the luminance distribution is converted into a spatial frequency domain by fast Fourier transform.

The invention according to claim 4 is the flatness measurement method for the plate material (11) according to any one of claims 1 to 3, wherein the luminance distribution is determined with respect to the longitudinal direction of the shape measurement lines (18a to 18e). A step of converting the luminance distribution obtained by inverting the distribution (19a, 19b, 19a ′, 19b ′) into the original luminance distribution, and / or extracting the spatial frequency from the spatial frequency, It has the process of setting it as the spatial frequency moved to the low frequency side.

The invention according to claim 5 is the method for measuring flatness of the plate material (11) according to any one of claims 1 to 4 , wherein a plurality of cameras (15a, 15b) having different sensitivities and / or photographing angles are used. When measuring the flatness of a plate material from a plurality of images taken by a camera, a plurality of luminance distributions (19a, 19b, 19a ′, 19b ′) in the same shape measurement lines (18a to 18e) by the images Among them, when at least one luminance distribution (19a ′) has a portion where the luminance is saturated, the luminance distribution (19a) having a small saturated portion is used, and any of the luminance distributions (19b, 19b ′) is used. When the luminance is not saturated, the luminance distribution (19b ′) having the highest luminance is used.
The invention according to claim 6 is the plate material flatness measurement method according to any one of claims 1 to 5, wherein the reference plate is adjacent to the shape measurement line at a plurality of heights in advance. A step of measuring a distance between the reference plate, a step of measuring the height of the plate material, and a step of determining a line interval of the reference plate to be used from the measured height of the plate material.

According to a seventh aspect of the present invention, in the flatness measuring method for the plate material (11) according to the sixth aspect , a pattern (16a to 16f) for measuring the height of the plate material is projected on the surface of the plate material, and the camera The height of the plate material is measured from the position of the pattern in the image photographed in (15a, 15b).

The invention according to claim 8 is the method for measuring the flatness of the plate material (11) according to any one of claims 1 to 7 , wherein the line of the linear pattern (12) photographed by the camera (15a, 15b). The number of (14, 14,...) Is 40 to 100 .

The invention according to claim 9 is the method for measuring the flatness of the plate material (11) according to any one of claims 1 to 8 , wherein the exposure time dt of the camera (15a, 15b) is set to the following (a): It is obtained by an expression and set.

ただし、Ｐｓ_０’は、板材が水平な場合に線の幅方向における、１本の線とこの線に隣接する一方の暗部とを合わせた大きさである。ｎは大きさＰｓ_０’における暗部の割合、αは鉛直方向に対するカメラの撮影角度、βは鉛直方向に対する投影機の投影角度である。Ｖは板材の速度、θ_ｍａｘは板材の水平方向に対する最大表面角度である。

However, Ps ₀ ′ is a size obtained by combining one line and one dark part adjacent to the line in the width direction of the line when the plate material is horizontal. n is the ratio of the dark part in the size Ps ₀ ′, α is the shooting angle of the camera with respect to the vertical direction, and β is the projection angle of the projector with respect to the vertical direction. V is the speed of the plate material, and θ _max is the maximum surface angle with respect to the horizontal direction of the plate material.

The dark portion ratio n in the size Ps ₀ ′ becomes 1/2 in the width direction of the line if, for example, one line and one dark portion adjacent to the line have the same size. Note that the value of n is not limited, and may be 1/3, 2/3, or the like depending on the linear pattern.

The invention according to claim 10 is a projector (13) that projects a linear pattern (12) composed of a plurality of parallel lines (14, 14,...) Onto the surface of the plate (11), and is projected onto the plate. A plurality of cameras (15a, 15b) having different sensitivities and / or observation angles for photographing a linear pattern, and according to the flatness measuring method for a plate material according to any one of claims 1 to 9, The problem is solved by providing a flatness measuring device (10, 20) for a plate material characterized by measuring the flatness.

The invention according to claim 11, in the flatness measuring apparatus of the plate material according to claim 10 (11) (10, 20), comprising a light source to maximum brightness wavelength 360~560nm the projector (13), The camera (15a, 15b) is provided with a light transmission filter having a wavelength of 360 to 560 nm.

According to a twelfth aspect of the present invention, in the flatness measuring device (10, 20) for the plate material (11) according to the tenth or eleventh aspect , a flat reference plate is used, and the reference plate is measured at a plurality of heights in advance. A storage medium storing a measured value of the interval between the lines (14, 14,...) Is provided.

  Here, “comprising a storage medium” means that not only the flatness measuring device is provided, but also other devices are provided with the storage medium, and the flatness measuring device is provided with the storage medium by an electric communication circuit such as a LAN. It also includes being connected to a device.

The invention according to claim 13 is the finish rolling mill row (100) comprising a plurality of finish rolling mills (101, 101, ...) in series, and any one of claims 10 to 12 between the finish rolling mills. The above-mentioned problems are solved by providing a finishing rolling mill comprising the flatness measuring device (10, 20) for the plate material (11) described in 1.

The invention according to claim 14, plate material and having a reference to flatness measurement method of the plate according to any one of claims 1-9 (11), to measure the flatness of the plate process The above-mentioned problem is solved by providing the manufacturing method.

Invention of Claim 15 has the process of measuring the flatness of a board | plate material using the flatness measuring apparatus (10, 20) of the board | plate material (11) as described in any one of Claims 10-12. The above-mentioned problem is solved by providing a manufacturing method of a plate material characterized by the above.

According to the first aspect of the invention, the width meter is not required by detecting the edge of the plate from the camera image. As a result, the flatness measuring device is reduced, so that the restriction of the installation location is reduced, so that the flatness can be easily measured. In addition, facility costs can be reduced.
Further, since the shape measurement line is set in the image with reference to the edge of the plate material detected from the camera image, the shape measurement line can be quickly set and the accuracy of the set position is improved. Thereby, even if the position of a board | plate material changes, since a shape measurement line can be set easily in the same position, the measurement precision of flatness improves. Further, since the shape in the longitudinal direction of the plate material is calculated only from the position of the shape measurement line instead of measuring the entire line, the amount of calculation can be reduced, and the calculation becomes quick.
Further, by using the reference linear pattern interval, the flatness of the plate material can be calculated from the measured linear pattern interval.
Further, by converting to the spatial frequency domain by Fourier transform, it is possible to calculate the change in the linear pattern interval with high resolution at a resolution higher than the image resolution and high performance.

According to the third aspect of the present invention, it is possible to use the fast Fourier transform by re-sampling the luminance distribution data to 2 ⁿ (n is a natural number). Thereby, the calculation which converts a luminance distribution into a spatial frequency domain can be performed rapidly.

According to the fourth aspect of the present invention, the luminance distribution is a luminance distribution obtained by combining the luminance distribution inverted with respect to the longitudinal direction of the shape measurement line with the original luminance distribution, whereby discontinuous points are formed at both ends of the luminance distribution. Does not occur. Thereby, it is possible to accurately calculate both ends of the luminance distribution in the Fourier transform. Therefore, the measurement accuracy of flatness is improved. Also, by extracting the spatial frequency as a spatial frequency and extracting the spatial frequency to the low frequency side, the discontinuity can be detected stably in the subsequent unwrapping process. . Thereby, the measurement accuracy of flatness improves.

According to the fifth aspect of the present invention, a clear luminance distribution can be obtained at all positions of the shape measurement line by using clear portions from a plurality of images. Thereby, the flatness can be measured at all positions of the shape measurement line.

According to the sixth aspect of the present invention, the reference linear pattern interval changes according to the height of the plate material. Therefore, the measurement accuracy of flatness can be improved by measuring the height of the plate material and using the reference linear pattern interval corresponding to the plate material height.

According to invention of Claim 7 , the height of a board | plate material is measured from the image of a camera. Accordingly, the height of the plate material is calculated by a personal computer (hereinafter referred to as “PC”) in the same manner as the flatness, so that the measurement becomes easy and is reflected in the measurement of the flatness quickly. Thereby, the measurement accuracy of flatness can be improved.

According to the invention described in claim 8 , the number of lines of the linear pattern photographed by the camera is set to 40 to 100, whereby each line is clearly decomposed from the image of the linear pattern in the analysis. Can do. This is because if the number of lines to be photographed is too small, the spatial resolution on the surface of the plate is deteriorated, while if the number of lines to be photographed is too large, the lines cannot be clearly decomposed. Thereby, the flatness measurement precision of a board | plate material improves. In order to further improve the measurement accuracy, it is preferable that the number of lines of the linear pattern photographed by the camera is 50 to 70.

According to the ninth aspect of the present invention, the exposure time of the camera is obtained and set by the above equation (a), so that the adjacent lines of the linear pattern do not interfere with each other even if the plate material moves, which is good. Images can be taken. Thereby, the precision of the flatness measurement of a board | plate material can be improved.

According to the tenth aspect of the present invention, a plurality of images with different luminances can be obtained by providing a plurality of cameras with different sensitivities and / or observation angles. Thereby, a clear luminance distribution can be obtained at all positions of the shape measurement line by using a portion having a clear luminance from a plurality of images. Therefore, the flatness can be measured at all positions on the shape measurement line.

According to the eleventh aspect of the present invention, even when the plate material emits radiation light because of high temperature, light having a wavelength different from the wavelength in the infrared region of the radiation light is used as the light source, and the wavelength of the light source by the filter is used. By photographing only the light, it is possible to photograph a clear linear pattern. Thereby, the flatness can be measured even with a high-temperature plate.

According to the twelfth aspect of the present invention, the height of the plate material is obtained by providing the storage medium storing the measurement values of the reference linear pattern intervals measured in advance at a plurality of heights in a PC or the like of the flatness measuring device. It is possible to measure and use a reference linear pattern interval corresponding to the plate material height. Thereby, the measurement accuracy of flatness can be improved.

According to the invention described in claim 13 , the flatness measuring apparatus for plate material according to any one of claims 10 to 12 does not require a width meter in order to detect the edge of the plate material from the image of the camera. As a result, the apparatus becomes simple. This makes it possible to install a flatness measuring device between finishing mills that have been difficult to install due to its narrow size. Accordingly, since the finish rolling mill and the flatness measuring device have a paired relationship, the finish rolling mill to be adjusted and the adjustment amount are clarified, so that the flatness of the plate material can be improved.

According to the invention described in claim 14 , by using the flatness measuring method for a plate material according to any one of claims 1 to 9 , the flatness can be measured, facilitated, or the measurement accuracy is improved. Therefore, a plate material with good flatness can be manufactured. Further, since the measurement of the flatness becomes rapid, the productivity of the plate material is improved.

According to the invention described in claim 15 , by using the flatness measuring device for a plate material according to any one of claims 10 to 12 , the flatness cannot be measured, so that the plate material is stably produced. can do. Moreover, since the measurement accuracy of flatness is improved, a plate material with good flatness can be manufactured.

  Hereinafter, based on the embodiment shown in the drawings, as an example of a plate material manufacturing method according to the present invention and a finish rolling mill row, a case where a flatness measuring device including two cameras is provided in a hot rolling line of a steel plate. Although described below, what is described below is an example of an embodiment of the present invention, and the present invention is not limited to the following description unless it exceeds the gist.

  FIG. 1 is a diagram schematically showing a flatness measuring apparatus 10 used in a method for manufacturing a steel sheet according to the present invention according to one embodiment. The steel plate 11 has a long shape in the left-right direction (hereinafter referred to as “longitudinal direction”). A projector 13 that projects the linear pattern 12 on the surface of the steel plate 11 is installed on the top of the steel plate 11. The linear pattern 12 is composed of a plurality of parallel lines 14, 14,. The linear pattern 12 is projected on the surface of the steel plate 11 such that the longitudinal direction of the lines 14, 14,... Is perpendicular to the longitudinal direction of the steel plate 11 (hereinafter referred to as “width direction”). Two cameras 15 a and 15 b facing the linear pattern 12 are arranged on the upper part of the steel plate 11. For the sake of easy viewing, some of the reference numerals of the lines 14, 14,... Are omitted.

  FIG. 2 is a diagram showing the arrangement of the projector 13 and the cameras 15a and 15b in the flatness measuring apparatus 10. As shown in FIG. The projector 13 sets the projection direction obliquely downward in the direction opposite to the rolling direction of the steel plate 11, and the projection angle is β with respect to the vertical direction. On the other hand, the cameras 15a and 15b set the shooting direction obliquely downward in the rolling direction of the steel plate 11, and the shooting angle is α with respect to the vertical direction. The cameras 15a and 15b are arranged at the same shooting angle α. The cameras 15a and 15b are connected to the PC 16.

  With this configuration, the linear pattern 12 (see FIG. 1) is projected onto the surface of the steel plate 11 by covering the projector 13 with a slide (not shown) depicting the linear pattern 12. The linear pattern 12 is photographed by the cameras 15a and 15b, and the flatness is calculated by the PC 16 from the images. Here, since the cameras 15a and 15b have different sensitivities, the brightness of the images is different. Therefore, a clear linear pattern 12 can be obtained by properly using the images of the cameras 15a and 15b as described later.

  FIG. 3 is a diagram illustrating a flatness measuring apparatus 20 according to the second embodiment. In addition, about the thing which takes the same structure as FIG. 1, the code | symbol used in FIG. 1 is attached | subjected and the description is abbreviate | omitted. The cameras 15a and 15b are arranged by changing the sensitivity and the shooting angle with respect to the longitudinal direction of the steel plate 11. With this configuration, the cameras 15a and 15b can obtain images having different luminance from the position where the reflection is received in addition to the sensitivity. Therefore, even if there are images where the luminance is saturated in the images of the cameras 15a and 15b or where the weak luminance is weak, by using the images of the cameras 15a and 15b properly as described later, a clear luminance distribution is obtained. It is possible. In addition, in the flatness measuring apparatus 20, although the imaging | photography angle with respect to the longitudinal direction of the steel plate 11 is changed, you may change the angle of the width direction. In addition, the sensitivity of the cameras 15a and 15b may be the same by changing only the shooting position according to the situation.

  In the flatness measuring apparatuses 10 and 20, it is preferable that the projector 13 includes a light source having a wavelength of 360 to 560 nm as a maximum luminance, and the cameras 15a and 15b include a light transmission filter having a wavelength of 360 to 560 nm. According to this, even when the plate material emits radiant light because of high temperature, light having a wavelength different from the wavelength in the infrared region of the radiated light is used as a light source, and the cameras 15a and 15b filter the light of that wavelength. It is possible to photograph the linear pattern 12 clearly by photographing only the image. Therefore, the flatness can be measured even with a high-temperature plate material. A metal halide lamp or a xenon lamp can be used as the light source.

  FIG. 4 is a flowchart 40 for measuring the flatness of the steel sheet. Hereinafter, the case where the flatness of the steel plate 11 is measured using the flatness measuring apparatus 10 will be described with reference to FIG. First, it is determined whether or not the steel plate 11 is in a position where the flatness is measured (step S1). The determination is performed by measuring the luminance of the images of the cameras 15 a and 15 b that capture the linear pattern 12 because the luminance is high when the linear pattern 12 is projected onto the steel plate 11. When an affirmative determination is made in step S1 (when there is a steel plate 11), the flatness of the steel plate 11 is measured in steps S2 to S9. When a negative determination is made in step S1 (when there is no steel plate 11), the flatness measurement of the steel plate 11 ends. Hereinafter, steps S2 to S9 will be described in detail.

(Step S2)
In step S2, the height of the steel plate 11 is measured. Since the steel plate 11 is lifted during traveling on the hot rolling line, the height changes. In the flowchart 40, the surface angle distribution is calculated using the height of the steel plate 11 in step S7 described later. Therefore, it is necessary to measure the height of the steel plate 11 first. FIG. 5 is a diagram showing a linear pattern 12 on the surface of the steel plate 11 taken by a camera. The left and right sides of the drawing are the width direction of the steel plate 11. The linear pattern 12 includes a plurality of parallel lines 14, 14,... And slits 16a to 16f. Each line 14, 14,... Is divided into a central portion and left and right portions in the longitudinal direction, and slits 16a to 16f are disposed therebetween. The linear pattern 12 is projected so that the longitudinal direction of the lines 14, 14,... And the slits 16 a to 16 f is the width direction of the steel plate 11. For the sake of easy viewing, the lines 14, 14,...

  With this configuration, the linear pattern 12 projected onto the steel plate 11 is photographed by the cameras 15a and 15b (see FIG. 1), and the height of the steel plate 11 is measured from the positions of the slits 16a to 16f in the image. FIG. 6A shows the relationship between the position of the slit 16 a and the height of the steel plate 11. Here, illustration of the camera 15b is omitted, and only the camera 15a will be described below, but the same applies to the camera 15b. When the height of the steel plate 11 changes by Δh, the position of the slit 16a on the surface of the steel plate 11 changes by ΔY. Therefore, the position of the slit 16a in the image of the camera 15a changes. Thereby, the height of the steel plate 11 can be calculated from the position of the slit 16a.

  FIG. 6B is a diagram showing the relationship between the vertical position (hereinafter referred to as “Y coordinate”) of the slits 16 a to 16 f in the image of the camera 15 a and the height of the steel plate 11. When the height of the steel plate 11 changes, the Y coordinates of the slits 16a to 16f on the surface of the steel plate 11 in the image of the camera 15a change accordingly. The Y coordinate is determined by the pixel position in the image of the camera 15a. Therefore, the height of the steel plate 11 can be measured by obtaining the relationship of FIG. 6B in advance and measuring the Y coordinates of the slits 16a to 16f.

(Step S3)
In step S3, the edge in the width direction of the steel plate 11 (hereinafter simply referred to as “edge”) is detected. In FIG. 5, the linear pattern 12 is projected so as to be larger than the steel plate 11 in the width direction of the steel plate 11. As a result, the linear pattern 12 is bright due to reflection on the surface of the steel plate 11, and becomes dark because the linear pattern 12 is not reflected at a position off the surface of the steel plate 11. Therefore, the edge side of the steel plate 11 can be detected from the difference in luminance. As an example of the detection method, the luminance of a plurality of lines 14, 14,... Is integrated in the longitudinal direction of the steel plate 11 from the images of the cameras 15a, 15b, and the integrated value is differentiated in the width direction of the steel plate. From the result of differentiation, the positions where the change in luminance is maximum and minimum are the end sides 17a and 17b of the steel plate 11. Thus, the detection accuracy can be increased by detecting the end sides of the steel plate 11 using the plurality of lines 14, 14,. Note that the edge of the steel plate 11 may be detected from the luminance change in the longitudinal direction for one line 14, for example, without depending on the above method.

(Step S4)
In step S4, the shape measurement line of the steel plate 11 is determined. The surface shape in the longitudinal direction of the steel plate 11 is obtained along this shape measurement line. Here, the shape measurement lines 18a to 18e are set in the longitudinal direction of the steel plate 11 from the edge of the steel plate 11 detected in step S3 as shown in FIG. The shape measurement lines 18a to 18e are set to other than the divided portions of the lines 14, 14,.

  In order to calculate the steepness λ of the steel plate 11, in order to obtain Δε in the equation (3), in addition to the shape measuring line 18a at the central portion of the plate width as shown in the equation (1), at least one shape measuring line is used. What is necessary is just to set to the position which measures a shape. However, in actual operation, since the steepness λ is measured at least at both ends in the width direction of the steel plate 11, it is preferable to install the shape measurement lines 18a to 18c. Furthermore, as shown in FIG. 5, it is preferable to set the shape measurement lines 18d and 18e between the shape measurement lines 18a and 18b and 18c. According to this, since the surface shape of the steel sheet is measured at five locations, it is easy to grasp the surface shape. In addition, although a shape measurement line can be further set according to the size of the steel plate 11 in the width direction, the amount of calculation increases according to the number of shape measurement lines, so that the measurement time of flatness becomes long.

  By setting the shape measurement lines 18a to 18e as described above, it is possible to significantly reduce the amount of calculation compared to calculating the shapes of all the lines 14, 14,. Further, by setting the shape measurement lines 18a to 18e on the image with reference to the edge side of the steel plate 11 detected from the images of the cameras 15a and 15b, the shape can be changed even when the edge position of the steel plate 11 changes such as when meandering. The measurement lines 18a to 18e can be accurately set at the same position on the steel plate 11. Thereby, the measurement accuracy of flatness can be improved.

(Step S5)
In step S5, a luminance distribution is created with the shape measurement lines 18a to 18e. The shape measurement lines 18a to 18e cross the plurality of lines 14, 14,. Therefore, the shape measurement lines 18a to 18e have brightness brightness. First, this luminance distribution is measured for each image of the cameras 15a and 15b. As an example of the measurement method, the luminance distributions of the shape measurement lines 18a to 18e are captured as luminance data strings in the PC 16 from the images of the cameras 15a and 15b. At this time, in order to improve noise resistance performance, it is preferable to measure and average the neighboring width direction data for each of the shape measurement lines 18a to 18e.

  FIG. 7A is a diagram showing an example of the luminance distribution measured from the image of the camera 15a, and FIG. 7B is a diagram showing the luminance distribution measured from the image of the camera 15b. The horizontal axis represents the Y coordinate in the images of the cameras 15a and 15b, and the vertical axis represents the luminance. The luminance distributions 19a and 19a 'of the shape measurement line 18a and the luminance distributions 19b and 19b' of the shape measurement line 18b are shown. Here, the Y coordinate is measured from the number of pixels to 351. On the other hand, if the linear pattern interval is collapsed due to saturation of the brightness beyond the camera sensitivity range, or if the linear pattern is buried in the dark current noise of the camera because the brightness is too low, the luminance measurement is not possible. It becomes impossible. Therefore, the sensitivity of the cameras 15a and 15b is changed, and an image with a clear luminance distribution is used for each of the shape measurement lines 18a to 18e. Here, the sensitivity of the camera 15b is set to four times the sensitivity of the camera 15a.

  For example, in the range A of FIG. 7B, the luminance of the luminance distribution 19a 'is saturated and the luminance cannot be measured. Therefore, for the shape measurement line 18a, the luminance distribution 19a in which the luminance is not saturated (the number of saturated portions is small) is used. Further, the luminance distributions 19b and 19b 'of the shape measurement line 18b are not saturated in luminance, but the luminance is not measurable because the luminance is low in the range B in FIG. 7A. Therefore, for the shape measurement line 18b, the luminance distribution 19b 'based on the image of the camera 15b having a high luminance is used. Thus, flatness can be measured by obtaining a clear luminance distribution for each of the shape measurement lines 18a to 18e.

(Step S6)
In step S6, the linear pattern interval Pm (y) is calculated from the luminance distribution of the shape measurement lines 18a to 18e. Details of step S6 will be described later.

(Step S7)
In step S7, the surface angle distribution of the steel plate 11 on the shape measurement lines 18a to 18e is calculated. FIG. 8 is a view of the steel plate 11 viewed from the width direction, and shows a method of calculating the surface angle distribution of the steel plate 11 from the linear pattern interval Pm (y) obtained in step S6. On the upper part of the steel plate 11, a projector 13 and cameras 15a and 15b are installed. Here, illustration of the camera 15b is omitted, and only the camera 15a will be described below, but the same applies to the camera 15b. On the surface of the steel plate 11, two lines 14 a and 14 b and a dark part 14 c therebetween are illustrated in the linear pattern 12 projected from the projector 13.

The camera 15a sets the shooting direction obliquely downward in the longitudinal direction of one of the steel plates 11, and the shooting angle is α with respect to the vertical direction. On the other hand, the projector 13 sets the projection direction obliquely downward in the other longitudinal direction of the steel plate 11, and the projection angle is β with respect to the vertical direction. Let the space | interval of line 14a, 14b be the magnitude | size which match | combined the line 14a and the dark part 14c in the width direction (longitudinal direction of a steel plate) of line 14a, 14b. Line 14a on the surface of the steel sheet 11 when the steel sheet 11 is horizontal, the spacing 14b and Ps _0, the line 14a in the image of the camera 15a, the distance 14b and Ps (y). When the surface angle with respect to the horizontal direction of the steel plate 11 is θ, the interval between the lines 14a and 14b on the surface of the steel plate 11 is Pm _0, and the interval between the lines 14a and 14b in the image of the camera 15a is Pm (y). . At this time, the following equations (4) to (6) are established geometrically.

（４）式に、（５）、（６）式を代入することで、次の（７）式を導くことができる。

By substituting the equations (5) and (6) into the equation (4), the following equation (7) can be derived.

ここで、α及びβは、設定値である。そのため、平坦な基準板で予め基準線状パターン間隔を測定し、これを線状パターン間隔Ｐｓ（ｙ）とする。そして、ステップＳ６により線１４ａ、１４ｂの間隔Ｐｍ（ｙ）を測定することで、線１４ａと１４ｂとの間における鋼板１１の表面角度θを求めることができる。同様にして、全ての線１４、１４、…の間隔から鋼板１１の表面角度θを求めることで、形状測定線１８ａ〜１８ｅにおける表面角度分布を求めることができる。

Here, α and β are set values. Therefore, the reference linear pattern interval is measured in advance with a flat reference plate, and this is defined as the linear pattern interval Ps (y). And the surface angle (theta) of the steel plate 11 between the lines 14a and 14b can be calculated | required by measuring the space | interval Pm (y) of the lines 14a and 14b by step S6. Similarly, the surface angle distribution in the shape measurement lines 18a to 18e can be obtained by obtaining the surface angle θ of the steel plate 11 from the intervals between all the lines 14, 14,.

  In addition, the linear pattern space | interval Ps (y) can be measured by implementing said step S2-6 about a reference | standard board. Since the linear pattern interval Ps (y) varies depending on the height of the steel plate 11, it is necessary to measure at a plurality of heights of the steel plates 11 in advance. And the linear pattern space | interval Ps (y) to be used is determined from the height of the steel plate 11 measured at step S2. An example of the measurement result of Ps (y) at the height of one steel plate 11 is shown in FIG. FIG. 9 shows the relationship between the Y coordinate of the image on the reference plate and the linear pattern interval Ps (y). Thereby, the linear pattern space | interval Ps (y) can be determined by the height of the steel plate 11 and the position (Y coordinate) in the steel plate 11 being determined.

(Step S8)
In step S8, the surface shape of the steel plate 11 is calculated from the surface angle distribution obtained in step S7. The surface shape of the steel plate 11 can be obtained for each of the shape measurement lines 18a to 18e by integrating the surface angle obtained in step S7.

(Step S9)
In step S9, the steepness λ of the steel plate 11 is calculated from the surface shape obtained in step S8. The elongation ε is calculated by calculating the surface length of the surface shape at the positions of the respective shape measurement lines 18a to 18e. Then, an elongation difference rate Δε is obtained from the elongation rate ε 18a in the shape measurement line 18a and the elongation rates ε _m in the other shape measurement lines 18b to _18e, and the steepness λ of the steel plate 11 is calculated. Here, the subscript m is 18a to 18e, and represents the value in the shape measurement line m. From the obtained elongation rate, the difference (elongation difference rate) Δε from the central portion is calculated by the following equation (9).

Here, the case where the steepness λ of the steel plate 11 is obtained from the surface shape of the shape measurement lines 18a and 18b will be described. FIG. 10A is a diagram showing the surface shape 20a at the shape measurement line 18a, and FIG. 10B is a diagram showing the surface shape 20b at the shape measurement line 18b. The elongation is calculated by calculating the surface lengths of the surface shapes 20a and 20b and the linear distance therebetween. The surface length is calculated by dividing the target section and calculating by approximating the polygonal line. According to this, the influence of minute measurement noise can be suppressed. Here, the target section is divided into 12 and the surface length is calculated by polygonal line approximation. Assuming that each point after division is P _i (i = _{0 to 12} ) and the linear distance between the surface shapes 20a and 20b is P ₀ P ₁₂ , the elongation ε _m is expressed by the following equation (8). .

得られた伸び率から、次の（９）式により中心部との差（伸び差率）△εを計算する。
△ε＝ε_１８ａ−ε_１８ｂ ・・・（９）
この△εから、（３）式により急峻度λを求めることができる。同様に、形状測定線１８ａと、形状測定線１８ｃ〜１８ｅとによっても急峻度λを測定することが可能である。

From the obtained elongation rate, the difference (elongation difference rate) Δε from the central portion is calculated by the following equation (9).
Δε = ε _18a −ε _18b (9)
From this Δε, the steepness λ can be obtained by the equation (3). Similarly, the steepness λ can be measured by the shape measurement line 18a and the shape measurement lines 18c to 18e.

  By calculating the steepness λ of the steel plate 11 by the above steps S1 to S9, the flatness of the steel plate 11 can be obtained.

  FIG. 11 is a diagram showing detailed steps of step S6 described above. In steps S6a to 6c, the luminance distributions of the shape measurement lines 18a to 18e are converted into the spatial frequency domain. In steps S6d and 6e, the converted function in the spatial frequency domain is inversely transformed into the spatial domain. In steps S6f to 6h, the interval Pm (y) between the linear patterns 12 is calculated from the spatial frequency of the inversely transformed space region. Hereinafter, steps S6a to S6h will be described in detail.

(Step S6a)
In step S6a, if the number of luminance data in the luminance distributions of the shape measurement lines 18a to 18e is not 2 ⁿ (n is a natural number; the same applies hereinafter), the number of luminance data is resampled to a luminance distribution k (x) of 2 ^n. . This makes it possible to apply fast Fourier transform, which is a high-speed calculation method of discrete Fourier transform, to the luminance distribution k (x). FIG. 12 shows an example of re-sampling the number of luminance data from 6 to 8. In the luminance data D1 to D6 of 6 points in FIG. 12A, adjacent data are connected by a line. Then, when the area between D1 to D6 is divided into eight equal parts, and the intersection of the line that divides into eight parts and the luminance data are set as new luminance data D1 ′ to D8 ′, the luminance data as shown in FIG. The number is 8. Here, due to re-sampling, there is a difference in position between the luminance data D1 to D6 and D1 ′ to D8 ′. However, the linear pattern interval Pm (y) and the linear pattern interval Ps (y) are used for measuring the flatness of the steel plate 11 as shown in the equation (3), and Pm (y), Ps (y) ) Since both are resampled, the flatness value is not affected. Note that it is preferable to increase the number of data in resampling in order not to reduce the accuracy of luminance data.

(Step S6b)
In step S6b, the resampled luminance distribution k (y) is converted into symmetric data. In the discrete Fourier transform, it is assumed that the same waveform is repeated outside the target signal waveform. The luminance distributions k (y) of the shape measurement lines 18a to 18e are different in both spatial frequency (linear pattern interval) and luminance at both ends, so that large discontinuities occur. This results in a measurement error at the end of the luminance distribution k (y) in the calculation result of the linear pattern interval. Therefore, with respect to the luminance distribution k (y), the luminance data string of the target luminance distribution k (y) is inverted and added after the original luminance data string, so that the luminance distribution g (y) having twice the number of data points. And As a result, since the end of the luminance distribution g (y) is not a discontinuous point, it is possible to accurately calculate the linear pattern interval up to the signal end. FIG. 13 is a diagram showing the conversion of luminance data into symmetric data. FIG. 13A shows a case where the luminance distribution k (y) before being converted into symmetric data is subjected to discrete Fourier transform. Both ends C of k (y) are discontinuous points due to differences in spatial frequency (linear pattern interval; horizontal axis) and luminance (vertical axis). In FIG. 13 (b), discrete Fourier transform is performed using g (y) obtained by inverting k (y) in the horizontal axis direction and combining after k (y). Thereby, since both ends D of g (y) do not become discontinuous points, it is possible to prevent a measurement error from occurring at the end D.

(Step S6c)
In step S6c, a fast Fourier transform is performed on the luminance distribution g (y) converted into symmetric data. Thereby, the spatial distribution of luminance is converted into a spatial frequency domain. If the converted function is G (f) and the fast Fourier transform is F [], it can be expressed by the following equation (10).
G (f) = F [g (y)] (10)

(Step S6d)
In step S6d, a spatial frequency region is extracted from the function G (f) that has been subjected to the fast Fourier transform. In the extraction, W (f) is added to G (f) and then moved to the lower frequency side by f _S. As shown in FIG. 14, W (f) is a function in which the frequency band f _{L to} f _{H of the} linear pattern interval Pm (y) is 1 (the unit is 1 / pixel), and the others are 0. By adding W (f) to g (f), the negative spatial frequency band value of g (f) becomes 0, and only the linear pattern existing in the positive spatial frequency band can be left. Thereby, the influence which the pattern produced by the scale production | generation nonuniformity of the steel plate 11 surface, and water riding have on a spatial frequency can be suppressed.

The frequency bands f _{L to} f _H of the linear pattern interval Pm (y) in W (f) can be obtained as follows. If the shooting angle α of the camera 15a in FIG. 8 and the projection angle β of the projector 13 are determined, the surface angle θ of the steel plate 11 and the linear pattern interval ratio Pm as shown in FIG. The relationship with (y) / Ps (y) can be obtained. The same applies to the camera 15b. Here, the measurement range of the surface angle θ is determined from the actual operation. The measurement range of the surface angle θ is determined by the sum of the range of the surface angle θ ₁ obtained from the required steepness measurement range and the range of the surface angle θ ₂ resulting from the inclination of the entire steel plate 11 that can occur during measurement. . Here, when the measurement range of the surface angle θ is −10 ° ≦ θ ≦ 10 °, Pm (y) / Ps (y) is 0.82 to 1.20 as shown in FIG.

The linear pattern interval Ps (y) is obtained in advance with the reference plate as described above. As shown in FIG. 9, when Ps (y) is 7 to 13, Ps (y) is integrated to Pm (y) / Ps (y), and Pm (y) is 5.74 (= 7 × 0.82). ) To 15.6 (= 13 × 1.2). Therefore, the reciprocal spatial frequency is 0.064 (= 1 / 15.6) to 0.174 (= 1 / 5.74). Thereby, it can be determined that f _L = 0.064 and f _H = 0.174.

Then, set the _{f S} such that _{f S} ≦ _{f L.} Then, G (f) × W (f) is moved to the low frequency side by f _S. Thereby, in the unwrapping process in step S6g described later, a case where it is not possible to determine whether the amount of change in the phase angle is large or a discontinuous point is suppressed, so that the discontinuous point can be detected stably. .

As described above, the function H (f) in the spatial frequency region after extraction can be expressed by the following equation (11).
H (f) = G (f + fs) · W (f + fs) (11)
Although it is not always necessary to move the frequency f _S , when the linear pattern interval Pm (y) is narrow (the spatial frequency is high), the discontinuity point is stabilized in the unwrapping process in step S6g described later. It is preferable to carry out the detection.

(Step S6e)
In step S6e, the function H (f) is converted from spatial frequency domain data to luminance spatial distribution by inverse fast Fourier transform. The converted result is defined as _gan (y). F ⁻¹ [] represents an inverse fast Fourier transform which is a transformation from the spatial frequency domain to the spatial distribution. Thereby, _gan (y) can be expressed by the following equation (12).
g _an (y) = F ⁻¹ [H (f)] (12)

(Step S6f)
In step _S6f, calculates the _g an, real part _Re of the _{(y) [g an (y} )] and the imaginary part _{Im [g} an (y)] from the phase angle phi (y). φ (y) can be expressed by the following equation (13).
φ (y) = tan ⁻¹ [Im [ _gan (y)] / Re [ _gan (y)]] (13)

(Step S6g)
The phase angle φ (y) is folded to −π / 2 to π / 2. This is said to be wrapped. Therefore, in step S6g, an unwrapping process for adding or subtracting π is performed so that the phase angle φ (y) is differentiated and smoothly connected at discontinuous points. This makes φ (y) a continuous wave. The differential value of the phase angle φ (y) is proportional to the spatial frequency −f _S of the linear pattern. Therefore, the spatial frequency distribution of the linear pattern can be obtained by the following equation (14).
f (y) = − dφ / dx / (2π) + f _S (14)

(Step S6h)
In step S6h, the reciprocal of f (y) is calculated to calculate the linear pattern interval Pm (y). Pm (y) can be expressed by the following equation (15).
Pm (y) = 1 / f (y) (15)
Here, since the latter half of the data is a portion where the inverted data is combined by the above-described step S6b, the first half is used as valid data.

  In the above-described embodiment, the flatness of the steel plate 11 is obtained by calculating the steepness. However, the shape of the line 14 in the longitudinal direction of the image of the camera 15a and / or 15b is generally used in image processing. By measuring the surface shape of the steel plate 11 in the width direction, it is possible to control an apparatus such as a rolling mill so as to obtain good flatness.

  Moreover, in the said embodiment, it is preferable that the number of the lines 14, 14, ... (refer FIG. 5) of the linear pattern 12 image | photographed with the cameras 15a and 15b shall be 40-100. According to this, each line can be clearly decomposed from the image of the linear pattern in the analysis. Therefore, the flatness measurement accuracy of the plate material is improved. For example, as shown in FIG. 16, the variation σ of the surface angle θ of the steel material 11 (see FIG. 1) obtained in step S7 described above can be reduced. Note that the variation σ of the surface angle of the steel material 11 in FIG. 16 is obtained by measuring the surface angle distribution of a completely flat measurement object whose surface angle is 0 ° over the entire surface, and measuring the angular distribution from an error (from 0 °). The standard deviation of the deviation) was calculated.

In the said embodiment, the measurement range of the rolling direction of the steel material 11 is 1400 mm. In addition, the interval between the linear patterns 12 (see FIG. 1) is a size obtained by combining one line 14 and one dark portion adjacent to the line 14 in the width direction of the lines 14, 14,. 8). Therefore, the linear pattern 12 interval can be obtained by the following equation (16).
Linear pattern 12 interval = measurement range in rolling direction / number of lines 14, 14,... (16)
Thereby, the space | interval of the linear pattern 12 becomes 35 mm-14 mm.
In order to further improve the measurement accuracy, it is preferable to set the number of lines 14, 14,... Of the linear pattern 12 photographed by the cameras 15a, 15b to 50 to 70.

Furthermore, it is preferable that the exposure time dt of the cameras 15a and 15b is obtained and set by the above equation (a). FIG. 17A is a diagram showing a method for obtaining the exposure time dt of the cameras 15a and 15b (not shown). FIG. 17B is an enlarged view of a portion A in FIG. The cameras 15a and 15b have the shooting direction obliquely below one longitudinal direction of the steel plate 11, and the shooting angle is α with respect to the vertical direction. Hereinafter, only the camera 15a will be described, but the same applies to the camera 15b. On the other hand, the projector 13 sets the projection direction obliquely downward in the other longitudinal direction of the steel plate 11, and the projection angle is β with respect to the vertical direction. A surface angle in the longitudinal direction of the steel plate 11 with respect to the horizontal direction is defined as θ. In the horizontal plate member 11, the size of the line 14a and the dark portion 14c in the width direction of the lines 14a and 14b (hereinafter referred to as “the size of the line 14a”) is defined as Ps ₀ ′. The size of the line 14a in the steel plate 11a having the surface angle θ is defined as Pm (y) ′. The steel plate 11a ′ is a position after the steel plate 11a has moved at the speed V to the exposure time dt of the camera 15a. The steel plate 11a ′ rises dh from the height of the steel plate 11a. In the steel plate 11a ′, the size of the line 14a increased during the exposure time dt is defined as dPm (y) ′.

Here, when dh ′ is set as shown in FIG.
dh ′ = dh + dh ′ · tan β · tan θ
Therefore, dh ′ can be expressed by the following equation (17).

これにより、ｄＰｍ（ｙ）’は

であることから、次の（１８）式で表すことができる。

As a result, dPm (y) ′ becomes

Therefore, it can be expressed by the following equation (18).

また、上記（７）式、（５）式を用いて、Ｐｍ（ｙ）’は、次の（１９）式で表すことができる。

Moreover, Pm (y) ′ can be expressed by the following equation (19) using the above equations (7) and (5).

In order to recognize the linear pattern 12 (see FIG. 1), it is necessary that the dark portion 14c remains. For that purpose, it is necessary to satisfy the following equation (20).
n · Pm ′ (y)> dPm (y) ′ (20)
Here, n is the ratio of the dark portion 14c in the size Ps ₀ ′ of the line 14a. Substituting equations (18) and (19) into equation (20) described above yields the following equation (21).

一方、高さｄｈは、次の（２２）式で表すことができる。
ｄｈ＝Ｖ・ｄｔ・ｔａｎθ ・・・（２２）
（２１）式、（２２）式から、カメラ１５ａの露光時間ｄｔは、次の（２３）式のとおりとなる。

On the other hand, the height dh can be expressed by the following equation (22).
dh = V · dt · tan θ (22)
From the equations (21) and (22), the exposure time dt of the camera 15a is as shown in the following equation (23).

（２３）式では、表面角度θが大きいほど露光時間ｄｔを小さくする必要がある。そのため、露光時間ｄｔを次の（２４）式とする。

In equation (23), the exposure time dt needs to be reduced as the surface angle θ increases. Therefore, the exposure time dt is set to the following equation (24).

ここで、θ_ｍａｘは、表面角度θの最大値である最大表面角度である。この（２４）式を用いてカメラ１５ａ、１５ｂの露光時間ｄｔを求め、設定することにより、最大表面角度θ_ｍａｘで鋼板１１が移動しても線状パターンの隣接する線同士が干渉せず、良好な画像を撮影することができる。これにより、鋼板１１の平坦度測定の精度を向上させることができる。

Here, θ _max is the maximum surface angle that is the maximum value of the surface angle θ. By obtaining and setting the exposure times dt of the cameras 15a and 15b using this equation (24), even if the steel plate 11 moves at the maximum surface angle θ _max , adjacent lines of the linear pattern do not interfere with each other, A good image can be taken. Thereby, the precision of the flatness measurement of the steel plate 11 can be improved.

  FIG. 18 is a diagram schematically showing the finish rolling mill row 100. The finish rolling mill row 100 has six finishing mills 101, 101,... Arranged in series. Further, flatness measuring devices 10, 10,... Are provided between the finishing mills 101, 101,. Since the space between the finishing mills 101, 101,... Is narrow, it has conventionally been difficult to install a flatness measuring device. The flatness measuring device 10 can be installed between the finishing mills 101, 101,... Thereby, since the finish rolling mill 101 and the flatness measuring apparatus 10 have a paired relationship, the finishing mills 101, 101,... To be adjusted from the measured values of the flatness measuring apparatuses 10, 10,. The amount of adjustment becomes clear. Therefore, the flatness of the steel plate 11 can be improved by controlling the finishing mills 101, 101,... With high accuracy. In addition to the flatness measuring device 10, the finishing rolling mill row can be provided with a flatness measuring device according to the present invention, such as the flatness measuring device 20.

  In the examples, the flatness measuring apparatus 10 (see FIG. 2) in the above embodiment was used in a hot rolling line for steel sheets. Hereinafter, a description will be given with reference to FIG. The flatness measuring apparatus 10 sets the projection angle β of the projector 13 to 15 °. In addition, the photographing angle α of the cameras 15a and 15b is the same at 40 °. As the projector 13, a metal halide lamp with an output of 2.5 kW was used. The light source had a luminance peak at a wavelength of 360 nm to 560 nm. The linear pattern 12 was projected on the surface of the steel plate 11 through the slide and the lens. The slide was generated by vapor deposition of Cr on a quartz glass substrate, and 60 lines 14, 14,... (See FIG. 1) having a line width of 0.72 mm were created at equal intervals. Since the projector 13 was installed at a site where a large amount of dust and mist-like water droplets were scattered, the entire projector 13 was housed in a dust-proof box made of stainless steel. In addition, air is supplied into the dustproof box with a large blower so that dust or mist-like water droplets do not enter the dustproof box from the opening where the linear pattern 12 (see FIG. 1) is projected, and is ejected from the opening. The structure.

  CCD cameras were used as the imaging cameras 15a and 15b. This CCD camera made it possible to output 50 images per second and to capture a plurality of cameras synchronously. The sensitivity ratio of the cameras 15a and 15b was 1: 4. In addition, a filter that transmits light having a wavelength of 360 nm to 560 nm was provided in front of the lenses of the cameras 15a and 15b. Similarly to the projector 13, the cameras 15a and 15b were also housed in a stainless steel dustproof box, and compressed air was supplied to the lens to prevent contamination.

The exposure times of the cameras 15a and 15b were obtained by the above equation (a). As described above, the shooting angle α of the cameras 15a and 15b is 40 °, and the projection angle of the projector 13 is 15 °. Further, the size Ps ₀ ′ of the lines 14, 14,... Is 60 mm on the average in the measurement range 1400 mm in the rolling direction of the steel plate 11, so that the average is 23 mm (1400 mm / 60). there were. The ratio n of the dark part in the size Ps ₀ ′ of the lines 14, 14,. The range of the surface angle θ is determined by the sum of the range of the surface angle θ ₁ obtained from the required steepness measurement range and the range of the surface angle θ ₂ resulting from the inclination of the entire steel plate 11 that can occur during measurement. Assuming that the shape of the steel plate 11 is a sine wave having a wave height δ and a pitch P of the plate material as shown in FIG. 19A, the surface angle θ ₁ is obtained by the following equation (25) as shown in FIG. 19B. Can do.
θ ₁ = πλ · cos (2πy / P) (25)
The unit of θ ₁ is rad. Here, λ is the steepness (δ / P). In the embodiment, since the steepness measurement range is ± 5%, λ is 0.05. y is the position of the steel plate 11 in the rolling direction. Since −1 ≦ cos (2πy / L) ≦ 1, the range of the surface angle θ ₁ was −9 ° ≦ θ ₁ ≦ 9 ° when the unit was converted to °. On the other hand, in the measurement place of an Example, the inclination of the steel plate 11 whole hardly arises. Therefore, the range of the surface angle θ is set to −10 ° ≦ θ ≦ 10 °. As a result, the maximum surface angle θ _max of the steel plate 11 was 10 °. The speed V of the steel plate was 25 m / sec at the maximum. Thus, since dt <2.80 msec was obtained from the above equation (a), the exposure time of the cameras 15a and 15b was set to 1 msec.

  The PC 16 can capture the images from the two cameras 15a and 15b into the memory at the same time with the brightness gradation of 0 to 255 by the built-in multi-channel image capturing board. From the number of pixels of the image, the number of luminance data in the luminance distribution of the shape measurement line was 351. The PC 16 can execute the above steps S1 to S9 by the created flatness analysis program.

Example 1
FIG. 20 is a diagram illustrating the results of Example 1 in which the edges 17a and 17b (see FIG. 5) of the steel plate 11 are detected from the images of the cameras 15a and 15b in step S2 (see FIG. 4) described above. FIG. 20A is a diagram showing the relationship between time and the actually measured meandering amount of the steel plate 11. FIG. 20B shows the shape measurement line 18a and the shape when the edge sides 17a and 17b of the steel plate 11 are detected from the images of the cameras 15a and 15b and the shape measurement lines 18a to 18c (see FIG. 5) are set. It is a figure which shows the measured value of the steepness degree (lambda) of the steel plate 11 by the measurement lines 18b and 18c. FIG. 20C shows the measurement of the steepness λ of the steel plate 11 by the shape measurement line 18a and the shape measurement lines 18b and 18c when the shape measurement lines 18a to 18c are set from the width data of the steel plate 11 according to the rolling schedule. It is a figure which shows a value. In FIG. 20A, an abnormal value E occurs in the steepness λ in the time when the meandering amount of the steel plate 11 becomes large in FIG. Therefore, in the method of setting the shape measurement lines 18a to 18c from the width data of the steel plate 11, it has been confirmed that the measurement accuracy of the steepness λ is deteriorated. On the other hand, in FIG. 20B, the abnormal value of the steepness λ did not occur. Thus, it was confirmed that the steepness λ can be measured with high accuracy by detecting the edges 17a and 17b of the steel plate 11 from the images of the cameras 15a and 15b and installing the shape measurement line.

(Example 2)
FIG. 21A to FIG. 21C are diagrams showing the results of Example 2 in which the above-described steps S6a to S6h (see FIG. 11) and step S7 (FIG. 4) are performed. The shape measurement line 18a was set, and the luminance distribution was measured. In step S6a, the luminance distribution of the shape measurement line 18a is resampled to the data number 512 to obtain the luminance distribution k (y). Next, when k (y) is inverted in step S6b and added after k (y) to obtain a luminance distribution g (y), the result is as shown in FIG. When step S6c to step S6h were executed for this g (y), the linear pattern interval Pm (y) was as shown in FIG. Then, when the surface angle distribution of the shape measurement line 18a was obtained in step S7, it was as shown in FIG. Thereby, it was confirmed that the surface angle distribution in the shape measurement line was obtained from the luminance distribution based on the camera image.

(Comparative Example 1)
FIG. 22A to FIG. 22C are diagrams showing the results of Comparative Example 1 with respect to Example 2 described above. In Comparative Example 1, as shown in FIG. 22A, the luminance distribution was not converted into symmetrical data in step S6b. The other steps were performed in the same manner as in Example 2. In Comparative Example 1, a large measurement error occurred near 0 in the Y coordinate as shown in FIGS. Therefore, it was confirmed that the surface angle distribution can be accurately calculated by converting the luminance distribution into symmetric data in step S6b.

(Comparative Example 2)
FIG. 23A to FIG. 23D are diagrams showing the results of Comparative Example 2 with respect to Example 2 described above. In Comparative Example 2, Steps S6a to S6c are executed in Example 2, and after W (f) is added to G (f) in Step S6d, it exists in the positive spatial frequency band as shown in FIG. 23 (a). The linear pattern was not moved to the low frequency side by f _S. Therefore, the function H (f) is obtained by the following equation (11) ′ instead of the above equation (11).
H (f) = G (f) · W (f) (11) ′
The other steps were performed in the same manner as in Example 2. For this reason, in step S6g, as shown in FIG. 23B, the unwrapping process could not be performed in the portion F where the spatial frequency is high. As a result, the linear pattern interval Pm (y) and the value of the surface angle distribution θ were not determined in the portion F in FIGS. 23 (c) and 23 (d). Therefore, in step S6d, the linear pattern interval Pm (y) and the surface angle distribution are calculated even in the portion where the spatial frequency is high by moving the linear pattern existing in the positive spatial frequency band by f _S toward the low frequency side. It was confirmed that

(Example 3)
FIG. 24 is a diagram illustrating the results of Example 3 in which the relationship between the measurement height h _{Ps (y) of} the linear pattern interval Ps (y), the height h _{Pm (y) of} the steel plate 11, and the steepness λ is measured. It is. The steel plate 11 for measuring the steepness λ was prepared with a steepness of 0%. The height h _{Pm (y)} of the steel plate 11 was set such that the surface height of the table roller was zero. The linear pattern interval Ps (y) was measured in advance by setting the measurement height h _{Ps (y)} of the reference plate to 18 mm, 63 mm, and 108 mm. Then, for each of the linear pattern interval Ps (y), by changing the height _{h Pm} of the steel sheet 11 _(y) to 0-200 mM, it was measured steepness lambda.

According to this, as the height h _{Pm (y) of} the steel plate 11 is closer to the height h _{Ps (y) obtained} by measuring the linear pattern interval Ps (y), the steepness becomes closer to 0%, and the flatness is reduced. It was confirmed that the accuracy was good. Therefore, it is necessary to use Ps properly according to the height of the steel plate. Specifically, in order to make the steepness measurement error within ± 0.05% at 24, the reference plate height h _{Ps (y)} is 18 mm when the steel plate height h _{Pm (y)} is about 40 mm or less. The linear pattern interval Ps (y) is used. Similarly, for the steel plate height h _{Pm (y)} of about 40 mm to about 80 mm, the linear pattern interval Ps (y) having the reference plate height h _{Ps (y)} of 63 mm is used. For a steel plate height h _{Pm (y)} of about 80 mm to about 130 mm, a linear pattern interval Ps (y) having a reference plate height h _{Ps (y)} of 108 mm is used. Thus, by determining h _{Ps (y)} to be used for the steel plate height h _{Pm (y)} in advance from the actual results, even when the steel plate height h _{Pm (y)} changes greatly, the steepness decreases. Can be measured with high accuracy.

  (Example 4)

  In Example 4, the steepness λ was actually measured using the flatness measuring apparatus 10. The flatness measuring apparatus 10 can measure the steepness λ about 15 times per second. As a result, it was confirmed that the response speed has no problem in actual operation.

  In addition, although the case where the flatness measuring apparatus provided with two cameras was provided in the hot rolling line of a steel plate was shown in the said embodiment, three or more cameras may be sufficient. Moreover, you may measure the flatness of a steel plate from one image by using one camera. Furthermore, the present invention can be applied not only to the steel plate but also to the flatness measurement of other materials such as aluminum and copper.

  While the present invention has been described in connection with embodiments that are presently the most practical and preferred, the present invention is not limited to the embodiments disclosed herein. However, it can be changed as appropriate without departing from the gist or concept of the invention that can be read from the claims and the entire specification, and the flatness measuring method, plate flatness measuring device, and finish rolling involving such changes. It should be understood that the manufacturing method of the machine train and the plate material is also included in the technical scope of the present invention.

It is the figure which showed the flatness measuring apparatus used for the manufacturing method of a steel plate in the schematic diagram. It is a figure which shows arrangement | positioning of a projector and a camera. It is a figure which shows the flatness measuring apparatus which concerns on 2nd Embodiment. It is a flowchart which measures the flatness of a steel plate. It is a figure which shows the linear pattern of the steel plate surface image | photographed with the camera. It is a figure which shows the relationship between the position of a slit, and the height of a steel plate. It is a figure which shows an example of the luminance distribution measured from the image of the camera. It is a figure which shows the method of calculating the surface angle distribution of a steel plate from a linear pattern space | interval. It is a figure which shows an example of the measurement result of the reference | standard linear pattern space | interval in one steel plate height. It is a figure which shows the surface shape in a shape measurement line. It is a figure which shows the detailed step of calculation (step S6) of a linear pattern space | interval. It is a figure which shows an example of the condition which resamples luminance data. It is a figure shown about symmetrical data conversion of luminance distribution. It is a figure which shows the function W (f) which leaves only the linear pattern which exists in a positive spatial frequency band from the function of a spatial frequency domain. It is a figure which shows the relationship between the surface angle (theta) of a steel plate, and the linear pattern space | interval ratio Pm (y) / Ps (y) in a reference | standard board and a steel plate. It is a figure which shows the relationship between the number of the lines in a linear pattern, and the dispersion | variation in the steel material surface angle. It is a figure which shows the method of calculating | requiring the exposure time dt of a camera. It is a figure which shows in a schematic diagram a finishing mill row | line | column provided with a flatness measuring apparatus between finishing mills. It is a figure which shows the shape of a steel plate, and the surface angle (theta) ₁ of a steel plate. It is a figure which shows the result of having detected the edge of the steel plate from the image of the camera. It is a figure which shows the result of having calculated | required the surface angle distribution of the steel plate from the luminance distribution of the shape measurement line. It is a figure which shows the result of having calculated | required the surface angle distribution of the steel plate, without making the luminance distribution symmetrical data from the luminance distribution of the shape measurement line. It is a diagram illustrating a result of obtaining the surface angle distribution of the steel plate a linear pattern present in the positive spatial frequency band without f _S moves to the low frequency side. It is a figure which shows the relationship between the measurement height of the measured reference linear pattern space | interval, the height of a steel plate, and steepness.

Explanation of symbols

Pm (y) Linear pattern interval Ps (y) Linear pattern interval when plate material is horizontal α Camera imaging angle β Projector projection angle θ Surface angle of plate material λ Steepness 10 Flatness measuring device 11 Plate material 12 Linear Pattern 13 Projector 14, 14a, 14b Line 14c Dark part 15a, 15b Camera 16a Slit 18a-18e Each shape measuring line 20 Flatness measuring device 100 Finishing mill row 101 Finishing mill

Claims (15)

A projector that projects a linear pattern composed of a plurality of parallel lines onto the surface of a plate material, and a camera that captures the linear pattern projected on the plate material, and an image of the linear pattern captured by the camera A flatness measurement method for a plate material for analyzing and measuring the flatness of the plate material,
Detecting the edge of the plate by making the size of the linear pattern larger than the plate in the longitudinal direction of the line;
A step of setting a shape measurement line that crosses the line on the surface of the plate material with the detected edge of the plate material as a reference ;
The surface angle distribution of the plate material is compared by comparing the interval between the adjacent lines at the shape measurement line in the plate material with the interval between the adjacent line at the shape measurement line in a flat reference plate measured in advance. And the process of calculating
Have
Calculating the surface angle distribution comprises:
Obtaining a spatial frequency of the linear pattern by converting the luminance distribution of the linear pattern in the shape measurement line into a spatial frequency domain by Fourier transform;
Calculating the reciprocal of the spatial frequency to determine the line spacing;
A method for measuring the flatness of a plate material. Measuring the flatness of the plate material while rolling the plate material in a rolling line,
Projecting the linear pattern so that the longitudinal direction of the parallel lines intersects the rolling direction of the plate,
The flatness measurement method for a plate material according to claim 1, wherein the shape measurement line is set in a rolling direction of the plate material. The number of data of the luminance distribution, 2 ⁿ by resampling ⁽ⁿ is. A natural number) and the number data of claim 1 or by fast Fourier transform and converting the luminance distribution to the spatial frequency domain The flatness measurement method of the board | plate material of 2 . The luminance distribution is a luminance distribution obtained by combining the luminance distribution obtained by inverting the luminance distribution with respect to the longitudinal direction of the shape measurement line, and / or the spatial frequency is extracted from the frequency band of the spatial frequency. And the flatness measuring method of the board | plate material as described in any one of Claims 1-3 which has the process made into the spatial frequency which moved the said frequency band to the low frequency side. It has multiple cameras with different sensitivity and / or shooting angle,
In measuring the flatness of the plate material from a plurality of images taken by the camera,
Among at least one of the plurality of luminance distributions in the same shape measurement line by the plurality of images, when at least one of the luminance distributions has a saturated portion of luminance, the luminance distribution with a small number of saturated portions is reduced. The flatness measurement method for a plate material according to any one of claims 1 to 4 , wherein when any of the luminance distributions is not saturated, the luminance distribution having the highest luminance is used. . For the reference plate, a step of measuring an interval between adjacent lines in the shape measurement line at a plurality of heights in advance;
Measuring the height of the plate,
The flatness measurement method for a plate material according to any one of claims 1 to 5 , further comprising a step of determining an interval between the lines of the reference plate to be used from the measured height of the plate material. . Projecting a pattern for measuring the height of the plate on the surface of the plate, from the position of the pattern in the image captured by the camera in claim 6, characterized in that to measure the height of the plate The flatness measuring method of the board | plate material as described. The plate material flatness measurement method according to any one of claims 1 to 7 , wherein the number of the lines of the linear pattern photographed by the camera is 40 to 100. The flatness measurement method for a plate material according to any one of claims 1 to 8 , wherein an exposure time dt of the camera is obtained and set by the following equation (a).

However, Ps ₀ ′: the size of one line in the width direction of the line and one dark part adjacent to the line when the plate is horizontal
n: Ratio of dark part in the size Ps ₀ ′
α: Shooting angle of the camera with respect to the vertical direction
β: Projection angle of the projector with respect to the vertical direction
V: Speed of the plate material
θ _max : Maximum surface angle with respect to the horizontal direction of the plate A projector that projects a linear pattern of parallel lines onto the surface of the plate,
A plurality of cameras having different sensitivities and / or observation angles for photographing a linear pattern projected on the plate material ;
The flatness measuring apparatus of the board | plate material which measures the flatness of the said board | plate material with the flatness measuring method of the board | plate material as described in any one of Claims 1-9 . 11. The flatness measuring apparatus for a plate material according to claim 10 , wherein the projector includes a light source having a wavelength of 360 to 560 nm as a maximum luminance, and the camera includes a light transmission filter having a wavelength of 360 to 560 nm. The flatness of the plate material according to claim 10 or 11 , further comprising a storage medium that uses a flat reference plate and stores measurement values of the line intervals measured in advance at a plurality of heights with respect to the reference plate. measuring device. In a finishing rolling mill row comprising a plurality of finishing rolling mills in series, finishing rolling comprising the plate material flatness measuring device according to any one of claims 10 to 12 between the finishing rolling mills. Machine train. A method for manufacturing a plate material, comprising the step of measuring the flatness of the plate material using the method for measuring flatness of a plate material according to any one of claims 1 to 9 . The manufacturing method of the board | plate material characterized by having the process of measuring the flatness of the said board | plate material using the flatness measuring apparatus of the board | plate material as described in any one of Claims 10-12 .

Images