MÉTODO DE MEDIR FORMATO DE OBJETO SEMELHANTE A TIRA, MÉTODO DE CONTROLE DE FORMATO DE OBJETO SEMELHANTE A TIRA, MÉTODO DE FABRICAR OBJETO SEMELHANTE A TIRA, MÉTODO DE GERENCIAMENTO DE QUALIDADE DE OBJETO SEMELHANTE A TIRA, APARELHO DE MEDIR FORMATO DE OBJETO SEMELHANTE A TIRA E EQUIPAMENTO DE FABRICAR OBJETO SEMELHANTE A TIRA

Abstract

This method for measuring the shape of a belt-like object is for measuring the shape of a belt-like object, and includes: an imaging step for capturing an image of the thermal radiation of a belt-like object such that, when designating a reference plane on the surface of the belt-like object as a plane α, an angle φ formed between an orthographic projection of the optical axis of a camera onto the plane α and a conveyance direction p of the belt-like object does not reach 90 degrees; and an image processing step for calculating an index of the surface shape of the belt-like object on the basis of the radiation luminance of the belt-like object in the obtained image.

Description

1 / 38 “METHOD FOR MEASURING THE SHAPE OF A STRIP-LIKE OBJECT, METHOD FOR CONTROLLING THE SHAPE OF A STRIP-LIKE OBJECT, METHOD FOR MANUFACTURING A STRIP-LIKE OBJECT, METHOD FOR MANAGING THE QUALITY OF A STRIP-LIKE OBJECT, MEASURING APPARATUS "STRIP-LIKE OBJECT FORMAT AND EQUIPMENT FOR MANUFACTURING STRIP-LIKE OBJECTS" Field

[001] The present invention relates to a method of measuring the shape of a strip-like object, a method of controlling the shape of a strip-like object, a method of manufacturing a strip-like object, a method of quality management of a strip-like object, an apparatus for measuring the shape of a strip-like object, and equipment for manufacturing a strip-like object. Background

[002] In the materials industry, the shape management of strip-like materials is important, and the quantification of product shapes has been required. For example, shape measurement of steel materials in a steelmaking process is rigorously demanded from the point of view of operational stability, product quality assurance, and the like. In particular, shape measurement during rolling to produce a product with a target shape is important because it leads to improvements in product quality and operational stabilization by initial setting of rolling conditions and feedback for rolling control during rolling.

[003] For example, in a hot-rolled steel sheet manufacturing line, a rectangular, parallelepiped-shaped semi-finished product called a slab, in a high-temperature state extracted from a heating furnace, is processed into a sheet shape through the steps of sizing, hot rolling, and finishing rolling, and coiled for Petition 870250081021, dated 09 / 09 / 2025, page 10 / 81 2 / 38 being a coil as a product. At this point, depending on the state of the lamination, the reduction of the lamination may become uneven and the steel sheet may partially extend in one width direction, resulting in a shape defect.

[004] For example, when compared to a central portion (central portion in the width direction) of the steel sheet, when only edge portions (end portions in the width direction) are extended, the edge portions will have a wavy shape. On the other hand, when only the central portion is extended, the central portion will have a wavy shape. The above shape defect not only causes a product defect but also deteriorates sheet workability in subsequent processes such as pickling and cold rolling, causing a problem. Therefore, it is strictly required to avoid the shape defect.

[005] In order to improve the shape defect, it is necessary to properly apply a load in the width direction during rolling, but there are several disturbances such as wear of a rolling roll, temperature distribution of the steel sheet, and variation in the characteristic distribution of the material. It is extremely difficult to obtain ideal rolling conditions by calculation alone. Therefore, in order to establish ideal rolling conditions, it is essential to perform pre-established control by conforming a steel sheet shape measurement to a rolling model to derive an ideal initial configuration, perform real-time feedback control of rolling conditions, and the like. By performing the above controls, the shape defect can finally be eliminated.

[006] However, it is difficult to measure the shape of the steel sheet after the steel sheet is formed into a roll-shaped product. In particular, it is necessary to measure the shape during rolling under feedback control. Therefore, the shape is preferably measured immediately after leaving a rolling mill. Petition 870250081021, dated 09 / 09 / 2025, page 11 / 81 3 / 38 finishing, which is the final rolling process. For example, when the extension of the central portion is significant relative to the edge portions of the steel sheet, bend control is performed to eliminate extension concentration in the central portion while allowing extension in the edge portions by adjusting a load balance in the width direction of the rolling mill roll. The format described in this document refers to sheet extension in a longitudinal direction that is generated locally in the width direction of the steel sheet, mainly at the time of rolling the steel sheet, and specifically includes extension in the central portion and extension in the edge portions.

[007] As a technique for measuring the shape of steel sheet immediately after leaving the finishing lamination, several techniques such as a method using a rod-like light source, or a magnetic sensor have been proposed in the past. As a particularly effective method among the above techniques, for example, Patent Literatures 1 to 3 disclose a method for measuring a shape by irradiating a target surface with a point-like or linear beam of light using a laser and measuring the light reflected therefrom.

[008] Furthermore, Patent Literature 4 discloses a technique of using three line lasers whose longitudinal direction is orthogonal to the transport direction of the steel plate with respect to a laser irradiation system. In the technique disclosed in Patent Literature 4, the laser line is irradiated parallel to the longitudinal direction of the steel plate so as to form three lines at equal intervals, and reflection images of these are acquired to compare profiles of respective lasers, thereby removing a vertical vibration influence from the steel plate.

[009] Furthermore, Patent Literature 5 to 7 and Non-Patent Literature 1 disclose a technique for increasing the number of irradiated lines at a lower cost than a laser by irradiating a target surface with a stripe pattern including Petition 870250081021, dated 09 / 09 / 2025, p. 12 / 81 4 / 38 a plurality of lines using a strong LED light source, thus stably measuring the shape despite a mirrored surface property or tilt of the target.

[010] A demand for measuring the shape of a steel sheet during manufacturing exists not only in hot-rolled steel sheets but also in other strip-like materials regardless of whether they are incandescent (e.g., 600°C or higher), hot (e.g., 300°C to 600°C), or cold (e.g., approximately at normal temperature) states. The term strip-like object as used herein refers to a long material. Examples of strip-like objects include products that are formed into a sheet, plate, frame, board, and rectangular slab, such as a steel plate, as well as finely coiled products such as iron, paper, cloth, or non-ferrous metal such as aluminum. List of Citations Patent Literature

[011] Patent Literature 1: JP S56-124006 A Patent Literature 2: JP S55-40924 A Patent Literature 3: JP S58-11708 A Patent Literature 4: JP S61-40503 A Patent Literature 5: JP 2008-58036 A Patent Literature 6: JP 2011-99821 A Patent Literature 7: JP 2016-65863 A Non-Patented Literature

[012] Non-Patent Literature 1: Isei and three others, Development of a shape gauge employing an LED pattern projection method for hot strip finishing mill, iron and steel, The Iron and Steel Institute of Japan, 2019, Vol. 105, No. 1, pp. 20-29 Non-Patent Literature 2: Kaneshige, Thermal Measurement with Radiation Petition 870250081021, dated 09 / 09 / 2025, p. 13 / 81 5 / 38 thermometer, molding, Japan Society of Polymer Processing, 2020, Vol. 32, No. 4, p. 121 to 124 Non-Patent Literature 3: Inoue, “Study on a method for measuring the non-contact temperature of metal in cold and hot bearings,” Grants-in-Aid for Scientific Research report AF-1997014 Summary Technical Problem

[013] Techniques disclosed in Patent Literatures 1 to 7 are a method for measuring the shape of an object by irradiating a hot-rolled steel sheet with light using a light source and capturing reflected light with a camera. However, for example, considering the influence of high temperature due to radiant heat, vapor adhesion, dust, oil, and the like on the steel sheet, an advanced technique is required to measure stably over a long period of time in a state where the light source and sensor are close to the steel sheet during transport. As a result, maintenance costs will also be high.

[014] In order to prevent this, it is conceivable to place the light source away from the object, but as the light is diffused and the amount of light decreases, it is difficult to design an optical system that stably condenses light in order to ensure the amount of light. Furthermore, although there is a possibility of using a magnetic sensor or similar, it is necessary to bring the sensor itself close to a measurement target, and it is similarly difficult to install the sensor and maintain its performance.

[015] The present invention was made in view of the foregoing, and an object of the present invention is to provide a method for measuring the shape of a strip-like object, a method for controlling the shape of a strip-like object, a method for manufacturing a strip-like object, a method for managing the quality of a strip-like object, a shape-measuring apparatus for a strip-like object, and strip-like object manufacturing equipment with the capability to operate easily and Petition 870250081021, dated 09 / 09 / 2025, p. 14 / 81 6 / 38 stably a light source or sensor on a target that is a portion other than an edge portion of a strip-like object without bringing the light source or sensor close to the target to be measured, and also eliminating a maintenance cost. Solution to the Problem

[016] (1) To solve the problem and achieve the objective, a method for measuring the shape of a strip-like object according to the present invention is the method for measuring the shape of a strip-like object. The method for measuring the shape of a strip-like object includes: a capture step to capture a thermal radiation light image of the strip-like object such that an angle φ formed by orthographic projection of an optical geometric axis of a camera onto a plane α and a transport direction p of the strip-like object does not become 90 degrees, the plane α being a reference plane of a surface of the strip-like object; and an image processing step to calculate an index of a surface shape of the strip-like object based on the radiance of the strip-like object in the image obtained.

[017] (2) Furthermore, in the method of measuring strip-like object shape, according to (1) above, the image processing step may include extracting a linear component from the radiance of the strip-like object, the linear component being parallel to the transport direction of the strip-like object.

[018] (3) Furthermore, in the method of measuring strip-like object shape, according to (1) or (2) above, in the capture step, the plane α and the optical geometric axis of the camera can form an angle θ that is less than or equal to 20 degrees.

[019] (4) Furthermore, in the method of measuring strip-like object shape, according to any one of (1) to (3) above, the image processing step may include calculating, as the index, any one or more of a slope, a wave height, a wave pitch, an extension amount, Petition 870250081021, dated 09 / 09 / 2025, page 15 / 81 7 / 38 and a percentage of extension at each position in a width direction on the surface of the strip-like object from a bright-dark pattern of the strip-like object obtained.

[020] (5) In addition, a method for controlling the shape of a strip-like object according to the present invention includes: measuring the shape of the strip-like object by the method of measuring the shape of a strip-like object, according to any one of (1) to (4) above; and controlling the shape of the strip-like object, based on a measurement result, to have a desired shape.

[021] (6) In addition, a method for manufacturing a strip-like object according to the present invention includes: measuring a shape of the strip-like object by the method of measuring the shape of a strip-like object, according to any one of (1) to (4) above; and manufacturing the strip-like object based on a measurement result.

[022] (7) In addition, a method for managing a quality of a strip-like object according to the present invention includes: measuring a shape of the strip-like object by the method of measuring the shape of a strip-like object, according to any one of (1) to (4) above; and managing the quality of the strip-like object based on a measurement result.

[023] (8) In addition, a strip-like object shape measuring apparatus according to the present invention is the apparatus that measures the shape of a strip-like object. The strip-like object shape measuring apparatus includes: an image capture unit configured to capture a thermal radiation light image of the strip-like object such that an angle φ formed by orthographic projection of an optical geometric axis of a camera onto a plane α and a transport direction p of the strip-like object does not become 90 degrees, the plane α being a reference plane of a surface of the strip-like object; and an image processing unit configured to calculate an index of Petition 870250081021, dated 09 / 09 / 2025, page 16 / 81 8 / 38 a strip-like object surface shape based on the radiance of the strip-like object in the obtained image.

[024] (9) In addition, the strip-like object making equipment according to the present invention includes the strip-like object shape measuring apparatus according to (8) above. Advantageous Effects of the Invention

[025] According to a method for measuring the shape of a strip-like object, a method for controlling the shape of a strip-like object, a method for manufacturing a strip-like object, a method for managing the quality of a strip-like object, a device for measuring the shape of a strip-like object, and equipment for manufacturing a strip-like object according to the present invention, it is possible to operate a light source or a sensor easily and stably without bringing the light source or the sensor close to a measurement target. It is also possible to eliminate a maintenance cost. Brief Description of the Drawings

[026] Figure 1 is a diagram illustrating a schematic configuration of a strip-shaped object measuring apparatus according to an embodiment of the present invention.

[027] Figure 2 is a diagram illustrating an example of a state in which a change in slope due to sheet elongation is captured as a bright-dark pattern in a steel sheet immediately after hot finish rolling.

[028] Figure 3 is a diagram illustrating an example of a state of an experiment to investigate angular emissivity characteristics of a steel sheet surface.

[029] Figure 4 is a graph illustrating an example of a relationship between radiance and a light reception angle on the surface of a steel sheet. Petition 870250081021, dated 09 / 09 / 2025, page 17 / 81 9 / 38

[030] Figure 5 is a diagram illustrating a mechanism for generating the bright-dark pattern on the steel sheet.

[031] Figure 6 is a diagram illustrating an example of a camera arrangement to capture sheet extension of steel sheet as the bright-dark pattern, in which (a) is a diagram illustrating the positional relationship viewed obliquely, (b) is a diagram illustrating the positional relationship viewed from above, and (c) is a diagram illustrating the positional relationship viewed from a direction δn in (b).

[032] Figure 7 is a diagram illustrating an example of a steel sheet image in a normal state and a shape defect state.

[033] Figure 8 is a flowchart illustrating a specific process flow of an image processing step by an image processing unit of a strip-like object shape measuring device according to an embodiment of the present invention.

[034] Figure 9 is a diagram illustrating a binarization process in the image processing step of a method for measuring strip-like object shape according to the embodiment of the present invention.

[035] Figure 10 is a diagram illustrating a rotation process in the image processing step of the method for measuring strip-like object shape, according to the embodiment of the present invention.

[036] Figure 11 is a diagram illustrating a luminance profile extraction process in the image processing step of the strip-like object shape measurement method, according to the embodiment of the present invention.

[037] Figure 12 is a diagram illustrating a steel sheet cut surface by a plane δ when a normal line n of a transport table is a geometric axis Y and a direction orthogonal to the geometric axis Y in the plane δ is a geometric axis X. Petition 870250081021, dated 09 / 09 / 2025, page 18 / 81 10 / 38

[038] Figure 13 is a graph illustrating an example of a relationship between radiance and the angle of reception of light on the surface of a steel sheet.

[039] Figure 14 is a graph illustrating an example of a relationship between slope and a bright-dark change rate.

[040] Figure 15 is a diagram illustrating an example of an image of the steel plate captured by a camera.

[041] Figure 16 is a diagram illustrating an example of a case where it is difficult to distinguish a sheet steel surface from a structure.

[042] Figure 17 is a diagram illustrating an example of a low-pass filter process performed when it is difficult to distinguish between the steel sheet surface and the structure in the method of measuring the shape of a strip-like object, according to the embodiment of the present invention.

[043] Figure 18 is an example of an application where the strip-like object shape measuring device according to the embodiment of the present invention is applied for shape acceptance determination.

[044] Figure 19 illustrates an example of the application when the strip-like object shape measuring device according to the embodiment of the present invention is applied for lamination feedback control.

[045] Figure 20 illustrates an example of the application when the strip-like object shape measuring apparatus according to the embodiment of the present invention is applied for lamination control using machine learning. Description of the Modalities

[046] A method for measuring the shape of a strip-like object, a method for controlling the shape of a strip-like object, a method for manufacturing a strip-like object, a method for quality management of a strip-like object, an apparatus for measuring the shape of a strip-like object, and equipment for manufacturing a strip-like object according to an embodiment of the present invention. Petition 870250081021, dated 09 / 09 / 2025, p. 19 / 81 11 / 38 The invention will be described with reference to the drawings. Note that the components in the following embodiment include those that can be easily replaced by those who are skilled in the art or those who are substantially skilled as well. (Format measuring device)

[047] A shape measuring apparatus for strip-like objects according to the embodiment will be described with reference to Figures 1 to 17. The shape measuring apparatus is an apparatus for measuring the shape of a strip-like object. An instance where the shape measuring apparatus is applied to hot finish rolling will be described below. Furthermore, an instance where the strip-like object to be measured is a steel sheet will be described below. Additionally, an instance where a shape measured by the shape measuring apparatus is an extension of a portion other than an edge portion of the steel sheet, specifically, the extension of a central portion of the steel sheet, will be described below.

[048] As illustrated in Figure 1, the shape measuring device according to the modality includes a camera 2 and an image processing unit 3. First, details of camera 2 will be described.

[049] As will be described later, camera 2 is arranged so that an angle φ formed by orthographic projection of an optical geometric axis of the camera onto plane α and a transport direction p of the strip-like object (steel sheet S) does not become 90 degrees (see Figure 6). Here, plane α refers to a reference plane of a surface of the strip-like object. Using camera 2 arranged in this way, a thermal radiation light image of steel sheet S on a hot finish rolling output side of steel sheet S rolled by rolling mill 1 is captured.

[050] When a digital camera or similar device is sold to the public in Petition 870250081021, dated 09 / 09 / 2025, page 20 / 81 12 / 38 general is used as camera 2, problems such as blurring or roughness of an image occur, and the shape of the S steel plate cannot be captured clearly. Therefore, technical points (1) to (4) for accurately capturing the shape of the S steel plate by camera 2 will be described below. (1) Ensure sufficient light quantity during imaging

[051] Firstly, a sufficient amount of light is ensured to obtain a clear image. On the hot-finished lamination exit side, the S-steel sheet can pass at a speed of, for example, 20 m / s or more, and thus the speed is very high. In order to clearly capture a high-speed object without blurring, it is preferable to shorten the exposure time. For example, when imaging is performed with a resolution of about 2 mm at 20 m / s and a pixel blur within 1 pixel, i.e., within 2 mm, a permissible exposure time is only 0.1 ms, and the amount of light received proportional to the exposure time is also very small.

[052] Furthermore, a steel sheet temperature on the exit side of hot finish rolling is about 900°C. As illustrated in Figure 5 of Non-Patent Literature 2, a peak wavelength of thermal radiation light at about 900°C (1200 K) is 2.5 μm, and sensitivity is low in the visible region (0.4 to 0.7 μm). Therefore, it is preferable to use a camera including an imaging element such as InGaAs, PbS, or PbSe as camera 2.

[053] On the other hand, a camera using an imaging element such as InGaAs, PbS, or PbSe is expensive. Moreover, the price becomes higher according to the resolution. Thus, to capture a high-resolution image, the inclusion cost increases. Therefore, like camera 2, it is preferable to use a camera including an inexpensive Si imaging element capable of using a near-infrared sensitivity region of 0.8 to 1.0 pm. Using such a camera 2, it is possible to obtain a sufficient amount of radiated light. Petition 870250081021, dated 09 / 09 / 2025, page 21 / 81 13 / 38 low-cost thermal

[054] Note that when the present method is applied to an object at a lower temperature, it is preferable to use camera 2 including an imaging element having sensitivity on the longer wavelength side. For example, when a measurement target is about 400°C, camera 2 including an InGaAs imaging element having sensitivity to 1.2 to 1.7 μm can be used without considering the inclusion cost. Furthermore, when the measurement target is about 200°C, camera 2 including an imaging element such as PbS or PbSe having sensitivity to 3 to 5 μm can be used. In this way, a sufficient amount of light can be ensured. Thus, it is preferable to select camera 2 including the imaging element having sensitivity to a suitable wavelength range according to the transport speed, resolution, temperature, depth of field (to be described later), and similar factors of the measurement target. (2) Positional relationship between steel plate and camera

[055] Secondly, in order to clearly image the shape of the steel plate S, a positional relationship between the steel plate S to be measured and camera 2 is examined. Here, it is considered to focus on a slight change in inclination in the longitudinal direction due to elongation that appears on a plate surface when plate elongation (specifically, elongation of a portion other than the edge portion of the strip-like object, and more particularly, elongation of the central portion) occurs, and to capture the change in inclination based on thermal radiation light due to heating.

[056] The inventors repeatedly imaged an incandescent steel plate S under various optical conditions to gain knowledge and study through heated imaging testing in a laboratory. As a result, the inventors found that a slight inclination on the surface of the steel plate S Petition 870250081021, dated 09 / 09 / 2025, page 22 / 81 14 / 38 can be captured as a change in luminance in an image by imaging the steel sheet S at a low angle (light receiving angle) formed between the reference plane of the steel sheet S surface and the optical geometric axis of camera 2. For example, Figure 2 illustrates a state in which a change in inclination due to sheet elongation is captured as a bright-dark pattern on the glowing steel sheet S immediately after hot finish rolling.

[057] Figure 3 illustrates an experiment in which an angular characteristic of the emissivity of the steel sheet S surface is investigated. In the shape measuring apparatus according to the modality, an attempt was made to simulate a heating process by an apparatus in Figure 3 in order to capture an image of thermal radiation light due to heating of the steel sheet S to be measured. Therefore, as illustrated in Figure 3, a sample of steel sheet SA cut into small pieces was placed on a heater 4, and an image was captured by camera 2. In addition, a biasing mechanism 5 used has the capacity to reduce an angle (light reception angle) formed between the reference plane of the surface of the steel sheet SA sample and the optical geometric axis of camera 2 while maintaining the surface of the steel sheet SA sample that is the measurement target as a visual field.As a result, although the SA steel sheet sample was heated by heater 4, the thermal radiation light due to heating of the SA steel sheet sample was stably imaged at an arbitrary light-receiving angle θ (vertical is defined as 90 degrees).

[058] Using the apparatus illustrated in Figure 3, thermal radiation light due to heating of the SA steel sheet sample was repeatedly imaged at each light reception angle θ, in order to investigate the relationship between radiance and the light reception angle θ. Figure 4 illustrates the relationship between radiance and the light reception angle θ obtained. In Figure 4, a vertical geometric axis Petition 870250081021, dated 09 / 09 / 2025, page 23 / 81 15 / 38 represents the radiance, and a horizontal geometric axis represents the angle of light reception θ. The radiance is normalized so that 100 is the maximum. Furthermore, since radiance and emissivity are in a proportional relationship, radiance can subsequently be referred to as emissivity.

[059] As illustrated in Figure 4, it can be seen that the emissivity is constant from immediately above the SA steel sheet sample (light receiving angle θ = 90 degrees) up to about 60 degrees, but after that, as the light receiving angle θ decreases, the emissivity decreases rapidly. A mechanism for generating the bright-dark pattern studied based on the data in Figure 4 will be described with reference to Figure 5.

[060] When imaging is performed at a low light reception angle θ, a slight change in inclination occurs on the surface according to an extension shape due to the occurrence of extension in the longitudinal direction of the steel plate S. As illustrated in (a) of Figure 5, the optical geometric axis of camera 2 rises on the front side of a sloped part, and thus the light reception angle θ increases. Conversely, as illustrated in (b) of Figure 5, the optical geometric axis of camera 2 rests on the back side of the sloped part, and thus the light reception angle θ decreases.

[061] According to the data on the angle dependence of thermal radiation light emissivity due to heating illustrated in Figure 4, radiance increases as the light receiving angle θ increases, and radiance decreases as the light receiving angle θ decreases at low angles (e.g., 0 degrees to less than 60 degrees). Therefore, when the steel plate S is imaged by camera 2, the bright-dark pattern is formed in the image.

[062] As described in Non-Patent Literature 3, the angle dependence of thermal radiation light emissivity due to heating can theoretically be described by a complex refractive index of the target. However, a Petition 870250081021, dated 09 / 09 / 2025, page 24 / 81 The 16 / 38 model described in Non-Patent Literature 3 has a problem because surface roughness, a coating film, and the like are not considered. Note that this phenomenon occurs similarly not only in the S-steel sheet but also in strip-like objects made of various metal and non-metal materials, and thus it is possible to accurately capture a slight variation in inclination also in strip-like objects other than the S-steel sheet.

[063] Furthermore, the angle dependence of the emissivity of thermal radiation light due to heating does not depend much on the wavelength of light reception. Therefore, it is preferable to select camera 2 according to the temperature of the steel plate S to be measured. When selecting camera 2, it is preferable to consider whether or not the imaging element included in the selected camera 2 can image, with respect to temperature, a wavelength that can ensure a quantity of light or a wavelength with less disturbance. In addition, there is a correlation between the intensity of the bright-dark pattern and the amount of change in inclination, and as will be described later, a degree of plate elongation (specifically, elongation of the non-edge portion of the strip-like object, and more particularly, elongation of the central portion) can also be used as an index.

[064] Subsequently, based on the mechanism above, optical conditions under which the bright-dark pattern can be precisely observed will be examined. Figure 6 illustrates an example of camera 2 arrangement for capturing the plate extension as the bright-dark pattern. In Figure 6, (a) is a diagram of the positional relationship viewed obliquely, (b) is a diagram of the positional relationship viewed from above, and (c) is a diagram of the positional relationship viewed from a direction δn in (b). Here, the direction δn is a Petition 870250081021, dated 09 / 09 / 2025, page 25 / 81 17 / 38 direction in which the optical geometric axis of camera 2 can be viewed from the right side. By defining the direction δn in this way, a magnitude of the light reception angle θ to be described later can be precisely visualized.

[065] In (a), (b), and (c) of Figure 6, plane α is parallel to a transport table of steel plate S, and plane β includes the transport direction p of steel plate S and a normal line n of the transport table. Furthermore, an angle θ (light receiving angle θ) is formed by the optical geometric axis of camera 2 and plane α, and an angle φ is formed by orthographic projection of the optical geometric axis of camera 2 onto plane α and the transport direction p. Plane α includes a surface of the strip-like object (steel plate S) in a state where the strip-like object (steel plate S) is stably transported in a substantially flat shape. Therefore, plane α is the same plane as the reference plane in Figure 3 and is the reference plane of the surface of the strip-like object (steel plate S).

[066] Furthermore, a normal line direction n of the transport table is the same as the normal direction of the surface of the steel plate S in a state where the transported steel plate S is stably transported in a substantially flat shape. Therefore, plane α includes the transport direction p of the strip-like object (steel plate S) and is perpendicular to the normal direction of the surface of the strip-like object (steel plate S). Furthermore, plane β is also a plane including the transport direction p of the strip-like object (steel plate S) and the normal direction of the surface of the strip-like object (steel plate S). Furthermore, in Figure 6, a point 0 is an intersection of plane α and the optical geometric axis. Furthermore, a plane γ is a plane perpendicular to the transport direction of the steel plate. In other words, since the angle φ described above approaches 90 degrees, plane γ and the optical geometric axis are nearly parallel.

[067] First, the light reception angle θ is established, which is an angle formed between the plane α of the reference plane of the plate surface. Petition 870250081021, dated 09 / 09 / 2025, page 26 / 81 18 / 38 steel plate S and the optical geometric axis of camera 2. The light receiving angle θ can be appropriately established according to a purpose or location where the shape measurement is performed. For example, as described above, the light receiving angle θ can be established from the relationship between the thermal radiation light due to heating and the light receiving angle θ obtained by the experiment. When an experimental result in Figure 4 is used, it is desirable that the light receiving angle θ, which is an angle formed by the plane α of the reference plane of the steel plate S surface and the optical geometric axis of camera 2, be 20 degrees or less.

[068] The inclination of the surface of the steel plate S due to plate extension occurs in the longitudinal direction of the steel plate S, that is, a direction parallel to the transport direction p. Therefore, when imaging is performed so that the optical geometric axis of camera 2 is in a direction perpendicular to the transport direction of the steel plate S, that is, a direction in which the angle φ is 90 degrees, it is very difficult to capture an image with inclination variation such as the bright-dark pattern. Therefore, the optical geometric axis of camera 2 is arranged so that the angle φ does not become 90 degrees.

[069] Furthermore, the angle φ is desirably 45 degrees or less. In addition, the angle φ is more preferably 0 degrees. On the other hand, when the angle φ approaches 0 degrees, the camera installation position approaches a transmission line. Therefore, the installation environment deteriorates due to high temperature or similar effects caused by radiant heat. Since it is necessary to perform engineering that enables stable operation even in an unsuitable environment, the angle φ can be made as small as possible by taking a distance from the transmission line to a point where the camera is not affected by thermal radiation or similar effects.

[070] A condition in which the bright-dark pattern with respect to Petition 870250081021, dated 09 / 09 / 2025, page 27 / 81 19 / 38 plate extension (specifically, extension of a portion other than the edge portion of the strip-like object, and more particularly, extension of the central portion) is most prominent when imaging from the same direction or the opposite direction of the transport direction of the steel plate S. However, in this case, it is necessary to install camera 2 immediately above the line or in the vicinity of the line. As a result, camera 2 is installed in an unsuitable environment exposed to fog, dust, high temperature, and the like, which is not preferable.

[071] Furthermore, as illustrated in Figure 6, by imaging the steel plate S while viewing the steel plate S at a low angle relative to the light-receiving angle θ, the distance between both ends of the steel plate S in the image is reduced, and an image of both edge portions can be acquired in a smaller field of view. In addition, the smaller the target field of view, the smaller the image size, and this is advantageous when high-speed transport of the steel plate S to be measured requires high-speed imaging or image processing. (3) Distance between steel plate and camera

[072] Thirdly, the distance between the steel plate S to be measured and camera 2 is increased. This is because, by separating camera 2 from the manufacturing line as much as possible, camera 2 can be installed in a favorable environment, and a difference in optical conditions such as the light reception angle θ and the angle φ in the width direction of the plate, i.e., on both edge portions, of the steel plate S can be reduced. When there is a difference in optical conditions in the width direction of the steel plate S, this difference appears as a difference in appearance. As a result, a shape defect level may be wrongly judged when the determination is made visually. A greater correction is required when the shape defect level is quantified by Petition 870250081021, dated 09 / 09 / 2025, page 28 / 81 20 / 38 image processing. In order to increase the distance, it is preferable to use a telephoto lens for imaging. By using a telephoto lens, a difference in optical conditions can be reduced. (4) Establishing the appropriate diaphragm value

[073] Fourthly, in order to ensure depth of field, a diaphragm value of camera 2 is appropriately determined. For example, it is assumed that the entire surface of the steel plate S is imaged through both the edge portions on the front side (bottom side) and the back side (top side) of the steel plate S. In this case, when a sheet width of the steel plate S is d (mm), a difference ΔL (mm) between a distance from camera 2 to the edge portion on the front side and a distance from camera 2 to the edge portion on the back side can be expressed by Formula (1) below. ΔL = dcosθ / sinφ ··· (1)

[074] In order to acquire an image in which both edge portions of the steel plate S are in focus, at least the depth of field (strip in focus) needs to be ΔLd (mm) or more. As the depth of field increases as the aperture value increases, it is preferable to set the aperture to ΔLd (mm) or more. However, when the aperture value increases, a quantity of light becomes insufficient, and there is a concern that blurring may occur. Therefore, it is preferable to determine the aperture value and a focal length that can balance blurring and depth of field according to the transport speed of the steel plate S to be measured and the required resolution. Next, image processing unit 3 will be described.

[075] The image processing unit 3 is implemented, for example, by a general-purpose computer such as a workstation or a personal computer. The image processing unit 3 can be installed near the camera 2, or it can be installed in a cloud if high speed is not possible. Petition 870250081021, dated 09 / 09 / 2025, page 29 / 81 21 / 38 required.

[076] As described below, image processing unit 3 calculates a surface shape index of steel plate S based on the radiance of steel plate S in an image captured and obtained by camera 2. As described below, image processing unit 3 appropriately selects, from the bright-dark pattern of steel plate S, any one or more of the slope, wave height, wave pitch, extension amount, and extension percentage at each position in the width direction on the surface of steel plate S, and calculates the index thereof. Subsequently, a process for calculating the steel plate S shape index from the image obtained by camera 2 using image processing unit 3 will be described.

[077] An image of steel plate S in a normal state is illustrated in (a) of Figure 7, and an image of steel plate S when a shape defect occurs is illustrated in (b) of Figure 7. As illustrated in Figure 7, a plate elongation state can be visually determined. Image processing unit 3 calculates the shape index of steel plate S from the image of steel plate S acquired according to, for example, a procedure illustrated in Figure 8. In the following description, with respect to a vertical geometric axis and a horizontal geometric axis of the image, a geometric axis closer to parallel with the transport direction of steel plate S is defined as the horizontal geometric axis for convenience of explanation.

[078] First, as illustrated in Figure 9, only a region of sheet metal from steel sheet S is extracted by binarization (Step S1 in Figure 8). An image before binarization is illustrated in (a) of Figure 9, and an image after binarization is illustrated in (b) of Figure 9.

[079] At this moment, there may be a case in which a disturbance occurs, such Petition 870250081021, dated 09 / 09 / 2025, page 30 / 81 22 / 38 such as a local decrease in luminance on the sheet surface due to water on the sheet surface, small cooling water droplets, or similar, or glare due to thermal radiation light scattered from the steel sheet S caused by small droplets sprayed into a space above the sheet. As these disturbances are often in a high-frequency strip and weak relative to a step of the shape to be calculated, it is desirable to remove connected / isolated points by expansion / contraction processing, a median filter or similar (Step S2 in Figure 8) in the extracted sheet region in a binarization step.

[080] Furthermore, there is a case where a plurality of bubbles (identified protrusions connecting surrounding pixels when binarized) that will be candidates for the sheet region is generated after removing the connection / isolated point. In this case, the sheet region can be determined by extracting a bubble (Step S3 in Figure 8) or by determining the sheet region based on a size, direction, and the like of the steel sheet S.

[081] Subsequently, for the steel plate region S obtained as described above, the bright-dark pattern generated perpendicular to the transport direction of the steel plate S is extracted. Note that extraction of the bright-dark pattern refers specifically to the calculation of a luminance profile. At this point, in order to easily extract the bright-dark pattern, for example, the image is rotated so that the transport direction of the steel plate S coincides with the vertical geometric axis or the horizontal geometric axis of the image (Step S4 in Figure 8), as illustrated in Figure 10. An image before rotation is illustrated in (a) of Figure 10, and an image after rotation is illustrated in (b) of Figure 10.

[082] At this point, it is necessary to determine the transport direction of the steel plate S. However, as long as the visual field of camera 2 does not change, the transport direction can be a direction fixed in advance or it can be Petition 870250081021, dated 09 / 09 / 2025, page 31 / 81 23 / 38 automatically calculated from the image. For example, a direction obtained by linearly adjusting the upper and lower contours of the plate region and calculating an average bias can be established as the transport direction of the steel plate S. When the plate region is considerably long in the longitudinal direction, elliptical approximation can be performed on the bubble of the plate region, and a direction of a principal geometric axis obtained can be established as the transport direction. A similar result can also be obtained by calculating a direction in which the Feret diameter is maximized with respect to the bubble described above.

[083] Subsequently, a vertical direction of the obtained image is established as a width direction of the plate, and positions of upper and lower edge portions are extracted (Step S5 in Figure 8).

[084] Subsequently, a luminance profile at a predetermined position in the width direction of the steel sheet S is calculated (Step S6 in Figure 8). In this step, for example, the luminance profile is calculated by extracting linear components parallel to the transport direction of the steel sheet S in the radiance of the steel sheet S, as illustrated in Figure 11. An image before the luminance profile is calculated is illustrated in (a) of Figure 11, and the calculated luminance profile is illustrated in (b) of Figure 11. In the case of the present embodiment, the transport direction p of the steel sheet S coincides with a rolling direction of the steel sheet S.

[085] In this step, coordinates of the upper and lower edge portions of the steel plate S are first calculated, and then coordinates in the width direction to calculate the luminance profile are calculated from the relative positions of both coordinates. For example, when y1 is the coordinate of the upper edge portion of the steel plate S and y2 is the coordinate of the lower edge portion of the steel plate S, a coordinate at the center in the width direction (center Petition 870250081021, dated 09 / 09 / 2025, page 32 / 81 The radiance of the 24 / 38 sheet can be calculated by (y1 + y2) / 2. As a result, the linear components parallel to the transport direction of the steel sheet S at the center in the width direction of the steel sheet S can be extracted from the radiance of the steel sheet S.

[086] Furthermore, when it is desired to calculate the luminance profile at a position α% from the upper edge portion of the steel plate S, a coordinate at that position can be calculated by (1 - a)y1 + ay2). As a result, it is possible to extract the linear components parallel to the transport direction of the steel plate S at a predetermined position in the width direction of the steel plate S from the radiance of the steel plate S.

[087] When the luminance profile at the predetermined position in the width direction of the steel plate S is calculated, local noise such as droplet noise may be added. Therefore, not only the luminance profile of a line at the predetermined position in the width direction of the steel plate S but also luminance profiles of several upper and lower lines can be extracted and divided proportionally in the longitudinal direction of the image, in order to reduce noise influence.

[088] Subsequently, an index of the steel plate S shape is calculated from the luminance profile obtained (Step S7 in Figure 8). As illustrated in Figure 6 described above, when the surface of the steel plate S is imaged from a direction at a low angle relative to the steel plate S and from a direction inclined relative to the transport direction of the steel plate S, the bright-dark pattern is generated according to a change in inclination caused by the shape. The intensity of the bright-dark pattern increases as the inclination changes widely. Therefore, by indexing the intensity of the bright-dark pattern caused by the shape, it is possible to index a degree of plate extension at the predetermined position in the width direction of the steel plate S. Petition 870250081021, dated 09 / 09 / 2025, page 33 / 81 25 / 38

[089] Furthermore, since the wave pitch of a wave generated in the shape of the steel sheet S depends on the diameter of the rolling mill roll 1, the wave pitch is within a certain range. Therefore, by performing frequency analysis and extracting only frequency components caused by sheet extension, it is possible to reduce high frequency or periodic noise. For frequency analysis, a low-pass filter such as a mean motion filter can be used, or Fourier transform, inverse Fourier transform, or similar can be used. Furthermore, as a high-pass filter, a frequency filter similar to the low-pass filter can be used, or a zero-order difference and first-order components can be taken after polynomial approximation, such as taking a difference after linear approximation.Thus, an index of the steel sheet shape S is calculated from the luminance profile obtained by extracting only one extension of the sheet component.

[090] A method for calculating the steel sheet format index S can be any method provided that the intensity of the bright-dark pattern can be quantified, and thus several methods are applicable. For example, methods (i) to (iii) described below can be used. (i) A difference or ratio between the maximum luminance and the minimum luminance is extracted and used as an index. A method for calculating the maximum and minimum luminance is not necessarily the maximum and minimum values, and may be a percentage or similar in which values ​​are ranked in descending order within an applicable range to extract the top 10%. (ii) An average luminance profile value is corrected to 0, and its variance or standard deviation is used as an index. (iii) A conversion table or formula is created by experimentally or theoretically associating a relationship between the intensity (maximum value, minimum value, or standard deviation) of the bright-dark pattern and the index of the format. Petition 870250081021, dated 09 / 09 / 2025, page 34 / 81 26 / 38 (slope or extension amount), and applying the conversion table or formula for actual measurement to calculate the steel sheet format index S.

[091] Furthermore, by studying a relationship between an angular characteristic of emissivity and a change in slope on the surface of the steel plate S (subsequently referred to as the steel plate surface) caused by plate extension (specifically, extension of a portion other than the edge portion of the strip-like object, and more particularly, extension of the central portion) in the shape index obtained by (i) above, it is also possible to calculate a slope from the bright-dark pattern based on a physical theory.

[092] First, the relationship between the slope and the change in inclination in the optical geometric axis direction of camera 2 (subsequently referred to as the optical geometric axis direction of the camera) on the surface of the steel plate is formulated. As a premise, the plate extension is a sinusoidal wave, its amplitude is h / 2, and its period is L. Furthermore, the slope λ at that moment is h / L.

[093] Here, in order to consider the variation of inclination of the sheet surface in the direction of the camera's optical geometric axis, a plane δ formed by the normal line n of the transport table and the optical geometric axis is assumed. Then, when the normal line n of the transport table is the geometric axis Y and the direction orthogonal to the geometric axis Y in the plane δ (direction parallel to the plane δ) is the geometric axis X, the surface of the steel sheet is cut as illustrated in Figure 12 by the plane δ and expressed by Formula (2) below. Note that an origin in Figure 12 is a midpoint in the width direction of the steel sheet.

[094] At this moment, when Si is a near lateral edge on the steel plate S with respect to camera 2, and S2 is a far lateral edge, the surface of the steel plate S cut by the plane δ is a line segment S1S2 as illustrated in (b) Petition 870250081021, dated 09 / 09 / 2025, page 35 / 81 27 / 38 of Figure 6. The line segment S1S2 is a sinusoidal wave having a period of φ, and its phase is established at ψ. More specifically, the near-side edge S1 is a point representing an edge on the camera 2 side of a line segment formed by the plane δ and the surface of the steel plate. Furthermore, the back-side edge S2 is a point representing an edge on the opposite side of camera 2 of the line segment formed by the plane δ and the surface of the steel plate. / 1 / Y(X) = -sin ^2π COS0 \ ,x—-ί-Χ + φ) ---(2) I read /

[095] In Formula (2) above, Y is the normal line n of the transport table, X is the direction orthogonal to the geometric axis Y in the plane δ, h / 2 is the amplitude, L is the period, and ψ is an arbitrary phase.

[096] Next, a relationship between the tilt in the direction of the camera's optical geometric axis and brightness will be studied. As illustrated in Figure 4, in low-angle imaging, radiance increases as the angle of light reception increases. In line segment S1S2, when the angle of light reception θ is the largest with respect to the direction of the camera's optical geometric axis, that is, when the tilt in Formula (2) above is the largest, the image becomes the brightest. The tilt at this moment is defined as 02. The maximum tilt of line segment S1S2 is ω2. Similarly, in line segment S1S2, when the angle of light reception θ is the smallest with respect to the direction of the camera's optical geometric axis, that is, when the tilt in Formula (2) above is the smallest, the image becomes the darkest. The tilt at this moment is defined as ωι. The minimum tilt of line segment S1S2 is ω-ι.

[097] In order to obtain the maximum slope ω2 and the minimum slope ωι, Formula (3) below can be obtained by differentiating from Formula (2) above with X. ,, , nh cos φ / cos φ \ Y (X) =--L cos2π——Χ + ψ) --(3)

[098] From Formula (3) obtained above, the maximum slope ω2 can be calculated by Formula (4) below, and the minimum slope ωι can be calculated by Petition 870250081021, dated 09 / 09 / 2025, page 36 / 81 28 / 38 Formula (5) below. Furthermore, as indicated in Formulas (4) and (5) below, absolute values ​​of the maximum slope ω2 and the minimum slope ωι are equal.xnhcosò ... ω2= Γ(-ψ) =--—- ---(4) Lt ω, = Υ'(-φ+—^\ 2 cos φ nhcos φ L

[099] At this point, an angle between a tangent of the maximum slope 02 and the geometric axis X and an angle between a tangent of the minimum slope ωι and the geometric axis X are defined as ΔΘ (degrees), respectively (see Figure 12). Then, when h / L is replaced by the slope λ and approximation is used as h«L, a relationship between Δθ and the slope λ can be expressed by Formula (6) below. As illustrated in Figure 12, Δθ in Formula (6) below is an angle formed by the reference plane (plane a) and the tangent of the maximum slope 02, and is an angle formed by the reference plane (plane a) and the tangent of the minimum slope ω-ι. 180z x180 18Ott / icos0 , . Δ0 ~---tan(Á0) =---ω2=--:---= 18OAcos0 ---(6) π π π L

[0100] Next, from the angular characteristic of emissivity illustrated in Figure 4, brightness variations at the maximum light reception angle 0 + ΔΘ and at the minimum light reception angle 0 - ΔΘ are calculated. At this point, when a luminance ratio of a bright portion and a dark portion is calculated in order to cancel the influence of background brightness variation caused by a change in target temperature, and the relationship between a bright-dark change rate and slope λ can be arranged as the bright-dark change rate r, the slope λ can be calculated from the image. Note that the bright-dark change rate r can be obtained by automatically calculating the luminance of the bright and dark portions of the image using image processing.

[0101] Firstly, in the graph in Figure 4, a mathematical formula is fitted and modeled in a range where the light reception angle θ is 20 degrees or less. In this range, the radiance I (Θ) also increases with an increase in Petition 870250081021, dated 09 / 09 / 2025, page 37 / 81 29 / 38 light receiving angle Θ. The mathematical formula to be fitted can be any formula, but here, linear approximation is performed as in Formula (7) below. In Formula (7) below, I (Θ) is obtained by linearly approximating a region of 0 < θ < 20 in the angular characteristic of the emissivity. 1(θ) = αθ + ό ---(7)

[0102] Next, the bright-dark change rate r is defined. A ratio between the radiance at the maximum light-receiving angle θ + Δθ at which the radiance is maximized on the steel plate surface and the radiance at the minimum light-receiving angle θ - ΔΘ at which the radiance is minimized can be expressed by Formula (8) below. However, θ in the following Formula (8) is limited to a case where θ is greater than Δ0. / (0+Δ0) / (0 + 1802 cos 0)Λ) / (0-Δ0) ~ / (0 - 18O2cos0)

[0103] Finally, from Formulas (7) and (8) above, Formula (9) below can be obtained by expressing the slope λ with the formula for the bright-dark change rate r. αθ + br - 1 2 = --- ---(9) 180α cos 0 r + 1

[0104] In Formula (9) above, the light receiving angle θ and the angle φ with respect to the reference plane (plane a) are predetermined values ​​based on the installation position of camera 2 with respect to the steel plate S, and thus parameters a and b are determined from an angle dependence of the radiance. Therefore, the slope λ can be physically estimated by calculating luminance values ​​of the bright and dark portions by image processing and substituting the ratio in Formula (9) above as the bright-dark change rate r.

[0105] An example in which the slope λ is estimated based on the bright-dark pattern actually generated in the image of steel plate S will be described. Camera 2 was installed on steel plate S so that the receiving angle of Petition 870250081021, dated 09 / 09 / 2025, page 38 / 81 30 / 38 light θ becomes 12.5 degrees and the angle φ becomes 45 degrees with respect to the reference plane (plane α). Then, from the angular characteristic of the emissivity illustrated in Figure 13, a = 3.5 and b = 20 were calculated by fitting Formula (7) above. A straight line I (θ) in Figure 13 is a straight line obtained as a result of fitting. Figure 14 is a graph obtained by substituting the parameters a and b obtained in Formula (9) above. By inserting the bright-dark change rate r (black point on the right side of Figure 14) based on this graph, the slope λ (black point on the left side of Figure 14) can be calculated.

[0106] Next, an image obtained by camera 2 is illustrated in Figure 15. The bright-dark pattern is generated, and a luminance value for the bright portion was calculated as 140 and a luminance value for the dark portion was calculated as 115 by image processing. Furthermore, the bright-dark change rate r was calculated to be 1.22 by taking a ratio of these values. The slope λ was successfully estimated to be 0.014 by substituting the calculated rate into Formula (9) above. Also in the graph of Figure 14, a point on the vertical geometric axis (slope λ) corresponding to the bright-dark change rate r of 1.22 is 0.014, and thus the slope λ can also be calculated from the graph of Figure 14.

[0107] Strictly speaking, since the distance from camera 2 is very different between the upper edge portion and the lower edge portion, the actual magnification and transport direction are different between the upper and lower portions of the plate region in the image. The difference becomes more conspicuous as the positions of the steel plate S to be measured and camera 2 get closer. Therefore, more strictly, it is desirable to perform trapezoidal correction on the image using linear transformation or similar from a positional relationship of camera parameters calculated beforehand and the measurement target. Furthermore, taking a sufficient distance between the steel plate S to be measured and the camera Petition 870250081021, dated 09 / 09 / 2025, page 39 / 81 31 / 38 2, a difference in imaging conditions in the visual field can be reduced.

[0108] In step S4 described above, the image is rotated so that the bright-dark pattern can be easily extracted, but the luminance profile can be extracted obliquely with respect to the transport direction of the steel plate S without rotation.

[0109] In binarization at step S1, the brightness of the thermal radiation light varies depending on the temperature of the measurement target. Therefore, when binarization is performed with a fixed value, there is a problem where part of the background is detected as the plate region when the steel plate S is bright, and on the other hand, the background cannot be detected when the steel plate S is dark.

[0110] For the problem above, it is desirable to perform luminance correction. Examples of a luminance correction method include a method for multiplying the luminance of the entire image by a fixed value so that a representative value such as a maximum value, an average value, a median value, or a percentage of the luminance of the entire image becomes a target value. In addition, correction for a dark current can be performed prior to multiplication.

[0111] Furthermore, for example, as illustrated in Figure 16, it can be difficult to distinguish a structure from the sheet surface when a structure such as a roller or a transfer table is irradiated with thermal radiation light emitted from the steel sheet S, and the structure is imaged with brightness in an image. In this case, in the image, the sheet surface is bright because a sheet surface directly receives thermal radiation light, but the structure is often darker than the sheet surface because the structure receives reflection of thermal radiation light emitted from the sheet surface.

[0112] Therefore, a binarization limit that can separate the structure and the sheet metal by performing the luminance correction described above can be determined with a fixed value, but it is possible to detect more stably a portion of Petition 870250081021, dated 09 / 09 / 2025, page 40 / 81 32 / 38 valley by automatically detecting the valley portion from a histogram of the entire image. To detect the valley portion, for example, it is possible to stably calculate the minimum value by differential processing, search processing, or similar by applying a low-pass filter to a luminance histogram itself. An example of the luminance histogram is illustrated in (a) of Figure 17, and (a) is a graph in which luminance is represented on the horizontal geometric axis and the number of pixels N is represented on the vertical geometric axis. An example in which the low-pass filter is applied to a luminance direction of the luminance histogram is illustrated in (b) of Figure 17. Note that, in the histogram, the valley portion includes two portions, which are a boundary between the sheet surface and the structure and a boundary between the structure and the background. Therefore, it is preferable to select the larger of the two valley portions.A position separated by a dashed line in (b) of Figure 17 corresponds to the binarization limit.

[0113] Furthermore, in the present embodiment, an example of processing the image from camera 2 by the image processing unit 3 to calculate the steel sheet format index S was described. However, a camera image can simply be presented to an operator as is to provide feedback for lamination control.

[0114] Furthermore, the shape index of steel plate S can be estimated from the image using a machine learning-based discriminator. Specifically, first, training data is created by associating the image of steel plate S obtained with the shape index measured visually or by some other methods as a correct answer. Then, using the created training data and the machine learning method, a discriminator that takes an image as input to output a shape index is generated. Using this discriminator, the shape index of steel plate S to be measured is Petition 870250081021, dated 09 / 09 / 2025, page 41 / 81 33 / 38 calculated. In addition to the image, data fed into the machine learning and discriminator can be a characteristic amount calculated from the image. Furthermore, the machine learning method is not limited, and a convolutional neural network or similar can be used as long as real-time processing is not required. Example of an application for determining format acceptance.

[0115] An example in which the strip-like object shape measuring apparatus according to the embodiment is applied for shape acceptance determination will be described with reference to Figure 18. The acceptance determination system in Figure 18 includes camera 2, image processing unit 3, and an acceptance determination unit 6.

[0116] First, image processing unit 3 calculates the shape index (shape data) of the steel sheet S from an image captured by camera 2, and transmits the index to the acceptance determination unit 6. The acceptance determination unit 6 determines acceptance or rejection of the shape based on information from the steel sheet (e.g., sheet thickness, sheet width, steel type, and temperature) acquired from a host system and the shape index of the steel sheet S. An acceptance determination result is transmitted to the host system. Based on the acceptance determination result, the host system determines whether correction is necessary, either to cut off the defective shape portion, to perform rolling in the subsequent process, and the like.As described above, using the S-shaped steel sheet index for an action in the subsequent process makes it possible to eliminate problems and contribute to improving product quality. Example of an application for controlling lamination feedback.

[0117] An example in which the strip-like object shape measuring device according to the modality is applied for lamination feedback control Petition 870250081021, dated 09 / 09 / 2025, page 42 / 81 Section 34 / 38 will be described with reference to Figure 19. A lamination control system in Figure 19 includes camera 2, image processing unit 3, and a lamination control unit 7.

[0118] First, the image processing unit 3 calculates the steel sheet shape index S from an image captured by camera 2, and transmits the shape index to the rolling control unit 7. The rolling control unit 7 calculates a control parameter using the shape index based on steel sheet information (e.g., sheet thickness, sheet width, steel type, and temperature) acquired from the host system, other measurement data such as sheet passage position, and the steel sheet shape index S. The rolling control unit 7 transmits a control signal to the rolling roll 1 to perform feedback control such as leveling. By performing the above feedback control, the product shape can be stabilized, and a shape defect can be reduced. Example of an application for lamination control using machine learning.

[0119] An example in which the modality-like strip-shaped object measuring device is applied to lamination control using machine learning will be described with reference to Figure 20. A lamination control system in Figure 20 includes camera 2, image processing unit 3, lamination control unit 7, a data server 8, a machine learning unit 9, and a control parameter estimation unit 10.

[0120] First, image processing unit 3 calculates the format index of the steel plate S from an image captured by camera 2, and transmits the format index to data server 8. Data server 8 Petition 870250081021, dated 09 / 09 / 2025, page 43 / 81 35 / 38 accumulates the format index associated with steel sheet information (e.g., sheet thickness, sheet width, steel type, and temperature) acquired from the host system and a control parameter at the time of rolling.

[0121] Accumulated data is transmitted to machine learning unit 9. In machine learning unit 9, a model to estimate a control parameter for rolling while simultaneously eliminating a shape defect is built by machine learning. An estimation model is transmitted to control parameter estimation unit 10, and estimates the control parameter based on the steel sheet information acquired from the host system. The control parameter is transmitted to rolling control unit 7. Rolling control unit 7 performs pre-established control such as leveling by transmitting a control signal to rolling roll 1. By performing the above pre-established control, the product shape can be stabilized, and a shape defect can be reduced.

[0122] Note that, in the present application example, the control parameter in the preset control was described. However, machine learning can also be similarly applied to calculate a control parameter for the feedback control. Furthermore, by combining the preset control and the feedback control, the product's form quality can be further improved.

[0123] Furthermore, in addition to the above application examples, the device for measuring the shape of a strip-like object according to the embodiment may be supplied as a part of equipment for manufacturing a strip-like object.

[0124] Furthermore, the method for measuring the shape of a strip-like object, according to the embodiment, is also applicable to a method for controlling the shape of a strip-like object. In this case, the shape of the strip-like object is measured by the method for measuring the shape of a strip-like object described above, and the shape Petition 870250081021, dated 09 / 09 / 2025, page 44 / 81 36 / 38 of the strip-like object is controlled to have a desired shape based on the measurement result.

[0125] Furthermore, the method of measuring the shape of a strip-like object, according to the embodiment, is also applicable to a method of manufacturing a strip-like object. In this case, the shape of the strip-like object is measured by the method of measuring the shape of a strip-like object described above, and the strip-like object is manufactured based on the measurement result.

[0126] Furthermore, the method for measuring the shape of a strip-like object, according to the embodiment, is also applicable to a method for managing the quality of a strip-like object. In this case, the shape of the strip-like object is measured by the method for measuring the shape of a strip-like object described above, and the quality of the strip-like object is managed based on the measurement result.

[0127] The method for measuring the shape of a strip-like object, the method for controlling the shape of a strip-like object, the method for manufacturing a strip-like object, the method for managing the quality of a strip-like object, the apparatus for measuring the shape of a strip-like object, and the equipment for manufacturing a strip-like object according to the embodiment described above have the following effects.

[0128] Firstly, by passively imaging the thermal radiation light from the steel plate S to be measured, imaging can be performed from a good remote installation environment without installing a light source or sensor around the passing steel plate S, and the shape of the steel plate S can be captured clearly.

[0129] Furthermore, by selecting an imaging element suitable for the target object temperature as the imaging element of camera 2, thermal radiation light can be received efficiently. For example, for a target Petition 870250081021, dated 09 / 09 / 2025, page 45 / 81 37 / 38 of measurement at approximately 900°C, sensitivity can be increased by using the near-infrared component of the Si imaging element. In this way, even for the S steel plate transported at high speed, a clear image without blurring can be captured by shortening the exposure time.

[0130] Furthermore, by performing imaging in such a way that the angle φ formed by the orthographic projection of the optical geometric axis of camera 2 onto the reference plane (plane α) and the transport direction p of the steel plate S does not become 90 degrees, it is possible to clearly capture plate extension at each position in the width direction generated on the surface of the steel plate S. Here, when the light reception angle θ is 20 degrees or less with respect to the reference plane (plane α) of the surface of the steel plate S, a higher effect can be obtained.

[0131] In the present invention, the steel sheet S on the exit side of the finishing rolling stage in the hot rolling process has been described. However, it is obvious that the present invention is also applicable to other high-temperature rectangular steel materials such as steel plates and slabs. Furthermore, it is obvious that the present invention is applicable not only to strip-like objects in a steel manufacturing process but also to various strip-like objects of different materials, provided that thermal radiation light from the measurement target can be obtained. In addition, the present invention is preferably applied to measuring the extension shape in a portion other than the edge portion of the strip-like object, so as to gain a greater effect. In particular, the present invention is more preferably applied to measuring the extension shape of the central portion of the strip-like object, so as to gain an even greater effect.

[0132] The method of measuring the shape of a strip-like object, the method of controlling the shape of a strip-like object, the method of manufacturing a strip-like object, the method of quality management of a strip-like object Petition 870250081021, dated 09 / 09 / 2025, page 46 / 81 38 / 38 strip, the apparatus for measuring the shape of a strip-like object, and the equipment for manufacturing a strip-like object according to the present invention have been specifically described above with reference to embodiments and examples for carrying out the invention. However, the essence of the present invention is not limited to these descriptions, and is broadly interpreted based on the description of the claims. It is obvious that various changes and modifications based on the descriptions are also included in the essence of the present invention. List of Numerical References 1. LAMINATION ROLL 2. CAMERA Image Processing Unit 4. HEATER 5. BIAS MECHANISM 6. UNIT FOR DETERMINING ACCEPTANCE 7. Lamination Control Unit 8. DATA SERVER 9. Machine Learning Unit 10. Control Parameter Estimation Unit STEEL PLATE SA SAMPLE OF STEEL SHEET Petition 870250081021, dated 09 / 09 / 2025, page 47 / 81

Claims

1 / 3 CLAIMS 1. Method for measuring the shape of a strip-like object, for measuring the shape of a strip-like object, the method for measuring the shape of a strip-like object CHARACTERIZED in that it comprises: a capture step for capturing a thermal radiation light image of the strip-like object such that an angle φ formed by orthographic projection of an optical geometric axis of a camera onto a plane α and a transport direction p of the strip-like object does not become 90 degrees, the plane α being a reference plane of a surface of the strip-like object; and an image processing step of calculating an index of a surface shape of the strip-like object based on the radiance of the strip-like object in the image obtained.

2. Method for measuring the shape of a strip-like object, according to claim 1, CHARACTERIZED in that the image processing step includes extracting a linear component from the radiance of the strip-like object, the linear component being parallel to the transport direction of the strip-like object.

3. Method for measuring the shape of a strip-like object, according to claim 1 or 2, CHARACTERIZED in that, during the capture step, the plane α and the optical geometric axis of the camera form an angle θ that is less than or equal to 20 degrees.

4. Method of measuring strip-like object shape according to any one of claims 1 to 3, CHARACTERIZED in that the image processing step includes calculating, as an index, any one or more of a slope, a wave height, a wave pitch, an amount of elongation, and a percentage of elongation at each position in a width direction on the surface of the strip-like object from a bright-dark pattern of the strip-like object obtained.

5. A method for controlling the shape of a strip-like object, the method being characterized in that it comprises: measuring the shape of the strip-like object by the method of measuring the shape of a strip-like object, as defined in any one of claims 1 to 4; and controlling the shape of the strip-like object, based on a measurement result, to have a desired shape.

6. Method for manufacturing a strip-like object, the method CHARACTERIZED in that it comprises: measuring a strip-like object shape by the method of measuring strip-like object shapes, as defined in any one of claims 1 to 4; and manufacturing the strip-like object based on a measurement result.

7. A method for managing the quality of a strip-like object, the method being characterized in that it comprises: measuring the shape of the strip-like object by the method of measuring the shape of a strip-like object, as defined in any one of claims 1 to 4; and managing the quality of the strip-like object based on a measurement result.

8. Strip-like object shape measuring apparatus, which measures the shape of a strip-like object, the strip-like object shape measuring apparatus CHARACTERIZED in that it comprises: an image capture unit configured to capture an image of thermal radiation from the strip-like object such that an angle φ formed by orthographic projection of an optical geometric axis of a camera onto a plane α and a transport direction p of the strip-like object does not become 90 degrees, the plane α being a reference plane of a surface of the strip-like object; and an image processing unit configured to calculate an index of a surface shape of the strip-like object based on the radiance of the strip-like object in the image obtained.

9. Equipment for manufacturing strip-like objects CHARACTERIZED in that it comprises an apparatus for measuring the shape of strip-like objects as defined in claim 8. Petition 870250081021, dated 09 / 09 / 2025, pp. 50 / 81

Images