JP2014037338A - Image processing method, image processing device, method of controlling electric melting tank, and method of manufacturing glass article - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing method capable of excellently monitoring the inside of an electric melting tank in which a cold top is formed by periodically supplying a glass material to the entire substrate surface of fused glass.SOLUTION: A state of a cold top at a position in a real space corresponding to a pixel at a common position is evaluated on the basis of variation in gradation value among pixels at the common position in a plurality of images obtained by photographing a range of a substrate surface in one cycle or a plurality of cycles of an operation to supply a glass material onto the substrate surface of fused glass. The process is performed for each of positions of pixels corresponding to the range of the substrate surface of the fused glass.

Description

  The present invention relates to an image processing method, an image processing apparatus, a method for controlling an electric melting tank, and a method for manufacturing a glass article, and in particular, an image processing method, an image processing apparatus, and an image for an image in an electric melting tank. The present invention relates to a method for controlling an electric dissolution tank using a treatment method and a method for producing a glass article.

  In the production of the glass article, there are an aspect in which the glass raw material is melted, for example, an aspect in which the glass raw material in the melting tank is heated with a burner, an aspect in which the glass raw material is melted by heating by energization, and the like. Hereinafter, a dissolution tank that is heated by energization is referred to as an electric dissolution tank. Examples of the electric dissolution tank are described in Patent Documents 1 and 2.

  As described in Patent Document 1, in the electric melting tank, utilizing the fact that the glass becomes a conductor at a high temperature, by directly energizing the molten glass from the electrode in the electric melting tank, by the generated Joule heat, Dissolve the glass raw material.

  FIG. 22 is a schematic cross-sectional view of the electrolysis tank as viewed from the side. An electrode 83 is provided in the electric dissolution tank 81. In addition, a raw material supply unit 82 is inserted into the electric melting tank 81 from the opening portion 88, and the raw material supply unit 82 supplies a glass raw material onto the base of the molten glass 85 from the tip of the raw material supply unit 82 itself. And it supplies with electricity directly to the molten glass 85 with the electrode 83, and the glass raw material deposited on the raw surface of the molten glass 85 is melt | dissolved by the generated Joule heat. Molten glass 85 generated by melting the glass raw material is supplied to the next step from the discharge port 87 (see FIG. 22).

  FIG. 23 is a schematic cross-sectional view of the electric dissolution tank as viewed from the insertion surface side of the raw material supply unit 82. For example, as shown in FIG. 23, the electrode 83 is disposed on the side surface side of the electrolysis tank 81. However, the arrangement of the electrode 83 shown in FIG. 23 is merely an example, and the electrode 83 may be arranged at another location. For example, the electrode 83 may be disposed on the bottom surface of the electrolysis tank 81. Alternatively, the electrode 83 may be disposed in the vicinity of the side surface of the electrolysis tank 81 so that the electrode 83 enters the molten glass 85 from above. 22 and 23 show a case where four electrodes 83 are arranged in the electric dissolution tank 81, but the number of electrodes 83 is not limited to four.

  Moreover, as described in Patent Document 1, in an electric melting tank, in order to effectively use heat energy generated by energization, a base of molten glass is covered with a layer of glass raw material. This layer of glass material is called a cold top. The cold top is sometimes referred to as a cold crown or blanket. By supplying the glass raw material on the ground surface of the molten glass 85 and providing the cold top 86 (see FIGS. 22 and 23), the heat radiation amount is reduced and the glass raw material (that is, the cold top 86) is efficiently melted. Can be made.

  As described above, it is preferable to maintain the state in which the cold top 86 is provided and melt the cold top 86 to obtain molten glass. Therefore, the raw material supply part 82 (refer FIG. 22) maintains the state in which the cold top 86 is formed by supplying a glass raw material periodically. At this time, in order to form the cold top 86 on the entire ground surface of the molten glass 85, the raw material supply unit 82 moves back and forth, right and left in the electric melting tank 81, or obliquely as necessary, and the glass raw material is disposed on the entire ground surface. Supply. The raw material supply unit 82 is configured, for example, to hold a glass raw material in an internal tank and discharge the raw material in the tank while moving. Alternatively, the raw material supply unit 82 may include a conveyor, move the glass raw material to the tip of the raw material supply unit 82 with the conveyor, and supply the glass raw material from the front end to the ground surface.

  FIG. 24 is an explanatory diagram illustrating an example of a movement path of the raw material supply unit 82 in the electric dissolution tank 81. FIG. 24 shows a path through which the tip of the raw material supply unit 82 passes.

  24A shows a movement path (outward path) of the raw material supply section 82 from a position 91 on the near side (opening portion 88 side shown in FIG. 1) to a position 92 on the back side (discharge port 87 side shown in FIG. 1). An example of As shown in FIG. 24 (a), for example, the raw material supply unit 82 moves to the left from the position 91, and when it reaches the vicinity of the left side surface, it moves to the back side by a predetermined distance and then moves to the right. When it reaches the vicinity of the right side, it moves to the far side and then moves to the left. The material supply unit 82 repeats this operation and moves to the position 92. It should be noted that the distance from the last reaching the side surface to the movement to the position 92 is, for example, 1/2 of the predetermined distance described above.

  FIG. 24B shows an example of a movement path (return path) of the raw material supply unit 82 from the position 92 on the back side to the position 91 on the near side. As shown in FIG. 24 (b), for example, when the raw material supply unit 82 moves to the right from the position 92 and reaches the vicinity of the right side surface, the raw material supply unit 82 moves to the near side by a predetermined distance and then moves to the left. When it reaches the vicinity of the left side surface, it moves to the near side by a predetermined distance and then moves to the right. The material supply unit 82 repeats this operation and moves to the position 91. It should be noted that the distance from the end of the side surface to the movement to the position 91 is, for example, ½ of the predetermined distance described above.

  As illustrated in FIG. 24, if the raw material supply unit 82 moves, the glass raw material supply position when moving in the left-right direction on the forward path and the return path is shifted, so that it is easy to form a cold top on the entire ground surface. In addition, the movement path | route of the raw material supply part 82 is not limited to the path | route illustrated in FIG. For example, the outbound path and the inbound path may overlap.

  The raw material supply unit 82 moves at a constant speed along the forward path and the return path, and continues to supply the glass raw material during that time. Moreover, when the raw material supply part 82 returns to the position 91, it will stop supply of a glass raw material, and will wait for a fixed time. Thereafter, the raw material supply unit 82 again moves at a constant speed along the path illustrated in FIG. 24 to supply the glass raw material. The time until the raw material supply unit 82 starts from the position 91 and returns to the position 91 is, for example, 30 to 40 minutes, and the standby time is, for example, 3 minutes. However, the time shown here is an example, and is not limited to these times.

  Thus, the raw material supply part 82 repeats the operation | movement of supplying a raw material to the whole ground surface, and standing by with a fixed period.

  Therefore, the cold top 86 is melted by the heat generated by energizing the molten glass 85, but the cold top 86 is maintained by periodically supplying the glass raw material.

  Note that a phenomenon may occur in which a part of the cold top 86 is broken through from below. This phenomenon is considered to be a phenomenon in which gas accumulates at the boundary between the cold top 86 and the molten glass 85 and the gas breaks through the cold top 86 and blows out. Hereinafter, this phenomenon is referred to as “blowout phenomenon”.

  Patent Document 2 describes a technique for measuring the thickness of the cold top by measuring the surface height of the cold top, the height of the boundary between the cold top and the molten glass, and the like.

In Patent Document 3, an image of a glass melting surface in a glass melting furnace is taken, and a time key is given to glass melting surface image information obtained by the photographing, glass melting furnace operation information corresponding to the photographing time, and the like. Are stored as time series data, a plurality of glass melt surface image information is extracted from the time series data, and a glass melt surface image at each photographing time is compared. In the method described in Patent Document 3, images of the glass melt surface are taken at intervals within a range of 10 seconds to 36 × 10 ² seconds. Moreover, in the method described in Patent Document 3, when comparing the glass melting surface images, the interface position between the glass raw material and the molten glass in the glass melting furnace and the interface position between the foam layer surface and the molten glass surface are compared. . Further, Patent Document 3 describes that an image taken as a perspective image inside the furnace is converted into a planar image by affine transformation. In addition, the glass melting furnace described in patent document 3 is not an electric melting tank, but a melting tank in a mode in which heating is performed by a burner.

  Patent Document 4 describes a plant monitoring device. The plant monitoring apparatus described in Patent Document 4 measures the temperature distribution of a monitoring target plant in the form of image data using an infrared camera, and inputs an image of the monitoring target plant into the monitoring apparatus main body at certain time intervals. And the monitoring apparatus calculates the average temperature from the average brightness | luminance of the pixel contained in each area | region about several area | regions on the image important for plant monitoring.

JP-A-5-163024 U.S. Pat. No. 4,718,931 JP 2009-161396 A JP-A-6-259678

  The state of the cold top in the electric melting tank means the state of melting the glass raw material, the thickness of the cold top, and other states. With regard to the state of the cold top, for example, “There is a region where the glass raw material is melted immediately and the cold top cannot be maintained.”, “There is a region where a blowing phenomenon occurs.”, “Blowing phenomenon may occur. It is preferable that various evaluations such as “There is a certain area” can be performed.

  In the techniques described in Patent Documents 3 and 4, the inside of a glass melting furnace or a plant is photographed, and the state in the glass melting furnace, the temperature inside the plant, and the like are evaluated based on the image. However, the techniques described in Patent Documents 3 and 4 are not intended to monitor the electric dissolution tank. That is, the situation where the glass material is periodically supplied to the entire ground surface of the molten glass 85 is not photographed.

In the technique described in Patent Document 3, images are taken at intervals within a range of 10 seconds to 36 × 10 ² seconds, and a plurality of images are compared. Even if such a photographing mode is applied to monitoring in the electric melting tank, the comparison result of the image varies depending on which timing image is used in the periodic glass raw material supply. The top status cannot be monitored well.

  In addition, as in the technique described in Patent Document 4, even if an image is taken with an infrared camera and the image is input to the monitoring apparatus main body at a certain time interval, an image at any timing in the periodic glass material supply is used. The evaluation result of the cold top changes depending on whether it is. Therefore, even the technique described in Patent Document 4 cannot satisfactorily monitor the state of the cold top in the electric dissolution tank.

  Furthermore, it is preferable to monitor the inside of the electrolysis tank well so that the electrolysis tank can be appropriately controlled.

  Therefore, the present invention provides an image processing method and an image processing apparatus capable of satisfactorily monitoring the inside of an electric melting tank in which a glass raw material is periodically supplied to the entire ground surface of molten glass to form a cold top. With the goal. Moreover, it aims at providing the control method of the electric dissolution tank which can control an electric dissolution tank appropriately using the image processing method, and the manufacturing method of a glass article.

  An image processing method according to the present invention includes a molten glass element in which a glass raw material is periodically supplied in an electric melting tank that melts a cold top formed on the molten glass by energizing the molten glass. An image processing method using an image obtained by periodically photographing a ground area, wherein the ground area is within one cycle or a plurality of cycles of an operation of supplying a glass material onto a molten glass substrate. Based on the change in the gradation value of the pixel at the common position in the plurality of images obtained by photographing the image, the condition of the cold top at the position in the real space corresponding to the pixel at the position is melted. This is performed with respect to each position of the pixel corresponding to the range of the glass substrate.

  As an evaluation value for evaluating the cold top state, a statistic of a gradation value of pixels at a common position in a plurality of images may be used.

  A maximum value specifying process for specifying a maximum value of gradation values of pixels at a common position in a plurality of images, and a minimum value specifying process for specifying a minimum value of gradation values of pixels at a common position in a plurality of images; At least one of them may be performed, and at least one of the maximum value and the minimum value may be an evaluation value for evaluating a cold top state at a position in the real space corresponding to the pixel at the position.

  A maximum gradation value of pixels at a common position in a plurality of images is specified, a minimum gradation value of pixels at the common position in the plurality of images is specified, and an actual value corresponding to the pixel at the position is specified. As an evaluation value for evaluating the cold top state at a position in the space, a difference value between the maximum value and the minimum value of the gradation values of the pixels at the common position may be calculated.

  An image in which the gradation value at the position of the pixel can be regarded as a desired statistic based on the time at which the raw material is introduced into the position of the real space corresponding to the pixel corresponding to the range of the molten glass substrate May be specified, a gradation value at the position of the pixel may be extracted from the image, and the gradation value may be used as an evaluation value for evaluating the cold top state.

  Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified The first extraction process for extracting the gradation value at the pixel position from the image, and the time when the raw material is introduced at the position in the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass Based on this, at least one of the second extraction processes for specifying an image that can be regarded as having the minimum gradation value at the pixel position and extracting the gradation value at the pixel position from the image. And performing at least one of the gradation value extracted in the first extraction process and the gradation value extracted in the second extraction process on the cold top at the position in the real space corresponding to the pixel at the position. As an evaluation value to evaluate the state Good.

  Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified The first extraction process for extracting the gradation value at the pixel position from the image is executed, and the gradation value at the pixel position can be regarded as the minimum value based on the time. As an evaluation value for performing a second extraction process for extracting the gradation value at the position of the pixel from the image and evaluating the cold top state at the position in the real space corresponding to the pixel at the position The difference value between the gradation value extracted in the first extraction process and the gradation value extracted in the second process may be calculated.

  When a change in gradation value of pixels at a common position in a plurality of images matches a predetermined pattern, a phenomenon corresponding to the predetermined pattern is a position in the real space corresponding to the pixel at the position. It may be determined that it has occurred.

  The image processing apparatus according to the present invention is a molten glass in which a glass raw material is periodically supplied in an electric melting tank that melts a cold top formed on the molten glass by energizing the molten glass. An image processing apparatus that uses an image obtained by periodically photographing a range of a ground surface of the glass substrate, wherein the ground surface is within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the ground surface of the molten glass Evaluating the cold top state at the position in the real space corresponding to the pixel at the position based on the change in the gradation value of the pixel at the common position in the plurality of images obtained by photographing the range of Further, it is characterized by comprising state evaluation means for each position of the pixel corresponding to the range of the ground surface of the molten glass.

  The state evaluation means may use a statistic of the gradation value of pixels at a common position in a plurality of images as an evaluation value for evaluating the cold top state.

  The state evaluation means specifies a maximum value specifying means for specifying a maximum value of gradation values of pixels at a common position in a plurality of images, and a minimum value for specifying a minimum value of gradation values of pixels at a common position in the plurality of images. It may be configured to include at least one of the value specifying means.

  The state evaluation means specifies the maximum value specifying means for specifying the maximum gradation value of the pixel at the common position in the plurality of images, and specifies the minimum value of the gradation value of the pixel at the common position in the plurality of images. And a difference calculating unit that calculates a difference value between the maximum value and the minimum value of the gradation values of the pixels at the common position.

  According to the state evaluation means, the gradation value of the pixel position is a desired statistic based on the time when the raw material is put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass A configuration may be used in which an image that can be regarded is specified and a gradation value at the position of the pixel is extracted from the image.

  The state evaluation means considers that the gradation value at the pixel position is the maximum value based on the time when the raw material is put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass. The first extraction means for identifying an image that can be captured and extracting the gradation value at the position of the pixel from the image, and the raw material is placed at a position in the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass A second extraction unit that identifies an image in which the gradation value at the pixel position can be regarded as a minimum value based on the time, and extracts the gradation value at the pixel position from the image; At least one of the configurations may be provided.

  The state evaluation means considers that the gradation value at the pixel position is the maximum value based on the time when the raw material is put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass. A first extraction unit that identifies an image that can be captured and extracts a gradation value at the position of the pixel from the image, and the gradation value at the position of the pixel is regarded as a minimum value based on the time. A second extraction unit that identifies a possible image and extracts a gradation value at the position of the pixel from the image; a difference between a gradation value extracted by the first extraction unit and a gradation value extracted by the second extraction unit; It may be configured to include difference calculation means for calculating a value.

  A state evaluation unit extracts a gradation value of a pixel at a common position from a plurality of images and generates time-series data of gradation values, and the time-series data matches a predetermined pattern. In this case, a configuration may be provided that includes a phenomenon occurrence determination unit that determines that the phenomenon corresponding to the predetermined pattern occurs at a position in the real space corresponding to the pixel at the position.

  In addition, the method for controlling an electric melting tank according to the present invention melts the cold top formed on the ground surface of the molten glass by energizing the molten glass, and periodically glass raw material on the ground surface of the molten glass. Is a method for controlling an electric melting tank, wherein a plurality of obtained ground images are captured within one cycle or a plurality of cycles of an operation of supplying a glass raw material onto a molten glass substrate. The evaluation of the cold top state at the position in the real space corresponding to the pixel at the position based on the change in the gradation value of the pixel at the common position in the image corresponds to the range of the ground surface of the molten glass. It is performed for each position of the pixel, and the electrolysis tank is controlled according to the evaluation result of the cold top state.

  Further, the method for producing a glass article according to the present invention is formed on the molten glass substrate by energizing the molten glass in an electric melting tank in which the glass raw material is periodically supplied to the molten glass substrate. A melting step for producing a molten glass by melting the cold top formed, a clarification step for removing bubbles of the molten glass in a clarification tank, a molding step for molding the molten glass from which bubbles have been removed, and molding A step of slowly cooling the molten glass, and the range of the ground surface within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the ground surface of the molten glass in the electric melting tank Based on a change in the gradation value of a pixel at a common position in a plurality of images obtained by shooting the image, a cold top at a position in the real space corresponding to the pixel at that position is displayed. Evaluating the state performed for each position of the pixel corresponding to the range of the matrix surface of the molten glass, according to the evaluation result of the cold-top condition, and controlling said electrical dissolving tank.

  ADVANTAGE OF THE INVENTION According to this invention, the inside of the electric melting tank in which a glass raw material is periodically supplied to the whole ground surface of a molten glass and a cold top is formed can be monitored favorably.

1 is an explanatory diagram illustrating an image processing apparatus according to the present invention. The graph which shows the example of the change of the gradation value of the pixel of the common position in several images. The graph which shows the example of the change of the gradation value in each image obtained by image | photographing the inside of an electrolysis tank, when the raw material supply part 82 has approached the rightmost side, or has approached the leftmost side. The block diagram which shows the state evaluation means 10 in the 1st Embodiment of this invention. 5 is a flowchart illustrating an example of processing progress of the image processing apparatus according to the first embodiment. The schematic diagram which shows the example of a maximum gradation image. The schematic diagram which shows the example of the minimum gradation image. The schematic diagram which shows the example of a difference image. FIG. 4 is a schematic diagram showing an example of an image change in conversion by a conversion unit 15. The schematic diagram which shows the process result of step S6. The graph which shows the example of an output of the average value of the gradation value in each area | region. The schematic diagram which shows one pixel in an image. The block diagram which shows the state evaluation means 10 in the 2nd Embodiment of this invention. The schematic diagram which shows the example of conversion with respect to a maximum gradation image. The schematic diagram which shows the example of conversion with respect to the minimum gradation image. The block diagram which shows the state evaluation means 10 in the 3rd Embodiment of this invention. The schematic diagram which shows the example of the change of the gradation value in the pixel which image | photographed the balloon generation | occurrence | production location in the image for 1 period which image | photographed the condition where a balloon generation phenomenon generate | occur | produces and the balloon generation phenomenon converges after that. The schematic diagram which shows the example of the change of the gradation value in the pixel which image | photographed the balloon phenomenon generation | occurrence | production part in the image for 1 period which image | photographed the situation where a balloon phenomenon generate | occur | produces and the balloon continues for a while after that. 10 is a flowchart illustrating an example of processing progress of an image processing apparatus according to a third embodiment. The schematic diagram which shows an example of the manufacturing line of the glass article used with the manufacturing method of the glass article of this embodiment. The flowchart which shows the example of the manufacturing method of the glass article of this embodiment. The typical sectional view at the time of seeing an electric dissolution tank from the side. The typical sectional view at the time of seeing an electric dissolution tank from the insertion side of raw material supply part 82. FIG. Explanatory drawing which shows the example of the movement path | route of the raw material supply part 82 in the electric dissolution tank 81. FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[Embodiment 1]
FIG. 1 is an explanatory diagram showing an image processing apparatus according to the present invention. In FIG. 1, the electric dissolution tank 81 shown together with the image processing apparatus 1 is the same as the electric dissolution tank 81 shown in FIG. 22, and the same elements as those shown in FIG. The detailed description is omitted.

  The electric melting tank 81 is directly energized to the molten glass 85 by the electrode 83 to melt the cold top 86 (glass raw material). Molten glass 85 produced by melting the glass raw material is supplied to the next process from the discharge port 87.

  The raw material supply part 82 repeats the operation | movement of supplying a raw material to the whole ground surface of the molten glass 85, and standing by with a fixed period. Moreover, the raw material supply part 82 may be capable of adjusting the supply amount of the glass raw material during movement. In the following description, the case where the raw material supply unit 82 moves at a constant speed along the forward path and the backward path illustrated in FIG. 24 and supplies the raw material to the entire ground surface of the molten glass 85 will be described as an example. However, if the raw material can be supplied to the entire ground surface of the molten glass 85, the movement path of the raw material supply unit 82 is not limited to the example shown in FIG.

  The raw material supply unit 82 periodically supplies the glass raw material to the entire ground surface of the molten glass 85, so that the state where the cold top 86 is formed on the entire ground surface of the molten glass 85 is maintained. However, for example, depending on the temperature of the molten glass 85 or the increase / decrease of the amount of the molten glass 85 in the electric melting tank 81, even if the glass raw material is supplied, there is a phenomenon that the glass raw material is immediately melted. Can occur. That is, it is not guaranteed that the cold top 86 is always formed.

  The camera 2 is, for example, a visible light camera or an infrared camera. The camera 2 is arranged in the vicinity of the opening portion 88 in a posture capable of photographing the range of the ground surface of the molten glass 85 in the electric melting tank 81. Further, the posture of the camera 2 is kept constant. The camera 2 captures a range of the ground surface of the molten glass 85 at regular intervals, and outputs an image obtained as a result of the capturing to the image processing apparatus 1 together with the capturing time of the image. Since the cold top 86 is present in the range of the ground surface of the molten glass 85, the cold top 86 is photographed, but the ground surface of the molten glass 85 is photographed at a location where the cold top 86 is melted. Accordingly, the cold top 86 and the molten glass 85 are shown in the image obtained as a result of photographing.

  The time information output unit 3 monitors the position of the raw material supply unit 82 and outputs the time when the raw material supply unit 82 exists at a predetermined position to the image processing apparatus 1. The time information output unit 3 outputs to the image processing apparatus 1 the time when the raw material supply unit 82 is closest to the right side and the time when it is closest to the left side in the movement route (see FIG. 24). . Note that “right” and “left” mean “right” and “left”, respectively, when the surface of the molten glass 85 is viewed from the camera 2.

  Further, the time information output unit 3 outputs information on the start time of the operation cycle of the material supply unit 82 to the image processing apparatus 1.

  The image processing apparatus 1 includes a state evaluation unit 10. The state evaluation unit 10 uses a plurality of images obtained within one cycle or a plurality of cycles of the movement of the raw material supply unit 82, and based on the change in the gradation value of pixels at a common position in the plurality of images. The state of the cold top 86 at the position in the real space corresponding to the pixel at that position is evaluated. The state evaluation means 10 performs this evaluation for each pixel in the image corresponding to the range of the ground surface of the molten glass 85. In each embodiment, an example in which the state evaluation unit 10 repeats the process of evaluating the state of the cold top 86 using a plurality of images taken within one cycle of the movement of the raw material supply unit 82 every cycle. I will explain.

  FIG. 2 is a graph illustrating an example of a change in gradation value of pixels at a common position in a plurality of images. The horizontal axis shown in FIG. 2 indicates the elapsed time. A time zone surrounded by a broken line is a time zone in which the raw material supply unit 82 is on standby. The raw material supply unit 82 moves from the position 91 (see FIG. 24) along the forward path, reaches the position 92, returns to the position 91 along the return path, and then waits for a certain period of time. And the raw material supply part 82 repeats this operation | movement. A period from the departure from the position 91 (see FIG. 24) to the next departure from the position 91 is defined as one cycle.

  The vertical axis shown in FIG. 2 is the gradation value. The camera 2 (see FIG. 1) captures an area of the ground surface of the molten glass 95 at regular intervals. Focusing on one pixel position among the plurality of images obtained as a result, FIG. 2 shows an example of a change in the gradation value of the pixel. Further, the time at which the glass material is supplied to the position on the ground surface in the real space corresponding to the pixel position is indicated by an arrow in FIG.

  In FIG. 2, the gradation value changes greatly up and down at successive shooting times. This is because the raw material supply unit 82 moves along the movement path illustrated in FIG. 24, so that the cold top 86 or the raw material supply unit 82 may appear in the pixel position of interest. This is because when the raw material supply unit 82 appears in the pixel, the gradation value of the pixel is greatly reduced.

  FIG. 3 shows the change in the gradation value of the pixel in each image obtained by photographing the inside of the electrolysis tank when the raw material supply unit 82 is on the rightmost side or the leftmost side. Represents. In FIG. 3, the gradation value at the photographing time when the raw material supply unit 82 is closest to the right side or the leftmost side is indicated by a marker. Since the raw material supply unit 82 is located in the vicinity of the wall surface in the electrolysis tank at this photographing time, a cold top is shown at the focused pixel position. Therefore, the change in the marker shown in FIG. 3 represents the change in the brightness of the cold top in the pixel position of interest. Then, unlike the change of the gradation value at the continuous photographing time, the change of the marker does not fluctuate greatly up and down.

  In FIG. 3, as in FIG. 2, the time at which the glass raw material is supplied to the position on the bare ground in the real space corresponding to the pixel position of interest is indicated by an arrow.

  There is a certain tendency in the change in the gradation value of the cold top (the change in the marker shown in FIG. 3) reflected in the pixel position of interest. This tendency will be described. Note that the position on the ground surface of the molten glass 85 in the real space corresponding to the pixel position of interest is referred to as the position on the ground surface of interest.

  Before the glass raw material is supplied to the position on the ground surface of interest (for example, immediately before), the gradation value becomes the highest (see FIG. 3). This is because the cold top is thinnest before the glass raw material is supplied.

  Then, after the glass raw material is supplied to the position on the ground surface of interest, the gradation value gradually decreases (see FIG. 3). This is because the thickness of the cold top is increased by supplying the glass raw material. However, since the glass raw material has not been supplied yet and there is a bright region, the gradation value does not drop rapidly. The glass raw material is sequentially supplied even in the region on the back side of the position on the ground surface of interest, and as the region where the thickness of the cold top increases around the position on the ground surface of interest, the gradation is increased. The value goes down.

  The cold top gradation value reflected at the pixel position of interest gradually decreases as described above, but thereafter, starts to increase and the gradation value gradually increases (see FIG. 3). This is because the cold top is melted and thinned by heating due to energization in the electric melting tank over time.

  However, since the raw material supply unit 82 supplies the glass raw material even in the return path, the gradation value temporarily decreases and then increases again (see FIG. 3).

  Then, before the glass raw material is supplied again to the position on the surface of interest, the gradation value becomes the maximum again.

  Depending on the position of the ground surface of the molten glass 85, the time from the start of the cycle to the glass raw material charging time, or the raw material supplying unit moving in the return path after the glass raw material is input by the raw material supplying unit 82 moving in the forward path There is a difference in the time until the glass raw material is charged depending on 82. Therefore, the general tendency of the gradation value change is as shown in FIG. 3, but the time difference between the time when the glass material is charged and the time when the gradation value is minimum, or the gradation value is the maximum when the glass material is charged. The time difference from the time varies depending on the position of the ground surface of the molten glass 85.

  FIG. 4 is a block diagram showing the state evaluation means 10 in the first embodiment of the present invention. In the first embodiment, the state evaluation unit 10 includes an image holding unit 11, a maximum gradation image generation unit 12, a minimum gradation image generation unit 13, a difference image generation unit 14, a conversion unit 15, and a region. Another average gradation value calculation means 16 and an output control means 17 are provided.

  The image holding means 11 stores each image when the raw material supply unit 82 is closest to the right side and each image when the raw material supply unit 82 is closest to the left side within one cycle. Specifically, each image captured by the camera 2 within one cycle and information of the shooting time are input from the camera 2 to the image holding unit 11. Further, the time information output unit 3 receives information about the time when the raw material supply unit 82 is closest to the right side and the time when the raw material supply unit 82 is closest to the left side. The image holding unit 11 stores an image whose shooting time matches the time when the material supply unit 82 is closest to the right side or the left side among the images shot within one period. The image holding unit 11 does not need to store other images.

  However, the time when the raw material supply unit 82 is closest to the right side or the left side may not completely match the photographing time. For example, if the difference between the time when the raw material supply unit 82 is on the rightmost side and the photographing time is equal to or smaller than a predetermined threshold, the image holding means 11 considers that the two times coincide with each other and the photographing is performed. An image corresponding to the time may be stored. The same applies to the time when the raw material supply unit 82 is on the leftmost side.

  The maximum gradation image generation unit 12 generates a maximum gradation image based on each image stored in the image holding unit 11. The maximum gradation image is an image obtained by selecting a pixel having the maximum gradation value from a plurality of images for each pixel position and combining the selected pixels.

  However, it is preferable that the maximum gradation image generating unit 12 selects each pixel in the region on the right side from the center of the maximum gradation image from among the images when the material supply unit 82 is closest to the left side. . Similarly, it is preferable to select each pixel in the region on the left side from the center of the maximum gradation image from among the images when the material supply unit 82 is closest to the right side. By selecting the pixels in this way, a maximum gradation image with little influence of the light reflected by the raw material supply unit 82 is obtained.

  The minimum gradation image generation unit 13 generates a minimum gradation image based on each image stored in the image holding unit 11. The minimum gradation image is an image obtained by selecting a pixel having a minimum gradation value from a plurality of images for each pixel position and combining the selected pixels.

  However, it is preferable that the minimum gradation image generation unit 13 selects each pixel in the region on the right side from the center of the minimum gradation image from among the images when the material supply unit 82 is closest to the left side. . Similarly, it is preferable to select each pixel in the region on the left side from the center of the minimum gradation image from among the images when the material supply unit 82 is closest to the right side. By selecting the pixels in this way, a minimum gradation image with little influence of the light reflected by the raw material supply unit 82 is obtained.

  The difference image generation means 14 generates a difference image. The difference image is an image in which a value obtained by subtracting the gradation value of the corresponding pixel in the minimum gradation image from the gradation value of the corresponding pixel in the maximum gradation image is assigned as the gradation value for each pixel. It is. That is, the difference image generation means 14 sequentially selects the pixels of the difference image for which the gradation value is to be determined, and corresponds to the selected pixel from the gradation value of the pixel in the maximum gradation image corresponding to the selected pixel. A difference image is generated by subtracting the gradation value of the pixel in the minimum gradation image to be determined and defining the subtraction result as the gradation value of the selected pixel.

  The gradation value of the pixel of the difference image is the difference between the maximum value and the minimum value of the gradation value of that pixel in each image for one period. In the difference image, the gradation value of the pixel corresponding to the range of the ground surface of the molten glass 85 represents the degree to which the cold top is melted. Focusing on one pixel, a large difference between the maximum value and the minimum value of the pixel within one period means that the amount of cold top dissolved is large. On the other hand, a small difference between the maximum value and the minimum value of the pixel within one period means that the amount of cold top dissolved is small. Therefore, the amount of cold top dissolution is large on the ground surface of the real space corresponding to the bright area in the difference image. In addition, the amount of cold top dissolution is small on the ground surface of the real space corresponding to the dark area in the difference image.

  The difference image is an image when the camera is the viewpoint. The converting means 15 converts the difference image into an image whose viewpoint is changed from the position of the camera to directly above the ground surface of the molten glass (in other words, above the surface facing the ground surface). As a result of this conversion, an image when the bare ground is observed from above is obtained as an image representing the amount of cold top dissolved. Hereinafter, an image in which the viewpoint is directly above the ground surface of the molten glass is referred to as a bird's-eye view image.

  The area-specific average gradation value calculation means 16 divides an area corresponding to the range of the molten glass background in the bird's-eye view image into a plurality of areas, and for each divided area, averages the gradation values of the pixels in the area. Calculate the value.

  The bird's-eye view image is an image in which the viewpoint of the difference image is changed to just above the ground surface. Therefore, if the average value of the gradation values calculated for each region is large, it can be said that the amount of cold top dissolution is large in that region. If the average value of the gradation values is small, it can be said that the amount of cold top dissolved is small in that region.

  The output control unit 17 causes the output device (for example, a display device, not shown) to output the average value of the pixel gradation values calculated for each region by the region-specific average gradation value calculation unit 16.

  The image holding means 11, the maximum gradation image generation means 12, the minimum gradation image generation means 13, the difference image generation means 14, the conversion means 15, the area-specific average gradation value calculation means 16, and the output control means 17 are, for example, It is realized by a CPU of a computer that operates according to a program. That is, the CPU may operate as each of the above-described units 11 to 17 according to the program. Moreover, said each means 11-17 may be implement | achieved by the separate unit.

  Next, an example of processing progress of the first embodiment will be described. FIG. 5 is a flowchart illustrating an example of processing progress of the image processing apparatus according to the first embodiment.

  In the electric melting tank 81, the raw material supply part 82 repeats the operation | movement which supplies a glass raw material on the ground surface of the molten glass 85, moving along the path | route illustrated in FIG. During this time, the camera 2 captures an area of the ground surface of the molten glass 85 at regular intervals, and outputs each image obtained as a result to the image holding means 11 together with information on the capturing time. Further, the time information output unit 3 outputs information on the start time of the operation cycle of the raw material supply unit 82 to the image holding unit 11 for each cycle. Further, the time information output unit 3 outputs to the image holding unit 11 information on the time when the raw material supply unit 82 is closest to the right side and the time when the raw material supply unit 82 is closest to the left side.

  The period from the start time of a certain period to the start time of the next period is one period. Among the images generated by the camera 2 within one cycle, the image holding unit 11 is an image when the raw material supply unit 82 is closest to the right side, and the raw material supply unit 82 is closest to the left side. Each image is stored (step S1). The image holding means 11 has an image in which the photographing time matches the time when the raw material supply unit 82 is closest to the right side, and the photographing time is when the raw material supply unit 82 is closest to the right side. An image that matches the time may be stored.

  The maximum gradation image generating means 12 is for each pixel position based on each image (each image when the material supply unit 82 is on the leftmost or rightmost side) stored in step S1. A pixel having the maximum gradation value is selected, and those pixels are combined to generate a maximum gradation image (step S2). In this example, the maximum gradation image generating unit 12 selects each pixel in the region on the right side from the center of the maximum gradation image from among the images when the material supply unit 82 is closest to the left side. Similarly, each pixel in the region on the left side of the center of the maximum gradation image is selected from the images when the material supply unit 82 is closest to the right side.

  An example of the maximum gradation image is schematically shown in FIG. In FIG. 6, the cold top is present and the dark spot is shown in black, and the cold top is present, but since the cold top is thin, the bright spot is shown with a dotted pattern. Show. Moreover, the place which does not have a cold top and is very bright is shown in white. Note that, although there may be slight variations in gradation values in the areas represented by the same color or pattern, they are schematically shown by the same color or the same pattern in the drawing. Further, portions corresponding to structures (for example, wall surfaces) of the electric dissolution tank are indicated by hatching. These points are the same in FIGS. 7, 8, 9, 10, 14, and 15. Since the time at which the glass raw material is charged is shifted for each position of the molten glass base, the time at which the cold top is thinnest is also shifted at each position. It can be said that the maximum gradation image is an image obtained by combining the situation when the cold top at each position is the thinnest into one image. In FIG. 6, a balloon phenomenon is shown at a position indicated by an arrow.

  The minimum gradation image generating means 13 is for each pixel position based on each image (each image when the material supply unit 82 is on the leftmost or rightmost side) stored in step S1. A pixel having the smallest gradation value is selected, and these pixels are combined to generate a minimum gradation image (step S3). In this example, the minimum gradation image generating unit 13 selects each pixel in the region on the right side from the center of the minimum gradation image from among the images when the material supply unit 82 is closest to the left side. Similarly, each pixel in the region on the left side of the center of the minimum gradation image is selected from the images when the material supply unit 82 is closest to the right side.

  An example of the minimum gradation image is schematically shown in FIG. It can be said that the minimum gradation image is an image in which the situation when the cold top at each position appears darkest is combined into one image.

  Next, the difference image generation means 14 sequentially selects the pixels of the difference image for which the gradation value is to be determined, and selects the selected pixel from the gradation value of the pixel in the maximum gradation image corresponding to the selected pixel. A difference image is generated by subtracting the gradation value of the pixel in the corresponding minimum gradation image and determining the result of the subtraction as the gradation value of the selected pixel (step S4).

  An example of the difference image is schematically shown in FIG. In the difference image illustrated in FIG. 8, the region observed in gray is represented by a vertical line pattern. This also applies to FIGS. 9 and 10. The difference image represents the degree of dissolution of the cold top at each position within one cycle.

  And the conversion means 15 converts a difference image into a bird's-eye view image (step S5). FIG. 9 shows an example of an image change in conversion by the conversion means 15. FIG. 9A shows a difference image before conversion. FIG. 9B shows a bird's-eye view image when the difference image shown in FIG. 9A is converted. And in the bird's-eye view image shown in FIG.9 (b), only the part applicable to the range of the ground surface of a molten glass is extracted and shown.

  The area-specific average gradation value calculation means 16 divides a portion corresponding to the range of the ground surface of the molten glass into a plurality of areas in the bird's-eye view image. Then, the area-specific average gradation value calculating means 16 calculates the average value of the gradation values of the pixels in the area for each divided area (step S6). FIG. 10 is a schematic diagram showing the processing result of step S6. FIG. 10 illustrates a case where the bird's-eye view image illustrated in FIG. 9B is divided into nine regions. Further, in each region, the code shown in the upper part is identification information of each region, and the value shown in the lower part is an average value of the gradation values of the pixels in the region. Note that the number of divisions of the bird's-eye view image is not limited to nine.

  Next, the output control means 17 causes the output device to output the average value of the gradation values calculated for each region in Step 6 (Step S7). Here, the case where it displays on a display apparatus is made into an example.

  The state evaluation means 10 repeats the processing of steps S1 to S7 for each movement cycle of the raw material supply unit 82. In step S7, the output control means 17 causes the display device to display the average value of the gradation value in each region calculated for each period as a graph, and visually shows the change in the average value of the gradation value in each region. May be represented. FIG. 11 is a graph showing an output example of an average value of gradation values in each region. FIG. 11 illustrates a case where the gradation values in the regions 1L, 2L, 1C, 2C, 3C, 1R, and 2R are output among the regions illustrated in FIG. As shown in FIG. 11, the change in the output at each step 7 is graphed and output, so that it becomes easy to grasp the change in the situation in each region or the entire range of the ground surface. For example, it can be seen from the graph illustrated in FIG. 11 that the brightness of the cold top once increases (in other words, the cold top becomes lighter) and then returns to the original state. From this, it is estimated that the temperature of the molten glass amount temporarily increased because the molten glass supply amount from the electric melting vessel 81 to the next process increased and the molten glass amount in the electric melting vessel 81 decreased. it can.

  In the present embodiment, an image (maximum gradation image) obtained by combining the situation when the cold top at each position is thinnest into one image, or the situation when the cold top at each position appears darkest is one image. The image (minimum gradation image) synthesized is generated, and the condition of the cold top existing in the range of the ground surface of the molten glass is evaluated based on these images. Therefore, even if the supply time of the glass raw material is different at each position of the molten glass base by supplying the glass raw material while the raw material supply unit 82 moves, the situation of the cold top on the base of the molten glass is determined. It can be evaluated uniformly.

  When the techniques described in Patent Documents 3 and 4 are applied to the monitoring of the cold top in the electric melting tank, there is a problem that the evaluation result changes depending on which timing image is used in the periodic glass material supply. there were. On the other hand, in the present invention, the cold top situation is evaluated using an image (a maximum gradation image or a minimum gradation image) obtained by combining the cold top situation at each position into one image. Therefore, as described above, the situation of the cold top on the bare ground can be evaluated uniformly. Accordingly, it is possible to satisfactorily monitor the inside of the electric melting tank in which the glass raw material is periodically supplied to the entire surface of the molten glass and the cold top is formed.

  Further, the gradation value (that is, brightness) of the image is used as an evaluation value for evaluating the cold top situation. Therefore, not only the thickness of the cold top but also the thickness of the cold top and the temperature of the molten glass existing under the cold top can be comprehensively evaluated.

[Embodiment 2]
As shown in FIG. 3, the change in the brightness of the cold top shown in one pixel within one cycle has the above-described tendency (see the marker shown in FIG. 3). In the second embodiment, using this tendency, for each pixel, an image that can be regarded as having the maximum gradation value is selected, and the pixels of the selected image are synthesized for each pixel. A maximum gradation image is generated. Similarly, for each pixel, an image that can be regarded as having the minimum gradation value of the pixel is selected, and the pixels of the selected image are combined for each pixel to generate a minimum gradation image.

  In the second embodiment, for each individual pixel position, a pixel that can be regarded as having the maximum gradation value is selected from a plurality of images, and an image obtained by combining the selected pixels is referred to as a maximum gradation image. Define. In addition, for each pixel position, a pixel that can be regarded as having the minimum gradation value is selected from a plurality of images, and an image obtained by combining the selected pixels is defined as a minimum gradation image.

  The image processing apparatus according to the second embodiment also includes a state evaluation unit 10 as in the case shown in FIG. However, the state evaluation means 10 in the second embodiment will be described later. The electric dissolution tank 81 has already been described and will not be described.

  The camera 2 is the same as the camera 2 in the first embodiment.

  Similarly to the first embodiment, the time information output means 3 is the time when the raw material supply unit 82 is closest to the right side and the time when it is close to the left side in the movement route (see FIG. 24). The time is output to the image processing apparatus 1. Information about the start time of each cycle is also output to the image processing apparatus 1.

  Furthermore, in the second embodiment, the time information output unit 3 sets the raw material input time to the position on the ground surface in the real space corresponding to the pixel of the image obtained by the camera 2 for each pixel. Output to. FIG. 12 is a schematic diagram showing one pixel in an image. For example, when the tip of the raw material supply unit 82 moves to a position on the ground surface of the molten glass in the real space corresponding to the pixel 21 in the image 20, the time information output unit 3 corresponds to the pixel 21 with the current time. It outputs to the image processing apparatus 1 as a raw material injection | throwing-in time to the position on the bare ground in real space. Here, the pixel 21 in the image 20 has been described as an example. However, the time information output means 3 is provided for each pixel corresponding to the range of the ground surface of the molten glass in the real space corresponding to the pixel. The raw material charging time at a position on the ground is output to the image processing apparatus 1. Hereinafter, the glass raw material charging time at a position on the ground surface in a real space corresponding to a certain pixel is simply referred to as a glass raw material charging time at a position corresponding to the pixel.

  FIG. 13 is a block diagram showing the state evaluation means 10 in the second exemplary embodiment of the present invention. Elements similar to those of the first embodiment are denoted by the same reference numerals as those in FIG. In the second embodiment, the state evaluation unit 10 includes an image holding unit 11, a maximum gradation image generation unit 12a, a minimum gradation image generation unit 13a, a difference image generation unit 14, a conversion unit 15, and a region. Another average gradation value calculation means 16 and an output control means 17 are provided.

  Each image taken by the camera 2 within one cycle and information on the photographing time are input from the camera 2 to the image holding unit 11. In addition, the time information output unit 3 inputs the time when the raw material supply unit 82 is closest to the right side, the time information when the raw material supply unit 82 is closest to the left side, and the start time of each week machine. Then, as in the first embodiment, the image holding unit 11 has each image when the raw material supply unit 82 is closest to the right side within one cycle, and the raw material supply unit 82 is closest to the left side. Each image is memorized.

  The maximum gradation image generating means 12a is inputted with the time for outputting the glass raw material at the position corresponding to the pixel from the time information output means 3 for each pixel corresponding to the range of the ground surface of the molten glass. In addition, the maximum gradation image generating means 12a inputs the glass raw material to the position corresponding to each pixel in the image 20 generated by the camera 2 (see FIG. 12) for each pixel corresponding to the range of the ground surface of the molten glass. A time that is a difference between the time and the time at which the gradation value of the pixel can be regarded as the maximum is stored in advance. This difference time is referred to as a difference time relating to the maximum gradation. The difference time related to the maximum gradation may be determined in advance for each pixel corresponding to the range of the molten glass, and stored in the maximum gradation image generating means 12a. As already described, the gradation value of the pixel of interest tends to be highest before the glass raw material is supplied to the position on the ground surface of interest (for example, immediately before). Therefore, with respect to the pixel of interest, an image immediately before the raw material is supplied to the position on the ground surface in the real space corresponding to the pixel position as “an image that can be regarded as the maximum gradation value of the pixel” Can be adopted. Therefore, the difference time regarding the maximum gradation may be defined as a negative value (for example, −x seconds). In this case, it means that the gradation value of the pixel can be regarded as the maximum in the image at the time x seconds before the glass raw material charging time at the position corresponding to the pixel. Note that the difference time may be set to be different for each pixel position.

  The maximum gradation image generating means 12a adds the difference time relating to the maximum gradation predetermined for the pixel to the glass raw material charging time at the position corresponding to the pixel, so that the gradation value of the pixel is maximized. Calculate the time that can be considered to be. The maximum gradation image generating unit 12a selects an image at the shooting time closest to the time from the images held by the image holding unit 11, and further selects a pixel at a pixel position of interest from the image.

  For example, the maximum gradation image generating unit 12a adds the difference time relating to the maximum gradation predetermined for the pixel 21 to the glass raw material charging time at the position corresponding to the pixel 21 shown in FIG. As a result, a time at which the gradation value of the pixel 21 can be regarded as the maximum is obtained. The maximum gradation image generating unit 12a selects an image at the photographing time closest to the time from the images held by the image holding unit 11, and further selects a pixel corresponding to the position of the pixel 21 from the image.

  Similarly, the maximum gradation image generation means 12a generates a maximum gradation image by selecting pixels for each pixel corresponding to the range of the ground surface of the molten glass and synthesizing those pixels. Since the pixels in the region other than the range of the molten glass substrate are not subject to evaluation, the gradation value may be set to a constant value.

  When selecting an image at the photographing time closest to the time at which the pixel gradation value can be regarded as the maximum, if the pixel is a pixel in the region on the right side of the center, the material supply unit 82 is on the leftmost side. It is preferable to select an image at the photographing time closest to the time at which the gradation value can be regarded as the maximum from each image when approaching. In addition, when the pixel is a pixel in the region on the left side from the center, the shooting time closest to the time at which the gradation value can be regarded as the maximum among the images when the raw material supply unit 82 is on the rightmost side. It is preferable to select an image.

  The glass material input time at the position corresponding to the pixel is input from the time information output unit 3 to the minimum gradation image generating unit 13a for each pixel corresponding to the range of the molten glass. Further, the minimum gradation image generation means 13a is configured to input a glass raw material to a position corresponding to each pixel corresponding to the range of the molten glass in the image 20 generated by the camera 2 (see FIG. 12). A time that is a difference between the time and the time at which the gradation value of the pixel can be regarded as the minimum is stored in advance. This difference time is referred to as a difference time relating to the minimum gradation. The difference time relating to the minimum gradation may be determined in advance for each pixel corresponding to the range of the molten glass, and stored in the minimum gradation image generating means 13a. As already described, after the glass raw material is supplied to the position on the ground surface of interest, the gradation value of the pixel of interest gradually decreases and then increases. Therefore, after the raw material is supplied to the position on the ground surface in the real space corresponding to the pixel position as the “image that can be regarded as the minimum pixel gradation value” for the pixel of interest. Images can be adopted. In this way, when the image after supplying the raw material to the position on the ground surface of interest is adopted as the “image that can be regarded as the minimum pixel gradation value”, the difference time is set to a positive value. It should be determined as. An image immediately after the raw material supply may be employed as the “image that can be regarded as having the smallest pixel gradation value”. In addition, as an “image that can be regarded as having the smallest pixel gradation value”, an image at a time prior to the time at which the raw material is supplied to the position on the ground surface of interest may be employed. Good. In this case, the difference time regarding the minimum gradation may be determined as a negative value (for example, −x seconds). In this case, it means that the gradation value of a pixel can be regarded as the minimum in an image at a time x seconds before the glass raw material charging time at a position corresponding to the pixel. Note that the difference time may be set to be different for each pixel position.

  The minimum gradation image generation means 13a adds the difference time relating to the minimum gradation predetermined for the pixel to the glass material charging time at the position corresponding to the pixel, thereby minimizing the gradation value of the pixel. Calculate the time that can be considered to be. The minimum gradation image generation unit 13a selects an image at the shooting time closest to the time from the images held by the image holding unit 11, and further selects a pixel at a pixel position of interest from the image.

  For example, the minimum gradation image generating unit 13a adds the difference time related to the minimum gradation that is predetermined for the pixel 21 to the glass material charging time at the position corresponding to the pixel 21 shown in FIG. As a result, a time at which the gradation value of the pixel 21 can be regarded as the minimum is obtained. The minimum gradation image generation unit 13a selects an image at the shooting time closest to the time from the images held by the image holding unit 11, and further selects a pixel corresponding to the position of the pixel 21 from the image.

  Similarly, the minimum gradation image generation means 13a generates a minimum gradation image by selecting pixels for each pixel corresponding to the range of the molten glass, and synthesizing those pixels. Since the pixels in the region other than the range of the molten glass substrate are not subject to evaluation, the gradation value may be set to a constant value.

  When selecting an image at the photographing time closest to the time at which the gradation value of the pixel can be regarded as the minimum, if the pixel is a pixel in the region on the right side of the center, the material supply unit 82 is on the leftmost side. It is preferable to select an image at the photographing time closest to the time at which the gradation value can be regarded as the smallest from among the images when approaching. Further, when the pixel is a pixel in the region on the left side from the center, the shooting time closest to the time at which the gradation value can be regarded as the minimum among the images when the raw material supply unit 82 is on the rightmost side. It is preferable to select an image.

  The difference image generation unit 14 generates a difference image using the maximum gradation image generated by the maximum gradation image generation unit 12a and the minimum gradation image generated by the minimum gradation image generation unit 13a. The difference image generation process is the same as that in the first embodiment.

  The conversion unit 15, the area-specific gradation value calculation unit 16, and the output control unit 17 are the same as those in the first embodiment.

  The maximum gradation image generation unit 12a and the minimum gradation image generation unit 13a are realized by a CPU of a computer that operates according to a program, for example. Alternatively, each may be realized by an independent unit.

  The processing progress of the image processing apparatus in the second embodiment can be represented by the same flowchart as in FIG. Step S1 is the same as step S1 in the first embodiment.

  In step S2, the maximum gradation image generation means 12a generates a maximum gradation image. Since the generation processing of the maximum gradation image in the second embodiment has already been described, the description is omitted here.

  In step S3, the minimum gradation image generation means 13a generates a minimum gradation image. Since the generation process of the minimum gradation image in the second embodiment has already been described, the description is omitted here.

  Steps S4 to S7 are the same as steps S4 to S7 in the first embodiment, and a description thereof will be omitted.

  Also in 2nd Embodiment, the state evaluation means 10 repeats the process of step S1-S7 for every period of a motion of the raw material supply part 82. FIG.

  In the second embodiment, it can be said that the maximum gradation image is an image obtained by selecting pixels that can be regarded as having the thinnest cold top from a plurality of images and combining them into one image. Further, it can be said that the minimum gradation image is an image obtained by selecting, from a plurality of images, a pixel in which the cold top can be regarded as the darkest image and combining it into one image. And the state evaluation means 10 evaluates the condition of the cold top which exists in the range of the ground surface of a molten glass based on those images. Therefore, even if the supply time of the glass raw material is different at each position on the molten glass base by supplying the glass raw material while the raw material supply unit 82 is moving, the molten glass of the molten glass is the same as in the first embodiment. The condition of the cold top on the ground surface can be evaluated uniformly.

Next, modifications of the first embodiment and the second embodiment will be described.
First, a first modification will be described. You may control an electrolysis tank based on the average value of the gradation value of the pixel in each area | region (refer FIG. 10) obtained by step S6 of 1st Embodiment and 2nd Embodiment.

  For example, in the region where the average value of the gradation values calculated in step S6 is high, it is considered that the temperature of the molten glass is high and the cold top is immediately melted. Therefore, in the region where the average value of the gradation values calculated in step S6 is higher than the predetermined first predetermined value, the amount of the glass raw material supplied from the raw material supply unit 82 may be increased. Or you may reduce the amount of electricity with respect to the molten glass of the area | region, and may reduce the temperature of the molten glass of the area | region. As a result, it is possible to prevent the cold top from being dissolved immediately.

  Conversely, in the region where the average gradation value calculated in step S6 is low, the temperature of the molten glass is low, and the cold top is considered to be thicker than the other regions. Therefore, in the region where the average value of the gradation values calculated in step S6 is lower than the predetermined second predetermined value, the amount of the glass material supplied from the material supply unit 82 may be reduced. Or the energization amount with respect to the molten glass of the area | region may be raised, and the temperature of the molten glass of the area | region may be raised. As a result, the cold top that is thicker than the other regions can have the same thickness as the cold top in the other regions. Note that the second predetermined value is a value lower than the first predetermined value.

Next, a second modification of the first embodiment and the second embodiment will be described.
In 1st Embodiment and 2nd Embodiment, the case where a difference image was produced | generated and the bird's-eye view image of a difference image was produced | generated was shown. On the other hand, one or both of the maximum gradation image and the minimum gradation image may be generated without generating the difference image. In this case, the state evaluation unit 10 does not include the difference image generation unit 14 illustrated in FIGS. 4 and 13 and does not have to execute Step S4 (see FIG. 5). In this case, the conversion unit 15 may convert either one or both of the maximum gradation image and the minimum gradation image into a bird's-eye view image in step S5. When both the maximum gradation image and the minimum gradation image are converted into bird's-eye images, the state evaluation unit 10 only needs to execute the processes of steps S6 and S7 for each bird's-eye image.

  In the case where only the maximum gradation image is generated among the maximum gradation image and the minimum gradation image, the state evaluation unit 10 uses the minimum gradation image generation unit 13 shown in FIG. 4 (or the minimum gradation image shown in FIG. 13). The image generation means 13a) is not provided, and step S3 (see FIG. 5) need not be executed. When only the minimum gradation image is generated among the maximum gradation image and the minimum gradation image, the state evaluation unit 10 is configured to output the maximum gradation image generation unit 12 shown in FIG. 4 (or the maximum gradation image shown in FIG. 13). The image generating means 12a) is not provided, and step S2 (see FIG. 5) need not be executed.

  FIG. 14 shows an example of conversion for a maximum gradation image. FIG. 14A shows a maximum gradation image before conversion. FIG. 14B shows a bird's-eye view image obtained by converting the maximum gradation image shown in FIG. However, in the bird's-eye view image shown in FIG. 14B, only the portion corresponding to the range of the ground surface of the molten glass is extracted and shown. With the bird's-eye view image illustrated in FIG. 14B, the situation when the cold top becomes the thinnest can be uniformly evaluated with respect to each position on the ground surface of the molten glass. In the example shown in FIG. 14 (b), it can be seen that the region where the molten glass is exposed increases when the cold top becomes the thinnest side from the center.

  FIG. 15 shows an example of conversion for a minimum gradation image. FIG. 15A shows a minimum gradation image before conversion. FIG. 15B shows a bird's-eye view image obtained by converting the minimum gradation image shown in FIG. However, in the bird's-eye view image shown in FIG. 15B, only the portion corresponding to the range of the ground surface of the molten glass is extracted and shown. With the bird's-eye view image illustrated in FIG. 15B, the situation when the cold top appears darkest at each position can be uniformly evaluated with respect to each position on the molten glass substrate. In a bright portion of the bird's-eye view image, even if the raw material is added, the cold top dissolves immediately, and the cold top is not maintained. In the example shown in FIG. 15B, it can be seen that such a region slightly exists on the left side.

  Further, the electrolysis tank may be controlled based on the conversion result of the maximum gradation image or the minimum gradation image obtained in this modification, as in the first modification described above.

  In the first and second embodiments, the maximum value (or a value that can be regarded as the maximum value) and the minimum value (or the value that can be regarded as the minimum value) are used as the statistic of the gradation value. Although the cold top state is evaluated using, the cold top may be evaluated using a statistic other than the maximum value and the minimum value. For example, a histogram may be created with the gradation values of pixels at a common position, and the cold top state may be evaluated based on the gradation values having a predetermined frequency order. Or you may evaluate the state of a cold top using the intermediate value of a gradation value.

[Embodiment 3]
As described above, when attention is paid to a certain pixel, the change in the gradation value of the cold top reflected at the pixel position has a certain tendency (see the change in the marker shown in FIG. 3). However, when a sudden phenomenon such as a balloon phenomenon occurs, the change in gradation value shows another tendency. In the third embodiment, information indicating a pattern of change in gradation value of a pixel that has captured a predetermined phenomenon is stored in advance, and when it is determined that the change in gradation value of a pixel corresponds to the pattern, It is determined that the phenomenon has occurred at the position in the real space corresponding to the pixel.

  The image processing apparatus according to the third embodiment also includes a state evaluation unit 10 as in the case shown in FIG. However, the components included in the state evaluation means 10 are different from those in the first and second embodiments. The electric dissolution tank 81 has already been described and will not be described.

  The time information output means 3 is the same as the time information output means 3 in the first embodiment. That is, in the movement route (see FIG. 24), the time when the raw material supply unit 82 is closest to the right side and the time when it is close to the left side are output to the image processing apparatus 1. Information about the start time of each cycle is also output to the image processing apparatus 1.

  The camera 2 is the same as the camera 2 in the first and second embodiments.

  FIG. 16 is a block diagram showing the state evaluation means 10 in the third embodiment of the present invention. In the third embodiment, the state evaluation unit 10 includes an image holding unit 11, a gradation value time-series data extraction unit 31, a phenomenon data storage unit 32, a phenomenon occurrence determination unit 33, and an output control unit 34. Prepare.

  The image holding unit 11 is the same as the image holding unit 11 in the first and second embodiments, and a description thereof will be omitted.

  The gradation value data extraction unit 31 is configured to calculate the gradation of the pixel corresponding to the pixel position for each pixel position corresponding to the range of the molten glass from the image for one period stored in the image holding unit 11. A value is extracted to generate time-series data of gradation values.

  The gradation value data extracting means 31 preferably extracts the gradation value of the pixel in the region on the right side of the center from each image when the material supply unit 82 is closest to the left side. Further, it is preferable that the gradation value data extraction unit 31 extracts the gradation value of the pixel in the left region from the center from each image when the material supply unit 82 is on the leftmost side.

  The phenomenon data storage means 32 stores data (hereinafter referred to as phenomenon data) indicating a pattern of change in gradation value in a pixel that has copied a predetermined phenomenon. In the present embodiment, the balloon phenomenon will be described as an example of the predetermined phenomenon.

  FIG. 17 is a schematic diagram illustrating an example of a change in gradation value in a pixel in which a balloon phenomenon occurrence location is captured in an image for one cycle in which a balloon phenomenon occurs and then the balloon phenomenon converges. When the balloon phenomenon occurs and then the balloon phenomenon converges, as shown in FIG. 17, the gradation value increases rapidly, and then the gradation value decreases rapidly. Therefore, in association with the phenomenon that “the balloon phenomenon occurs and then the balloon phenomenon converges”, data indicating that the gradation value suddenly increases and then the gradation value rapidly decreases is used as the phenomenon data. The phenomenon data storage means 32 may be stored. Note that the sudden increase in the gradation value and the decrease in the gradation value may be quantified and represented by the gradient (differential value) of the gradation value.

  FIG. 18 is a schematic diagram illustrating an example of a change in gradation value in a pixel in which a balloon phenomenon occurrence location is captured in an image for one cycle in which a balloon phenomenon occurs and the balloon continues for a while thereafter. When the ballooning phenomenon occurs and the ballooning continues for a while after that, as shown in FIG. 18, the gradation value increases rapidly, and then the gradation value becomes a substantially constant value. Therefore, in association with the phenomenon that “the balloon phenomenon occurs and the balloon continues for a while after that”, the gradation value suddenly increases, and thereafter, the data indicating that the gradation value becomes almost constant What is necessary is just to memorize | store in the phenomenon data storage means 32 as data.

  In this embodiment, the phenomenon data corresponding to the phenomenon “the balloon phenomenon occurs and then the balloon phenomenon converges” and the phenomenon data corresponding to the phenomenon “the balloon phenomenon occurs and the balloon continues for a while thereafter” Will be described by taking the phenomenon data storage means 32 as an example. However, the phenomenon data stored by the phenomenon data storage unit 32 is not limited to these phenomenon data.

  The phenomenon occurrence determination unit 33 determines whether the time-series data of the gradation values generated for each pixel position by the gradation value data extraction unit 31 matches the phenomenon data stored in the phenomenon data storage unit 32. To do. If the time-series data matches the phenomenon data, it is determined that a phenomenon corresponding to the phenomenon data has occurred on the ground surface in the real space corresponding to the pixel position.

  The output control unit 34 causes an output device (for example, a display device, not shown) to output a determination result of whether or not a phenomenon has occurred at each position.

  The image holding unit 11, the gradation value data extraction unit 31, the phenomenon occurrence determination unit 33, and the output control unit 34 are realized by a CPU of a computer that operates according to a program, for example. That is, the CPU may operate as each of the above means according to the program. Each of the above means may be realized by separate units.

  FIG. 19 is a flowchart illustrating an example of processing progress of the image processing apparatus according to the third embodiment. Among the images generated by the camera 2 within one cycle, the image holding unit 11 is an image when the raw material supply unit 82 is closest to the right side, and the raw material supply unit 82 is closest to the left side. Each image is stored (step S1). This process is the same as step S1 in the first and second embodiments.

  The gradation value data extracting means 31 determines whether or not there is a pixel position that has not yet been selected in step S12 to be described later among the pixel positions corresponding to the range of the molten glass background in each image (step S12). S11).

  When it is determined that there is an unselected pixel position (Yes in step S11), the gradation value data extraction unit 31 sets the unselected pixel position to 1 among the pixel positions corresponding to the range of the ground surface of the molten glass. Is selected (step S12).

  Then, the gradation value data extraction unit 31 extracts the gradation value of the pixel corresponding to the selected pixel position from the image stored by the image holding unit 11 and arranges the gradation values in order of photographing time. Data is generated (step S13). Here, as an example, the gradation value of the pixel in the region on the right side of the center is extracted from each image when the material supply unit 82 is closest to the left side. Further, as an example, the gradation value of the pixel in the region on the left side of the center is extracted from each image when the material supply unit 82 is on the leftmost side.

  The phenomenon occurrence determination unit 33 determines whether the time series data generated in step S13 matches the phenomenon data. If the time-series data matches the phenomenon data, the phenomenon occurrence determination unit 33 has a phenomenon corresponding to the phenomenon data at a position on the ground surface in the real space corresponding to the pixel position selected in step S12. Is determined to have occurred. If the time-series data does not match the phenomenon data, it is determined that the phenomenon corresponding to the phenomenon data does not occur at the position on the ground surface in the real space corresponding to the pixel position selected in step S12. (Step S14).

  For example, if the time-series data matches the phenomenon data that “the gradation value suddenly increases and then the gradation value suddenly decreases”, the phenomenon occurrence determination unit 33 determines that the “blowing phenomenon”. It is determined that the phenomenon that the ballooning phenomenon has converged thereafter occurs.

  Further, for example, if the time series data matches the phenomenon data that “the gradation value suddenly increases and then the gradation value becomes substantially constant”, the phenomenon occurrence determination means 33 , It is determined that the phenomenon that “the balloon phenomenon occurs and the balloon continues for a while after that” has occurred.

After step S14, the process of shifting to step S11 is repeated. In step S11, when it is determined that there is no unselected pixel position among the pixel positions corresponding to the range of the molten glass base (No in step S11), the output control means 15 The output device (for example, display device) is made to output the determination result of the occurrence / non-occurrence of the phenomenon at each position in the real space corresponding to each pixel position corresponding to the range (step S15). In step S15, the position where it is determined that the phenomenon has occurred and the content of the phenomenon may be output.

  The state evaluation means 10 repeats the processing of steps S1 to S15 shown in FIG. 19 for each movement cycle of the raw material supply unit 82.

  According to the present embodiment, for example, it can be determined whether or not a predetermined phenomenon such as a blowing phenomenon occurs on a cold top of a molten glass. Further, when such a predetermined phenomenon occurs, the place where the phenomenon occurs can also be specified.

  Moreover, you may control the electrolysis tank 81 according to the phenomenon which generate | occur | produced. For example, the amount of glass raw material supplied to a place where a predetermined phenomenon has occurred may be changed. Alternatively, the energization amount at a place where a predetermined phenomenon has occurred may be changed.

  In each of the above embodiments, the camera 2 captures an area of the ground surface of the molten glass 85 in the electric melting tank 81 at regular time intervals. At this time, the camera 2 may continuously capture images at each capturing timing, or may capture a single image.

[Embodiment 4]
Next, a glass article manufacturing method will be described as a fourth embodiment of the present invention. The image processing method described in the first to third embodiments is applied to the glass article manufacturing method of the present invention. In the following description, a method for manufacturing a glass article to which the first embodiment is applied will be described. FIG. 20 is a schematic diagram illustrating an example of a glass article production line used in the method for producing a glass article of the present embodiment. In FIG. 20, the camera 2 is shown, but the time information output means 3 and the image processing apparatus 1 (see FIG. 1) are not shown. An electrolysis tank 81 and a clarification tank 50 are provided in the glass article production line.

  The electric dissolution tank 81 has already been described and will not be described. The molten glass 85 in the electric melting tank 81 is supplied to the clarification tank 50.

  The clarification tank 50 removes bubbles generated in the molten glass 85. The clarification tank 50 may be a reduced pressure type clarification tank that removes bubbles by reducing the inside of the tank to a reduced pressure state. Or the high temperature type clarification tank etc. which make the inside of a tank high temperature and remove a bubble may be sufficient. The molten glass from which the bubbles have been removed moves to a molding process and a slow cooling process.

  FIG. 21 is a flowchart showing an example of a method for manufacturing a glass article of the present embodiment. First, in the electric melting tank 81, the raw material supply unit 82 periodically supplies the glass raw material to the entire ground surface of the molten glass 85, thereby forming the cold top 86 on the ground surface of the molten glass 85. And the cold top 86 is melt | dissolved by the heat | fever which arises by supplying with electricity to the molten glass 85 using the electrode 83, and the molten glass 85 is newly manufactured (step S31, glass melting process).

  In step S31, the camera 1 photographs the inside of the electrolysis tank 81, and the image processing apparatus 1 performs the same processing as the first embodiment on the image obtained as a result (see steps S1 to S7, see FIG. 5). Do. And the electrolysis tank 81 is controlled based on the average value of the gradation value of the pixel in each area | region (refer FIG. 10) obtained by step S6.

  The molten glass 85 manufactured in step S <b> 31 is flowed to the clarification tank 50. Bubbles are present in the molten glass 85, and a bubble layer is formed on the surface of the molten glass 85. Inside the clarification tank 50, bubbles generated in the molten glass 85 are removed (step S32, clarification step).

  After step S32, the molten glass from which bubbles have been removed is molded (step S33, molding process). In the forming step, for example, molten glass may be formed by a float process. Specifically, the molten glass 85 from which bubbles have been removed is floated on molten tin (not shown) and is advanced in the conveying direction to form a continuous plate-like glass ribbon. At this time, in order to form a glass ribbon having a predetermined plate thickness, a rotating roll is pressed against both side portions of the glass ribbon, and the glass ribbon is stretched outward in the width direction (direction perpendicular to the conveying direction).

  Next, the glass ribbon formed in step S33 is gradually cooled (step S34, slow cooling step). In the slow cooling step, the glass ribbon is drawn out from the molten tin, and the glass ribbon is gradually cooled inside a slow cooling furnace (not shown). Even after being transported outside the slow cooling furnace, the glass ribbon is gradually cooled to near room temperature.

  After the slow cooling step, the glass ribbon solidified in the slow cooling step is processed as necessary (step S35, processing step). Examples of the processing in step S35 include cutting and polishing. However, the present invention is not limited to cutting and polishing, and other processing may be performed.

  According to the method for manufacturing a glass article described above, the inside of the electric melting tank 81 can be well monitored, and the electric melting tank 81 can be appropriately controlled according to the cold top state. As a result, the quality of the glass article can be improved.

  Moreover, in said 4th Embodiment, the case where 1st Embodiment was used by step S31 was shown. In step S31, the same processing as in the second embodiment may be performed. In step S31, the same processing as that in the third embodiment (steps S1 to S15, see FIG. 19) may be performed.

  The present invention can be suitably applied to monitoring the inside of an electric dissolution tank.

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 2 Camera 3 Time information output means 10 State evaluation means 11 Image holding means 12, 12a Maximum gradation image generation means 13, 13a Minimum gradation image generation means 14 Difference image generation means 15 Conversion means 16 Average floor by area Gradation value calculation means 17, 34 Output control means 31 Gradation value time-series data extraction means 32 Phenomenon data storage means 33 Phenomenon occurrence determination means 81 Electric dissolution tank 82 Raw material supply section 83 Electrode

Claims (18)

Periodically photographing the range of the molten glass substrate, where the glass raw material is supplied periodically, in an electric melting tank that melts the cold top formed on the molten glass substrate by energizing the molten glass. An image processing method using an image obtained by
Changes in gradation values of pixels at a common position in a plurality of images obtained by photographing the range of the ground surface within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the molten glass substrate An image processing method comprising: evaluating a cold top state at a position in a real space corresponding to a pixel at the position with respect to each position of a pixel corresponding to a range of a ground surface of molten glass. . The image processing method according to claim 1, wherein a statistic of a gradation value of pixels at a common position in a plurality of images is used as an evaluation value for evaluating a cold top state. A maximum value specifying process for specifying a maximum value of gradation values of pixels at a common position in a plurality of images; and
Performing at least one of minimum value specifying processing for specifying the minimum value of gradation values of pixels at a common position in a plurality of images;
The image processing method according to claim 1, wherein at least one of the maximum value and the minimum value is an evaluation value for evaluating a cold top state at a position in the real space corresponding to the pixel at the position. Specify the maximum gradation value of pixels at a common position in multiple images,
Specifying a minimum value of gradation values of pixels at the common position in the plurality of images;
The difference value between the maximum value and the minimum value of the gradation values of the pixels at the common position is calculated as an evaluation value for evaluating the cold top state at the position in the real space corresponding to the pixel at the position. The image processing method according to any one of claims 1 to 3. An image in which the gradation value at the position of the pixel can be regarded as a desired statistic based on the time at which the raw material is introduced into the position of the real space corresponding to the pixel corresponding to the range of the molten glass substrate The image processing method according to claim 1, further comprising: extracting a gradation value at a position of the pixel from the image and using the gradation value as an evaluation value for evaluating a cold top state. Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified A first extraction process for extracting a gradation value at the position of the pixel from the image; and
Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, an image that can be considered that the gradation value at the position of the pixel is the minimum value is specified And performing at least one of the second extraction processing for extracting the gradation value at the position of the pixel from the image,
At least one of the gradation value extracted in the first extraction process and the gradation value extracted in the second extraction process is used to evaluate a cold top state at a position in the real space corresponding to the pixel at the position. The image processing method according to claim 1, wherein the evaluation value is an evaluation value. Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified And executing a first extraction process for extracting a gradation value at the position of the pixel from the image,
Based on the time, an image in which the gradation value at the pixel position can be regarded as the minimum value is specified, and a second extraction process is performed to extract the gradation value at the pixel position from the image And
As an evaluation value for evaluating the cold top state at the position in the real space corresponding to the pixel at the position, the difference between the gradation value extracted by the first extraction process and the gradation value extracted by the second process The image processing method according to any one of claims 1, 5, and 6. When a change in gradation value of pixels at a common position in a plurality of images matches a predetermined pattern, a phenomenon corresponding to the predetermined pattern is a position in the real space corresponding to the pixel at the position. The image processing method according to claim 1, wherein the image processing method is determined to have occurred. Periodic imaging of the range of the molten glass substrate to which glass raw material is periodically supplied in an electric melting tank that melts the cold top formed on the molten glass substrate by energizing the molten glass. An image processing apparatus using an image obtained by
Changes in gradation values of pixels at a common position in a plurality of images obtained by photographing the range of the ground surface within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the molten glass substrate On the basis of this, it is provided with a state evaluation means for evaluating the state of the cold top at the position in the real space corresponding to the pixel at the position with respect to each position of the pixel corresponding to the range of the ground surface of the molten glass. An image processing apparatus. The image processing apparatus according to claim 9, wherein the state evaluation unit uses a statistic of a gradation value of pixels at a common position in a plurality of images as an evaluation value for evaluating a cold top state. State evaluation means
Maximum value specifying means for specifying the maximum value of gradation values of pixels at a common position in a plurality of images;
The image processing apparatus according to claim 9 or 10, comprising at least one of minimum value specifying means for specifying a minimum value of gradation values of pixels at a common position in a plurality of images. State evaluation means
Maximum value specifying means for specifying the maximum value of gradation values of pixels at a common position in a plurality of images;
Minimum value specifying means for specifying a minimum value of gradation values of pixels at the common position in the plurality of images;
The image processing apparatus according to any one of claims 9 to 11, further comprising difference calculation means for calculating a difference value between a maximum value and a minimum value of the gradation values of the pixels at the common position. According to the state evaluation means, the gradation value at the pixel position is a desired statistic based on the time when the raw material is introduced into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass. The image processing apparatus according to claim 9, wherein an image that can be regarded is specified, and a gradation value at the position of the pixel is extracted from the image. State evaluation means
Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified First extracting means for extracting a gradation value at the position of the pixel from the image;
Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, an image that can be considered that the gradation value at the position of the pixel is the minimum value is specified The image processing apparatus according to claim 9, further comprising at least one of second extraction means for extracting a gradation value at the position of the pixel from the image. State evaluation means
Based on the time when the raw material was put into the position of the real space corresponding to the pixel corresponding to the range of the ground surface of the molten glass, the image whose gradation value at the position of the pixel can be regarded as the maximum value is identified First extracting means for extracting a gradation value at the position of the pixel from the image;
Based on the time, a second extraction unit that identifies an image in which the gradation value at the pixel position can be regarded as a minimum value, and extracts the gradation value at the pixel position from the image;
The difference calculation means for calculating the difference value between the gradation value extracted by the first extraction means and the gradation value extracted by the second extraction means. The image processing apparatus according to one item. State evaluation means
Gradation value extraction means for extracting gradation values of pixels at a common position from a plurality of images and generating time-series data of gradation values;
A phenomenon occurrence determination unit that determines that a phenomenon corresponding to the predetermined pattern occurs at a position in the real space corresponding to the pixel at the position when the time-series data matches the predetermined pattern; An image processing apparatus according to claim 9. A method for controlling an electric melting tank in which a cold top formed on a ground surface of the molten glass is melted by energizing the molten glass, and a glass material is periodically supplied to the ground surface of the molten glass. ,
Changes in gradation values of pixels at a common position in a plurality of images obtained by photographing the range of the ground surface within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the molten glass substrate Based on the position of the cold top at the position in the real space corresponding to the pixel of the position based on each position of the pixel corresponding to the range of the ground surface of the molten glass,
The method for controlling an electrolysis tank, comprising: controlling the electrolysis tank according to an evaluation result of a cold top state. Molten glass is melted by melting the cold top formed on the molten glass substrate by energizing the molten glass in an electric melting tank in which glass raw materials are periodically supplied to the molten glass substrate. A melting step to manufacture,
A clarification step of removing bubbles of the molten glass in a clarification tank;
Forming step of forming molten glass from which bubbles have been removed;
A slow cooling step of slowly cooling the molded molten glass, and
Pixels at common positions in a plurality of images obtained by photographing the range of the ground surface within one cycle or a plurality of cycles of the operation of supplying the glass raw material onto the ground surface of the molten glass in the electric melting tank Based on the change in the gradation value, evaluating the cold top state at the position in the real space corresponding to the pixel at the position is performed for each position of the pixel corresponding to the range of the molten glass background,
The method for producing a glass article, wherein the electrolysis bath is controlled in accordance with an evaluation result of a cold top state.

Images