Classification device, classification method, program, and learned model

Abstract

To provide a classification device capable of performing classification related to fragility of a glass plate due to foreign matter such as brick-like foreign matter.SOLUTION: A classification part comprises a determination part which inputs image data obtained by imaging a glass plate manufactured in a plate shape to a machine-learned model, and acquires an output result from the learned model as a determination result related to whether the glass plate is easy to break with foreign matter having different heat shrinkage characteristics.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to a classification device, a classification method, a program, and a trained model. [Background technology]

[0002] Glass sheets are manufactured by a float process or the like. In the float process, for example, molten glass continuously supplied onto molten tin in a bath is caused to flow on the molten tin and formed into a strip shape.

[0003] Patent Document 1 describes an inspection device that detects defects in optical materials that result in irregular colors and shapes (see Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2022-56389 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the prior art, the influence of debris-like foreign matter on glass sheets has not been fully considered. That is, in glass manufacturing methods such as the float process, brick-like foreign matter may adhere to the surface of a glass sheet, causing cracks. In this case, since the thermal shrinkage characteristics (e.g., thermal shrinkage rate) of the glass sheet and the brick-like foreign matter are different, when the temperature of the glass sheet drops, the glass sheet may break due to thermal shrinkage. If a foreign object with a different thermal shrinkage rate adheres to a glass plate, when the temperature of the glass plate drops, a phenomenon may occur in which cracks start from the foreign object and spread in all directions.

[0006] The present disclosure has been made in consideration of these circumstances, and aims to provide a classification device, classification method, program, and trained model that can classify the susceptibility of glass plates to breakage caused by foreign matter such as brick-like foreign matter. [Means for solving the problem]

[0007] One aspect of the present disclosure is a classification device that includes a determination unit that inputs image data of a glass plate manufactured into a plate-shaped glass plate into a machine learning trained model and obtains an output result from the trained model as a determination result regarding whether the glass plate is susceptible to breakage due to foreign matter with different thermal shrinkage characteristics.

[0008] One aspect of the present disclosure is a classification method in which a classification device inputs image data of a glass plate manufactured into a plate-shaped form into a trained model of machine learning, and obtains an output result from the trained model as a determination result as to whether the glass plate is susceptible to breakage due to foreign matter with different thermal shrinkage properties.

[0009] One aspect of the present disclosure is a program for enabling a computer to input image data of a glass plate being manufactured into a plate-shaped object into a trained machine learning model, and obtain the output result from the trained model as a determination result as to whether the glass plate is susceptible to breakage due to foreign matter with different thermal shrinkage characteristics.

[0010] One aspect of the present disclosure is a machine learning trained model that inputs image data of a glass plate manufactured into a plate shape and outputs judgment result data regarding whether the glass plate is susceptible to breaking due to foreign matter with different thermal shrinkage characteristics. [Effects of the Invention]

[0011] The information processing device, classification device, classification method, program, and trained model according to the present disclosure can classify glass sheets according to their susceptibility to breakage caused by foreign matter such as brick-like foreign matter. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a diagram illustrating a schematic configuration example of a glass manufacturing apparatus according to an embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of the arrangement of an imaging unit according to the embodiment. [Figure 3] FIG. 1 is a diagram illustrating an example of the configuration of a classification device according to an embodiment. [Figure 4A] FIG. 10 is a diagram showing an example of an image including brick-like foreign matter according to the embodiment; [Figure 4B] FIG. 10 is a diagram showing another example of an image including brick-like foreign matter according to the embodiment. [Figure 5A] 10A, 10B, and 10C are diagrams showing an example of the change over time in the influence of a brick-like foreign substance on a glass plate according to an embodiment. [Figure 5B] 10A, 10B, and 10C are diagrams showing other examples of changes over time in the influence of brick-like foreign matter on a glass plate according to the embodiment. [Figure 6] FIG. 10 is a diagram showing an example of a change over time in the distribution of brick-like foreign matter in the width direction of a glass sheet according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0014] [Glass manufacturing equipment] FIG. 1 is a diagram showing a schematic configuration example of a glass manufacturing apparatus 1 according to an embodiment. For convenience of explanation, FIG. 1 shows an XYZ Cartesian coordinate system, which is a three-dimensional Cartesian coordinate system. The glass manufacturing apparatus 1 includes a melting furnace 11, a float bath 12, an annealing furnace 13, and an inspection and cutting section 14.

[0015] In the melting furnace 11, the melting process is carried out. In the melting process, glass raw materials prepared by mixing multiple types of raw materials are melted to obtain molten glass. The glass raw materials are charged into the melting furnace 11 and then melted by heating means such as radiant heat from a flame sprayed from a burner or electric melting to become molten glass.

[0016] In the float bath 12, the molding process is carried out. In the forming process, the molten glass obtained in the melting process is continuously supplied onto molten tin in a bath, and the molten glass is caused to flow and form on the molten tin to obtain a glass sheet (so-called glass ribbon), which is a sheet of glass. The glass sheet is cooled while flowing in a predetermined direction and is pulled up from the molten tin.

[0017] In the annealing furnace 13, a annealing process is carried out. In the annealing step, the glass sheet obtained in the forming step is annealed inside the annealing furnace 13. Inside the annealing furnace 13, the glass sheet is annealed while being transported horizontally on rolls from the entrance to the exit of the annealing furnace 13.

[0018] In the inspection and cutting section 14, the inspection process and the cutting process are carried out. In the inspection step, the glass sheet that has been slowly cooled in the slow cooling step is inspected. In the cutting process, the glass plate that has been annealed in the annealing process is cut into predetermined dimensions using a cutting machine. In the cutting process, both widthwise edge portions (so-called edge portions) of the glass plate are cut off. Both widthwise edge portions of the glass plate are cut off because they become thick due to the influence of surface tension, etc. After the cutting process, the glass sheets are shipped.

[0019] <Example of imaging unit placement> FIG. 2 is a diagram illustrating an example of the arrangement of the imaging unit according to the embodiment. For convenience of explanation, FIG. 2 shows XY coordinates in an XYZ orthogonal coordinate system similar to that shown in FIG. Fig. 2 schematically shows the state of the molten glass G when viewed from above (the positive side of the Z axis) of the float bath 12. Specifically, Fig. 2 shows a glass width region E that represents the width region of the molten glass G in the float bath 12. The glass width region E in the flow direction represents the change in glass width in the flow direction.

[0020] Tensile forces F1 to F4 directed outward in the width direction (to both outer sides) are applied to the molten glass G. As a result, the width of the glass width region E gradually increases from the upstream side to the downstream side of the flow.

[0021] FIG. 2 shows a schematic diagram of a defect inspection section 311 (the inspection section of the inspection and cutting section 14 in FIG. 1). In the example of FIG. 2, the inspection unit 311 has three imaging units C1 to C3 arranged in the width direction. The imaging unit C1 captures an image of the inspection region R1. The imaging unit C2 captures an image of the inspection region R2. The imaging unit C3 captures an image of the inspection region R3. Here, the inspection region R1, the inspection region R2, and the inspection region R3 are aligned in the width direction corresponding to the imaging units C1 to C3, respectively. 2, a defect H12 exists in inspection area R1, and a defect H11 exists in inspection area R3. Here, the defect H11 corresponds to the defect H1 and the defect H1a, and the defect H12 corresponds to the defect H2 and the defect H2a. In the example of FIG. 2, there is no defect in the inspection area R2. In this way, each of the imaging units C1 to C3 captures an image of a partial area of ​​the glass plate. If the area contains a defect, the image contains a portion where the defect is captured (defective portion).

[0022] Here, in this embodiment, three imaging units C1 to C3 are shown, but the number of imaging units may be arbitrary. 2, the inspection regions R1 to R3 arranged in the width direction are shown spaced apart from one another to make the inspection regions R1 to R3 easier to see, but they may be adjacent to one another, that is, they may be capable of inspecting the entire width direction of the molten glass G. Alternatively, as shown in FIG. 2, a plurality of inspection regions spaced apart from one another may be used. Furthermore, although each of the imaging units C1 to C3 is shown simply in FIG. 2, each of the imaging units C1 to C3 may be a combination of devices such as a plurality of cameras.

[0023] [Classification device] FIG. 3 is a diagram illustrating an example of the configuration of the classification device 111 according to the embodiment. The classification device 111 is, for example, configured by a computer. The classification device 111 includes an input unit 131 , an output unit 132 , a communication unit 133 , a storage unit 134 , and a control unit 135 . The output unit 132 includes a display unit 151 . The control unit 135 includes a learning control unit 191 and a determination unit 192 .

[0024] The input unit 131 has a function of inputting information in response to an operation by a user (for example, a person), and a function of inputting information from an external device. The output unit 132 has a function of outputting information to the outside. For example, the display unit 151 has a function of displaying information to be displayed on a screen. The communication unit 133 has a function of communicating with an external device.

[0025] The storage unit 134 stores various types of information. In this embodiment, the storage unit 134 stores a learning model 171, training data 172, and the like. The learning model 171 may be set in advance in the storage unit 134, or may be set externally at any timing. The teacher data 172 may be set in advance in the storage unit 134, or may be set externally at any timing.

[0026] Here, any model may be used as the learning model 171, for example, a neural network model may be used. The learning model 171 extracts, for example, feature quantities (image feature quantities) of input image data and outputs an estimation result (determination result) based on the feature quantities. In this embodiment, the feature quantities are information representing the breakability of the surface of the glass plate.

[0027] In this embodiment, the function of the feature extraction unit that extracts the features of image data and the function of the estimation unit that outputs an estimation result (judgment result) regarding the ease of breaking based on the extracted features are integrated and included in the learning model 171. As another example, the function of the feature extraction unit and the function of the estimation unit may be configured using separate learning models. In this case, image data is input to a learning model having the function of the feature extraction unit, and the output result (feature extraction result) from the learning model is input to a learning model having the function of the estimation unit, and the output from the learning model is used as the estimation result (determination result).

[0028] The control unit 135 performs various processes and controls. The learning control unit 191 uses the teacher data 172 to learn the learning model 171. This learning is machine learning. Note that, in the learning stage, for example, instructions based on the judgment results of the user may be used for learning.

[0029] The determination unit 192 inputs image data to be determined to the trained learning model 171, and outputs the output from the learning model 171 as a determination result via the output unit 132. For example, the judgment unit 192 may directly adopt the output (estimation result) from the trained learning model 171 as the judgment result by the judgment unit 192, or may make a further judgment (for example, trend judgment) based on the output (estimation result) from the trained learning model 171, and use the result of that judgment as the judgment result by the judgment unit 192.

[0030] Here, the image data to be determined may be input from the input unit 131, or may be stored in advance in the storage unit . For example, the judgment unit 192 may always perform judgment processing in real time by inputting image data acquired by the inspection unit 311 or image data processed from the image data in real time from the input unit 131. The determination process may be performed automatically in the classifier 111, for example.

[0031] <Breakability of glass sheets containing brick-like foreign matter> It is believed that the susceptibility of a glass plate to breakage changes depending on the state of cracks that occur around a brick-based foreign object (brick-based foreign object) that is caused by the brick-based foreign object adhering to the glass plate. For example, it may be considered that the presence of cracks makes the glass sheet more susceptible to breaking than would be the case without cracks. It may also be considered that, for example, the larger the brick-like foreign matter (the larger the area), the more likely the glass plate will break. It may also be considered that, for example, the greater the number of cracks, the more likely the glass sheet is to break. It may also be considered that, for example, the larger the crack (the larger the area), the more likely the glass plate is to break. It may also be assumed that, for example, the longer the crack (extending further from the central foreign object), the more likely the glass sheet is to break. It is believed that such cracking conditions are related to, for example, the temperature gradient in the glass plate during the manufacturing process.

[0032] <Example of an image of a glass plate containing brick-like foreign matter> 4A and 4B, examples of images of a glass plate containing brick-like inclusions are shown. FIG. 4A is a diagram showing an example of an image 511 including a brick-like foreign substance 521 according to the embodiment. FIG. 4B is a diagram showing another example of an image 512 including a brick-like foreign matter 531 according to the embodiment. The examples of FIGS. 4A and 4B are merely illustrative and are not necessarily limited to these examples.

[0033] Here, image 511 shown in FIG. 4A and image 512 shown in FIG. 4B are images that have been captured by a diffuse transmission camera and then have their gradation values ​​enhanced by a predetermined enhancement process.

[0034] In the example of Fig. 4A, the glass plate contains a brick-like foreign matter 521, but no cracks have occurred originating from the brick-like foreign matter 521. For this reason, the glass plate is considered to be in a state where it is difficult to break. On the other hand, in the example of Fig. 4B, the glass plate contains a brick-like foreign matter 531, and three cracks 531a, 531b, and 531c have occurred starting from the brick-like foreign matter 531. For this reason, the glass plate is considered to be in a state where it is easily broken.

[0035] This will be explained further. In this example, the material that is the crack initiation point is a material such as a brick, and has an irregular shape that is a polygon with more than a square or a circle. The size of the material is, for example, at most a few millimeters. In contrast, in this example, when a crack occurs in the glass centered on the substance, the crack extends linearly and has a sharp triangular shape at the tip, which is classified as such. Generally, as a crack extends it may meander, but if it does, the glass may break into large pieces before it reaches the inspection machine, causing it to slide off the rolls in the roll transport section, or the broken glass may flow on top of the unbroken glass, making it impossible to detect.

[0036] <Example of the change over time in the impact of brick-like foreign matter on a glass plate> 5A(A), 5A(B), and 5A(C) are diagrams showing examples of changes over time in the influence of brick-like foreign matter on a glass plate according to the embodiment. The examples in FIGS. 5A(A), 5A(B), and 5A(C) are outlines for the purpose of explanation, and specific values ​​on the horizontal and vertical axes are omitted.

[0037] In each of Figures 5A(A), 5A(B), and 5A(C), the horizontal axis represents the passage of time (in this example, dates). The horizontal axis is common to Figures 5A(A), 5A(B), and 5A(C). In FIG. 5A(A), the vertical axis represents the number of brick-like foreign matter in which cracks have not occurred (in this example, the number per day). In FIG. 5A(B), the vertical axis represents the number of brick-like foreign objects in which cracks have occurred (in this example, the number per day).

[0038] In FIG. 5A(C), the vertical axis represents the crack ratio [%]. In this example, the cracked ratio [%] is the ratio [%] of brick-like foreign matter in which cracks have occurred to the total number of brick-like foreign matter. The total number of brick-like foreign matter is the sum of the number of FIG. 5A(A) and the number of FIG. 5A(B).

[0039] The characteristic 1011 shown in FIG. 5A(A) is the characteristic of the time change resulting from counting the number of crack-free brick-like foreign objects every day in a specified region of a glass plate in a specified glass manufacturing process. A characteristic 1021 shown in FIG. 5A(B) is a characteristic of time change resulting from counting the number of cracked brick-like foreign matter per day in the relevant region in the glass manufacturing process. A characteristic 1031 shown in FIG. 5A(C) is a characteristic of the change over time in the crack ratio [%] per day for the region in question in the glass manufacturing process in question. It should be noted that the examples of FIG. 5A(A), FIG. 5A(B), and FIG. 5A(C) are merely illustrative examples for the purpose of explanation, and the present invention is not necessarily limited to these examples.

[0040] 5B(A), 5B(B), and 5B(C) are diagrams showing other examples of changes over time in the influence of brick-like foreign matter on the glass plate according to the embodiment. Note that the examples in FIGS. 5B(A), 5B(B), and 5B(C) are outlines for the purpose of explanation, and specific numerical values ​​on the horizontal and vertical axes are omitted.

[0041] In each of Figures 5B(A), 5B(B), and 5B(C), the horizontal axis represents the passage of time (in this example, time per hour). The horizontal axis is common to Figures 5B(A), 5B(B), and 5B(C). In FIG. 5B(A), the vertical axis represents the number of brick-like foreign matter in which cracks have not occurred (number / hour in this example). In FIG. 5B(B), the vertical axis represents the number of brick-like foreign objects in which cracks have occurred (number / hour in this example).

[0042] In FIG. 5A(C), the vertical axis represents the crack ratio [%]. In this example, the cracked ratio [%] is the ratio [%] of brick-like foreign matter in which cracks have occurred to the total number of brick-like foreign matter. The total number of brick-like foreign matter is the sum of the number of FIG. 5B(A) and the number of FIG. 5B(B).

[0043] The characteristic 1111 shown in Figure 5B(A) is the characteristic of the time change resulting from counting the number of crack-free brick-like foreign objects every hour in a specified region of a glass plate during a specified glass manufacturing process. A characteristic 1121 shown in FIG. 5B(B) is a characteristic of time change resulting from counting the number of cracked brick-like foreign matter every hour in the relevant region in the glass manufacturing process. A characteristic 1131 shown in FIG. 5B(C) is a characteristic of the time change in the crack ratio [%] per hour for the region in question in the glass manufacturing process in question. Note that the examples of FIG. 5B(A), FIG. 5B(B), and FIG. 5B(C) are merely illustrative examples for the purpose of explanation, and the present invention is not necessarily limited to these examples.

[0044] Here, in the series of examples of Figures 5A(A), 5A(B), and 5A(C), and the series of examples of Figures 5B(A), 5B(B), and 5B(C), trends are visualized for the status of brick-like foreign matter and its cracks. For example, it may be considered that the glass plate is more likely to break in places where the crack ratio [%] is high, and less likely to break in other places.

[0045] For example, in the conventional technology, the results of Figure 5A(A) and Figure 5(B) cannot be separated, or the results of Figure 5B(A) and Figure 5(B) cannot be separated, but in this embodiment, these can be separated. In this embodiment, the crack presence ratio [%] can be obtained as shown in Figure 5A(C) or Figure 5B(C).

[0046] The aggregation time intervals on the horizontal axes in Figures 5A(A), 5A(B), and 5A(C), and the aggregation time intervals on the horizontal axes in Figures 5B(A), 5B(B), and 5B(C), are not particularly limited, and other time intervals may be used.

[0047] <An example of time-dependent changes in distribution of brick-like foreign particles across the width of a glass plate> FIG. 6 is a diagram showing an example of the change over time in the distribution of brick-like foreign matter in the width direction of the glass plate according to the embodiment. The horizontal axis shown in FIG. 6 represents the passage of time (date in this example), and the vertical axis represents the width direction of the glass plate.

[0048] Here, the width direction of the glass sheet is, for example, a direction parallel to the X axis in the example of FIG. 2, and is a direction perpendicular to the flow direction. As in the example of Figure 2, multiple imaging devices (in the example of Figure 2, imaging devices of imaging units C1 to C3) are installed in the width direction, and the example of Figure 6 is information corresponding to the captured image at the inspection point of one imaging device (or it may be the inspection point of two or more imaging devices).

[0049] In the example of Figure 6, brick-based foreign matter in which cracks have not occurred (crack-free brick-based foreign matter) is indicated by a predetermined mark K1, and brick-based foreign matter in which cracks have occurred (cracked brick-based foreign matter) is indicated by another predetermined mark K2. For example, it may be considered that glass sheets are more likely to break in places where there are many cracked brick-like foreign objects. Furthermore, for example, it may be considered that even if there are many crack-free brick-like foreign matters, the glass plate is less likely to break in a location where there are few cracked brick-like foreign matters. It is considered that the glass plate is less likely to break in areas where neither cracked brick-like foreign matter nor cracked brick-like foreign matter is present. The example in FIG. 6 is an example for the purpose of explanation, and the present invention is not necessarily limited to this example.

[0050] Here, in the example of Figure 6, the value (time) on the horizontal axis is a predetermined constant value, and the information along the vertical axis is a collection of information along the width direction at the same timing (same period), and based on this, it is possible to determine whether or not the glass plate is prone to breaking for each timing (period). In addition, in the example of Figure 6, the value on the vertical axis (position in the width direction) is a predetermined constant value, and the information along the horizontal axis is a collection of time-series information at the same position (same location) in the width direction, and based on this, it is possible to determine whether or not the glass plate is prone to breaking for each position (location).

[0051] <First example of learning and judgment> As input data (teaching data 172, data to be determined) of the learning model 171, image data of a glass plate containing brick-like foreign matter is used. Then, the output data of the learning model 171 is used as data on the determination result regarding the breakability of the glass plate.

[0052] Here, the judgment result regarding the fragility of the glass plate may be, for example, one or more of binary information indicating whether the glass plate is fragile or not, or three or more values ​​(discrete or continuous value information) indicating the degree of fragility of the glass plate. As information indicating the degree of fragility of a glass plate, for example, a plurality of classes that distinguish the degree of fragility of a glass plate into a plurality of stages may be used.

[0053] As a specific example, the determination unit 192 may determine an index of the fragility of a glass plate from an image captured (scanned) of information related to rubble-like foreign matter using a trained model (trained learning model 171). The determination unit 192 may, for example, determine an index of the glass plate's fragility for a large number of images that change over time in the same inspection areas R1 to R3, and perform trend aggregation of the determination results, thereby more accurately identifying (determining) the index.

[0054] Here, in this example, in machine learning, a judgment result such as, for example, that the presence of cracked brick-like foreign matter makes the glass plate more likely to break, and that the presence of cracked brick-like foreign matter does not make the glass plate more likely to break (i.e., less likely to break) may be used as training data 172. Alternatively, for example, it may be possible to actually inspect whether the glass plate is broken or not, and use the inspection result (whether the glass plate is broken or not) as the training data 172.

[0055] In the trained model (trained learning model 171), for example, one or more of the following may be used as features: the size of the rubble-based foreign object (e.g., area), the number of cracks caused by the rubble-based foreign object (e.g., the number of cracks extending from one rubble-based foreign object), the length of extension of the cracks caused by the rubble-based foreign object, or the size of the cracks caused by the rubble-based foreign object (e.g., the area of ​​the cracks).

[0056] Furthermore, the determining unit 192 may determine, for example, one or more of a location on the glass plate that is prone to breaking (for example, a position in the width direction) and a timing (period) when the glass plate is prone to breaking. In this case, for example, the judgment unit 192 may make only such a judgment (here, for convenience of explanation, referred to as judgment B), or may make a predetermined judgment in a previous stage (here, for convenience of explanation, referred to as judgment A) and then make judgment B based on the result of judgment A. Here, the judgment A may be, for example, a judgment of binary information representing whether or not the glass plate is fragile, or a judgment of ternary or more value information representing the degree of fragility of the glass plate.

[0057] <Second example of learning and judgment> As input data (teacher data 172, data to be determined) to the learning model 171, data of an image of a glass plate containing rubble-like foreign matter and auxiliary data (auxiliary data) may be used. Then, the output data of the learning model 171 is used as data on the determination result regarding the breakability of the glass plate. Here, data representing various characteristics may be used as the auxiliary data.

[0058] <Example of evaluation results for glass plate breakability> In this embodiment, the judgment result regarding the breakability of the glass plate is an index of the breakability of the glass plate, and may be information on a predetermined value itself, such as a binary value indicating whether the glass plate is breakable or not, or the degree of breakability of the glass plate, or may be information used to derive such a value. Furthermore, such an index may be, for example, a ratio (crack ratio [%]) as shown in Figure 5A(C) or Figure 5B(C), or information on the distribution as shown in Figure 6 (for example, widthwise information at a certain timing, or time-series information at a certain position in the widthwise direction).

[0059] Here, as the information regarding the fragility of the glass plate, for example, information regarding each individual brick-like foreign matter may be used, or information regarding a group of a plurality of brick-like foreign matters may be used. For example, if a unit area to be judged (judgment unit area) includes a plurality of brick-like foreign objects, a judgment result regarding the ease of cracking for a collection of these plurality of brick-like foreign objects may be acquired. As a specific example, image data obtained by capturing an image of the unit area to be judged (judgment unit area) may be used as image data to be used during machine learning learning and judgment (image data to be input to the learning model 171).

[0060] <Example of imaging device> In this embodiment, a preferred example is a diffuse transmission camera used as the imaging device (the imaging device constituting the imaging units C1 to C3 shown in FIG. 3). However, the imaging device is not limited to this, and other devices such as a reflective optical system camera may also be used. Here, images captured by a diffuse transmission camera are preferable because, for example, shadows are not captured, which increases the accuracy with which rubble-like foreign objects and cracks can be recognized from the captured images. Furthermore, for example, a process may be performed to enhance the gradation values ​​in image data captured by an imaging device. As an example, this process may be a process of converting gradation values ​​so that the original gradation values ​​and the changed gradation values ​​match at a predetermined boundary value (predetermined gradation value), and the changed gradation values ​​are smaller than the original gradation values ​​at gradation values ​​smaller than the boundary value, and the changed gradation values ​​are larger than the original gradation values ​​at gradation values ​​larger than the boundary value.

[0061] The positions at which the imaging devices are installed (the positions of the imaging units C1 to C3 shown in FIG. 3) are not necessarily limited to the example in FIG. 3, and they may be installed at other positions (other locations) in the manufacturing process of the glass plate. However, in this embodiment, since the presence or absence of cracks caused by debris-based foreign matter is to be recognized, the position at which the imaging device is installed is preferably a position at which the temperature of the glass plate has dropped to a certain extent, for example, a position downstream of the annealing process.

[0062] As described above, the classification device 111 according to this embodiment can use machine learning to classify glass sheets according to their susceptibility to breakage caused by foreign matter such as brick-like foreign matter. As a result, the classification device 111 according to this embodiment can improve the manufacturing efficiency of glass sheets by, for example, considering measures to prevent glass sheets from breaking based on the classification results. Here, during machine learning, for example, brick-like foreign matter and cracks resulting from the brick-like foreign matter in the glass plate may be counted by screening. In addition, when making a judgment using a trained model, it is possible to investigate relationships such as trend data on the presence or absence of cracks, temperature gradients in the glass plate, and breakage of the glass plate, and to consider countermeasures to prevent breakage of the glass plate.

[0063] It should be noted that the classification device 111 may be, for example, a non-online device or an online device (for example, a cloud server or an on-premise server).

[0064] Furthermore, in this embodiment, the case where the foreign matter is a brick-type foreign matter has been described, but the present invention may also be applied to other foreign matter that has different thermal shrinkage characteristics (e.g., thermal shrinkage rate) compared to a glass plate. Furthermore, in the present embodiment, the float method is used as a method for manufacturing a flat glass sheet. However, as another example, a fusion method may also be used.

[0065] As one configuration example, in the classification device 111, the judgment unit 192 inputs image data of a glass plate manufactured into a plate-shaped form into a machine learning trained model, and obtains the output result from the trained model as a judgment result regarding whether the glass plate is susceptible to breaking due to foreign matter with different thermal shrinkage properties. Therefore, the classifying device 111 can classify glass sheets according to their breakability caused by foreign matter such as brick-like foreign matter.

[0066] As one configuration example, in the classification device 111, the trained model uses, as a feature amount, one or more of the size of a foreign object or the extent of extension of a crack caused by the foreign object. Therefore, the classification device 111 can improve the accuracy of classification.

[0067] As an example of configuration, the classification device 111 includes an imaging device (imaging units C1 to C3 in the example of FIG. 2) that captures images of the image data. The imaging device is a diffuse transmission camera. Therefore, the classification device 111 can easily recognize foreign matter and cracks in the image data, and the accuracy of classification can be further improved.

[0068] As one configuration example, in the classification device 111, the determination unit 192 determines one or more of a location on the glass plate that is likely to break or a time when the glass plate is likely to break. Therefore, the classification device 111 can provide useful information regarding the breakability of the glass plate.

[0069] As one configuration example, in the classification device 111, the image data is data of an image of a part of a glass plate, and includes a part in which a foreign object is captured. Therefore, the classification device 111 can classify the glass plate in terms of the fragility of the glass plate for each portion where the foreign matter is present on the glass plate.

[0070] As an example configuration, in the classification device 111, the learning control unit 191 generates a trained model using training data. Therefore, the classification device 111 can generate a trained model and use the trained model to classify glass plates according to their breakability.

[0071] In this embodiment, the classification device 111 is shown to have both the learning control unit 191 and the judgment unit 192. However, as another example, the learning control unit 191 may not be provided in the classification device 111 but in another device, and the classification device 111 may use a trained model obtained by the other device.

[0072] A program for implementing the functions of any of the components of any of the above-described devices may be recorded on a computer-readable recording medium and loaded into a computer system for execution. The term "computer system" as used herein includes hardware such as an operating system or peripheral devices. The term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and compact discs (CDs) or read-only memories (ROMs), as well as storage devices such as hard disks built into computer systems. The term "computer-readable recording medium" also includes devices that retain a program for a certain period of time, such as volatile memory within a computer system that acts as a server or client when a program is transmitted over a network such as the Internet or a communication line such as a telephone line. Such volatile memory may be, for example, random access memory (RAM). The recording medium may also be, for example, a non-transitory recording medium.

[0073] The above program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by transmission waves in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has the function of transmitting information, such as a network such as the Internet or a communication line such as a telephone line. The above program may also be one that realizes part of the above-mentioned functions. Furthermore, the above program may be a so-called differential file that can realize the above-mentioned functions in combination with a program already recorded in a computer system. A differential file may also be called a differential program.

[0074] Furthermore, the functions of any of the components in any of the above-described devices may be implemented by a processor. For example, each process in the embodiments may be implemented by a processor operating based on information such as a program and a computer-readable recording medium storing information such as the program. Here, the functions of each unit of the processor may be implemented by, for example, individual hardware, or may be implemented by integrated hardware. For example, the processor may include hardware, and the hardware may include at least one of a circuit for processing digital signals and a circuit for processing analog signals. For example, the processor may be configured using one or more circuit devices mounted on a circuit board, or one or both of one or more circuit elements. An integrated circuit (IC) or the like may be used as the circuit device, and a resistor or a capacitor may be used as the circuit element.

[0075] Here, the processor may be, for example, a CPU. However, the processor is not limited to a CPU, and various types of processors such as a GPU (Graphics Processing Unit) or a DSP (Digital Signal Processor) may be used. The processor may also be, for example, a hardware circuit such as an ASIC (Application Specific Integrated Circuit). The processor may also be, for example, composed of multiple CPUs, or may be, for example, composed of a hardware circuit such as a multiple ASIC. The processor may also be, for example, composed of a combination of multiple CPUs and a hardware circuit such as a multiple ASIC. The processor may also include, for example, one or more of an amplifier circuit or a filter circuit that processes analog signals.

[0076] The embodiments of this disclosure have been described in detail above with reference to the drawings, but the specific configuration is not limited to this embodiment, and includes designs within the scope that do not deviate from the gist of this disclosure.

[0077] [Note] (Configuration example 1) to (Configuration example 9) are shown.

[0078] (Configuration example 1) A determination unit inputs image data of a glass plate to be manufactured into a machine learning trained model, and obtains an output result from the trained model as a determination result as to whether the glass plate is susceptible to breakage due to foreign matter with different thermal shrinkage characteristics. Classification device.

[0079] (Configuration example 2) In the trained model, one or more of the following features are used: the size of the foreign matter; the length of extension of a crack caused by the foreign matter; the number of cracks; and the size of the crack area. The classification device according to (Configuration Example 1).

[0080] (Configuration example 3) an imaging device that captures an image of the image data; The imaging device is a diffuse transmission camera. The classification device according to (Configuration Example 1) or (Configuration Example 2).

[0081] (Configuration example 4) The determination unit determines one or more of a location on the glass plate that is prone to breakage or a time when the glass plate is prone to breakage. The classification device according to any one of (Configuration Example 1) to (Configuration Example 3).

[0082] (Configuration Example 5) The image data is data of an image of a part of the glass plate, and includes a part in which the foreign matter is captured. The classification device according to any one of (Configuration Example 1) to (Configuration Example 4).

[0083] (Configuration Example 6) A learning control unit that generates the trained model using training data. The classification device according to any one of (Configuration Example 1) to (Configuration Example 5).

[0084] It is also possible to provide a method of processing (classification method) performed by the classification device. (Configuration Example 7) A classification device inputs image data of a glass plate manufactured into a plate-shaped object into a trained model of machine learning, and obtains an output result from the trained model as a determination result regarding whether the glass plate is susceptible to breakage due to foreign matter having different thermal shrinkage characteristics. Classification method.

[0085] It is also possible to provide a program that is executed by a computer (computer program). (Configuration Example 8) On the computer, A function to input image data of glass sheets being manufactured into a machine learning model. A function of acquiring an output result from the trained model as a determination result regarding whether the glass plate is susceptible to breakage due to a foreign object with different thermal shrinkage characteristics; A program to achieve this.

[0086] It is also possible to provide pre-trained machine learning models. (Configuration Example 9) Inputting image data of a glass plate to be manufactured into a plate shape, and outputting determination result data regarding whether the glass plate is susceptible to breakage due to foreign matter having different thermal shrinkage characteristics. A trained machine learning model. [Explanation of symbols]

[0087] 1...glass manufacturing apparatus, 11...melting furnace, 12...float bath, 13...annealing furnace, 14...inspection and cutting unit, 111...classification device, 131...input unit, 132...output unit, 133...communication unit, 134...storage unit, 135...control unit, 151...display unit, 171...learning model, 172...teaching data, 191...learning control unit, 192...judgment unit, 311...inspection unit, 511-512...image, 521, 531...brick-based foreign matter, 531a to 531c...cracks, 1011, 1021, 1031, 1111, 1121, 1131...characteristics, C1 to C3...imaging area, E...glass width area, F1 to F4...tensile force, K1...mark (brick-based foreign matter without cracks), K2...mark (brick-based foreign matter with cracks), H1, H1a, H2, H2a, H11, H12...defective area, R1 to R3...inspection area

Claims

1. A determination unit inputs image data of a glass plate to be manufactured into a machine learning trained model, and obtains an output result from the trained model as a determination result as to whether the glass plate is susceptible to breakage due to foreign matter with different thermal shrinkage characteristics. Classification device.

2. In the trained model, one or more of the size of the foreign matter, the length of extension of a crack caused by the foreign matter, the number of cracks, and the size of the area of ​​the crack are used as feature quantities. The classification device of claim 1 .

3. an imaging device that captures an image of the image data; The imaging device is a diffuse transmission camera. The classification device according to claim 1 or claim 2.

4. The determination unit determines one or more of a location on the glass plate that is prone to breakage or a time when the glass plate is prone to breakage. The classification device according to claim 1 or claim 2.

5. The image data is data of an image of a part of the glass plate, and includes a part in which the foreign matter is captured. The classification device according to claim 1 or claim 2.

6. A learning control unit that generates the trained model using training data. The classification device according to claim 1 or claim 2.

7. A classification device inputs image data of a glass plate manufactured into a plate-shaped object into a trained model of machine learning, and obtains an output result from the trained model as a determination result regarding whether the glass plate is susceptible to breakage due to foreign matter having different thermal shrinkage characteristics. Classification method.

8. On the computer, A function to input image data of glass sheets being manufactured into a machine learning model. A function of acquiring an output result from the trained model as a determination result regarding whether the glass plate is susceptible to breakage due to a foreign object with different thermal shrinkage characteristics; A program to achieve this.

9. Inputting image data of a glass plate to be manufactured into a plate shape, and outputting determination result data regarding whether the glass plate is susceptible to breakage due to foreign matter having different thermal shrinkage characteristics. A trained machine learning model.

Images