Flaw identification method and flaw identification system used in the flaw identification method

Abstract

To provide a flaw identification method that, when a damaged conveyance roll is included in a product conveyance line, makes it possible to easily and accurately identify the damaged roll and suppress the operation of conveyance rolls from being stopped, and a flaw identification system that is used for the flaw identification method.SOLUTION: A flaw identification method using a flaw identification system comprises: a determination step for determining whether or not a damaged product is detected; an information acquisition step for acquiring the ID of a damaged product, the result of flaw detection, the ID of a flaw-present image capturing camera, and the ID of a flaw-absent image capturing camera, when it is determined that a damaged product has been detected; and a notification step for notifying the ID of the damaged product and that a damaged conveyance roll is included in conveyance rolls in a specific range. In the information acquisition step, it is known at the time a product has passed through one camera whether the product having passed through the other camera installed immediately after the one camera in a prescribed direction is the damaged product or not.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a flaw identification method for identifying a damaged transport roll, and a flaw identification system used in the flaw identification method. [Background technology]

[0002] A flaw detection method and flaw detection device are known that can detect with high accuracy flaws formed on the surface of a billet in relation to its edge and estimate its position. The flaw detection method includes a marking step in which a mark having a predetermined pattern is applied to the surface of the billet at a location where a flaw has been formed using a colorant having a color in the visible range, an imaging step in which the image capturing device captures color images of the surface of the billet, including the edge, in the visible range while the billet is moved in a moving direction relative to the image capturing device, an edge detection step in which the edge is detected in the image obtained in the imaging step, and a mark detection step in which the mark pattern and the position relative to the edge are detected (see, for example, Patent Document 1).

[0003] Also known is a defect inspection method and apparatus for an object to be inspected that is easy to handle and rarely erroneously identifies non-defective portions as defects. This defect inspection method detects defects such as protrusions, scratches, bubbles, deposits, and stains present on the object to be inspected, and is characterized in that it images a plurality of objects to be inspected using an imaging means while continuously conveying them, extracts defect image data by binarizing the image data of the objects to be inspected output from the imaging means, and when detecting defects from the defect image data, it links individual defect images in the defect image data that are within a predetermined distance from each other to form a single defect area, and determines defect areas with areas larger than a predetermined area as edge images of the object to be inspected (see, for example, Patent Document 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2017-072435 [Patent Document 2] Japanese Patent Application Publication No. 09-072853 Summary of the Invention [Problem to be solved by the invention]

[0005] Both Patent Document 1 and Patent Document 2 disclose that an imaging means (a camera in Patent Document 1 and a CCD line sensor in Patent Document 2) captures an image of the surface of an object (a copper piece in Patent Document 1 and a glass plate in Patent Document 2) flowing on a series of conveying rollers (corresponding to a roller table in Patent Document 1), and that defects (for example, scratches, stains, etc.) are optically detected based on the image of the object's surface. However, when one of the conveying rollers is damaged and the object is damaged by passing over the damage, it can be difficult to identify which conveying roller is damaged.

[0006] For example, in Patent Documents 1 and 2, when it is discovered that an object that has finished flowing on a transport roller (hereinafter referred to as a "transport roll") is damaged, if multiple cameras are installed along the direction in which the transport rolls are lined up, it is possible to identify the image when the object begins to be damaged and the image when the object is not damaged by checking the images taken by all of those cameras based on the time when the damaged object passed through each camera, etc. This allows the camera that generated the image when the object began to be damaged and the camera that generated the image when the object is not damaged to be identified, and it is found that the transport roll lined up between those cameras is damaged. Then, by checking those transport rolls individually, the damaged transport roll can be identified. can.

[0007] However, this method of identifying damaged transport rolls requires a great deal of time and effort. For example, by comparing the time when a damaged object passed through each camera with the images from each camera, it is possible to identify the image when the object began to be damaged, but this requires a great deal of time and effort. As a result, the operation of the transport rolls will be stopped for a long period of time, which may lead to production delays, etc. In addition, there is a risk that maintenance of the transport rolls cannot be performed efficiently.

[0008] The present invention has been made in consideration of the above-mentioned problems, and has as its ultimate objective to provide a scratch identification method and a scratch identification system for use in the scratch identification method that can easily and accurately identify a damaged conveying roll when the damaged conveying roll is included in a product conveying line, and prevent the conveying roll from stopping operation. [Means for solving the problem]

[0009] The present disclosure for solving the above problems is as follows: A method for identifying a damaged conveying roll, which is a conveying roll having a scratch that may cause a scratch on a product when the product passes through the damaged conveying roll, when the damaged conveying roll is included in a conveying line consisting of a plurality of conveying rolls, comprising the steps of: a determination step of capturing an image of the underside of the product using a plurality of cameras installed at intervals along a predetermined direction in which the product is transported on the transport line, and determining whether or not a damaged product, which is the product that has been damaged by the damaged transport roll, has been detected based on the image capturing results and position information of the product; an information acquisition process for acquiring, when it is determined in the determination process that the damaged product has been detected, the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera among the multiple cameras that first captured an image of the damaged product, and the ID of the non-damaged imaging camera installed one behind the damaged imaging camera with reference to the predetermined direction; a notification step of notifying the user of the ID of the damaged product and the fact that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, The information acquisition process includes a damage identification method characterized in that, at the time the product passes one of the multiple cameras, it is known whether the product that passes another camera located one camera behind the one camera with the specified direction as the reference is the damaged product.

[0010] According to the present disclosure, it is possible to easily identify the damaged and undamaged imaging cameras from among multiple cameras, and to easily and accurately identify a specific range of transport rolls that includes damaged transport rolls based on the positional relationship between these cameras and the transport rolls. Therefore, it is possible to prevent the operation of the transport rolls from being stopped, and even if the operation of the transport rolls is stopped, the downtime can be shortened and the cost of finding the damaged transport roll can be reduced. Furthermore, if the specific range of transport rolls can be identified, it is easier to plan maintenance.

[0011] The present disclosure may also provide a scratch identification method further comprising a second determination step of determining whether or not to stop operation of the conveying line based on the scratches on the damaged product notified in the notification step and a predetermined index of the severity of the scratches, and if the operation of the conveying line is to be stopped, notifying the same, and if the operation of the conveying line is not to be stopped, notifying the same that all or part of the plurality of conveying rolls will be replaced after a certain date and time has passed. If the operation of the conveying line is stopped every time it is determined that a damaged product has been detected, product productivity will decrease. By including the cutting process, it is possible to suppress such a decrease in productivity. Furthermore, even if it is determined that a damaged product has been detected, if the damage is not serious, the operation of the conveying line will not be stopped. However, for example, if the damage on the damaged conveying roll that caused the damage to the damaged product gradually increases with age, there is a risk that it may cause serious damage to other products. By replacing all or some of the multiple conveying rolls, it is possible to prevent such serious damage to other products.

[0012] The present disclosure may also provide a flaw identification method further comprising a storage step of storing the image capture results acquired from each of the plurality of cameras for each product ID. This allows, when checking image capture results of products transported in the past, the camera that acquired the image capture results and the product ID displayed in the image capture results to be easily identified, making it easy to check, for example, the occurrence of flaws in products transported in the past.

[0013] The present disclosure may also provide a flaw identification method characterized in that the number of the plurality of cameras is smaller than the number of the plurality of transport rolls. If at least two cameras are installed, it is possible to identify transport rolls within a specific range. By minimizing the number of cameras installed, it is possible to minimize the communication costs of the plurality of cameras.

[0014] The present disclosure may also provide a flaw identification method, characterized in that, when it is determined in the second determination step that the operation of the conveyance line should be stopped, a sample is conveyed on the conveyance rolls in the specific range in the specific direction a specific distance, and then, using the specific direction as a reference, the sample is conveyed in the opposite direction to the specific direction until the leading edge of the sample is positioned in front of the imaging range of the flaw-free imaging camera, and this procedure is repeated while changing the specific distance until the flaw is found. This makes it possible to identify a damaged conveyance roll among the conveyance rolls in the specific range. Furthermore, if the rotation of the conveyance rolls can be controlled using a device such as a control panel, the damaged conveyance roll can be accurately identified even if it is located in a location that is difficult to visually confirm.

[0015] The present disclosure may also provide a flaw identification method characterized in that the sample is transported without slipping on the transport rolls, and the transport rolls all have the same rotation speed, rotation direction, and rotation start and end timings. This eliminates the need to consider idling of the transport rolls. Furthermore, the leading edge of the sample can be accurately calculated.

[0016] The present disclosure may also provide a flaw identification method, in which n = 1, the sample is placed on the transport roll so that the leading edge of the sample is positioned in front of the imaging range of the flaw-free imaging camera relative to the predetermined direction, the sample is transported in the predetermined direction by n times the distance between two adjacent transport rolls, and the sample is again transported in the opposite direction to the predetermined direction so that the leading edge is positioned in front of the imaging range of the flaw-free imaging camera. If no flaw is found, the value of n is incremented by 1. If the value of n is less than or equal to the number of transport rolls in the specific range, the sample is again transported in the predetermined direction by n times the distance between two adjacent transport rolls, and the sample is again transported in the opposite direction to the predetermined direction so that the leading edge is positioned in front of the imaging range of the flaw-free imaging camera. If the value of n is greater than the number of transport rolls in the specific range, an output is output indicating that the damaged transport roll is not included in the transport rolls in the specific range. This improves the accuracy of identifying damaged transport rolls among the transport rolls in the specific range.

[0017] In the present disclosure, a pusher pushes the product onto the conveying line along the longitudinal direction of the plurality of conveying rolls, and the damaged product is detected in the determining step. The method may further include a second information acquisition step of acquiring the position of the scratch on the damaged product obtained from the imaging results and a predetermined reference position from which the pusher pushes the product when it is determined that the scratch has been detected; a calculation step of calculating the width of the product in a direction parallel to the longitudinal direction of the multiple transport rolls as the product is transported on the multiple transport rolls based on the position of the scratch on the damaged product obtained from the imaging results and the reference position; and a second notification step of notifying the user that the damaged transport roll is included in the specific range of transport rolls, as well as the ID and width of the damaged product. Changing the product's width can prevent the product from coming into contact with the scratch even when passing through the damaged transport roll. This allows only products that can avoid coming into contact with the scratch even when passing through the damaged transport roll according to their width to be selected and only those products can be produced without stopping the operation of the transport line.

[0018] Furthermore, the present disclosure may also provide a scratch identification method, characterized in that in the calculation step, the reference position can be changed, and if changed, a new width is calculated based on the position of the scratch on the damaged product obtained from the imaging results and the changed reference position, and in the second notification step, the changed reference position and the new width are notified. Simply changing the width of the product may not prevent the product from coming into contact with the scratch when passing through a damaged transport roll, but changing the reference position may solve such a problem.

[0019] The present disclosure also provides: a conveying line consisting of a plurality of conveying rolls; a plurality of cameras installed at intervals along a predetermined direction in which the products are conveyed on the conveying line; a plurality of sensors that detect the position of the product when the product being conveyed on the conveyor line passes by; A damage identification system for identifying a damaged transport roll, which is a transport roll having a damage that may cause damage to the product when the product passes through the damaged transport roll, when the damaged transport roll is included in the transport line, determining whether or not a damaged product, which is the product that has been damaged by the damaged transport roll, has been detected based on the image capture results of the product acquired from the plurality of cameras and the position information of the product acquired from the plurality of sensors; a control unit that, when it is determined that the damaged product has been detected, acquires the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera among the multiple cameras that first captured an image of the damaged product, and the ID of the non-damaged imaging camera installed one behind the damaged imaging camera with reference to the predetermined direction; and a notification unit that notifies the user of the ID of the damaged product and that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, The system may include a damage identification system characterized in that, at the time the product passes through one of the multiple cameras, the control unit has already acquired information as to whether the product that has passed through another camera located just before the one camera based on the specified direction is the damaged product.

[0020] According to the present disclosure, it is possible to easily identify the damaged and undamaged imaging cameras from among a plurality of cameras, and to easily and accurately identify the transport rolls in a specific range from the positional relationship between these cameras and the transport rolls. As a result, it is possible to prevent the transport rolls from stopping operation, and even if the transport rolls do stop operation, the stop time can be shortened and the cost of finding the damaged transport rolls can be reduced. .

[0021] The present disclosure may also provide a flaw identification system, wherein the control unit determines whether to stop operation of the conveying line based on the flaw in the damaged product notified by the notification unit and a predetermined indicator of the flaw severity, and the notification unit notifies the user if the operation of the conveying line is to be stopped, and if the operation of the conveying line is not to be stopped, notifies the user that all or some of the plurality of conveying rolls will be replaced after a certain date and time has elapsed. By including a control unit in the flaw identification system of the present disclosure, it is possible to suppress a decrease in productivity due to excessive stoppage of the conveying line. Furthermore, even if the flaw in the damaged product is not severe, by replacing all or some of the plurality of conveying rolls, it is possible to prevent, for example, the flaw in the damaged conveying roll that caused the flaw in the damaged product from gradually increasing due to aging.

[0022] The present disclosure may also provide a flaw identification system further comprising a storage unit that stores the image capture results acquired from each of the plurality of cameras for each product ID. This allows easy identification of the camera that captured the image capture results and the product ID displayed in the image capture results when checking image capture results of products previously transported.

[0023] In addition, the present disclosure may provide a flaw identification system characterized in that the number of the plurality of cameras is smaller than the number of the plurality of transport rolls. By limiting the number of cameras to a necessary minimum, it is possible to minimize communication costs from the plurality of cameras to the control unit.

[0024] The present disclosure also provides a flaw identification system, further comprising a pusher that pushes the product onto the conveyor line along the longitudinal direction of the multiple conveyor rolls. When the control unit determines that the damaged product has been detected, the control unit acquires the location of the flaw on the damaged product obtained from the image capture and a predetermined reference position where the pusher pushes the product. Then, based on the location of the flaw on the damaged product obtained from the image capture and the reference position, the control unit calculates the width of the product in a direction parallel to the longitudinal direction of the multiple conveyor rolls as the product is conveyed on the multiple conveyor rolls. The notification unit notifies the system that the damaged conveyor roll is included in the specific range of conveyor rolls, as well as the ID and width of the damaged product. Changing the product width can prevent the product from coming into contact with the flaw even when passing through the damaged conveyor roll. This allows the system to select only products that will not come into contact with the flaw even when passing through the damaged conveyor roll according to their width, without stopping the conveyor line, and to produce only those products.

[0025] In addition, in the present disclosure, the reference position may be changeable, and when changed, the control unit calculates a new width based on the position of the scratch on the damaged product obtained from the imaging results and the changed reference position, and the notification unit notifies the changed reference position and the new width. Simply changing the width of the product may not prevent the product from coming into contact with the scratch when passing through a damaged transport roll, but changing the reference position may solve this problem.

[0026] The above means for solving the problems can be used in combination with each other whenever possible. [Effects of the Invention]

[0027] According to the present invention, when a damaged transport roll is included in a product transport line, the damaged transport roll can be easily removed. It is possible to easily and accurately identify the transport roll and prevent the operation of the transport roll from being stopped, which makes it possible to efficiently perform maintenance of the transport roll, for example. [Brief explanation of the drawings]

[0028] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of a flaw identification system according to an embodiment. [Figure 2] FIG. 2 is a flowchart showing the procedure of a flaw identification method using the flaw identification system according to the embodiment. [Figure 3] 3A to 3C are diagrams illustrating the flow of a method for detecting a damaged transport roll from transport rolls within a specific range in a flaw detection system in which the distance between two adjacent transport rolls is greater than the circumference of the transport roll, in relation to a flaw detection method according to an embodiment. Fig. 3A is a first explanatory diagram. Fig. 3B is a second explanatory diagram. Fig. 3C is a third explanatory diagram. [Figure 4] Figures 4A and 4B are diagrams illustrating a continuation of the flow of Figure 3C. Figure 4A is a fourth explanatory diagram. Figure 4B is a fifth explanatory diagram. [Figure 5] 5A to 5C are diagrams illustrating the flow of a method for identifying a damaged transport roll from transport rolls within a specific range in a flaw identification system in which the distance between two adjacent transport rolls is equal to or less than the circumference of the transport roll, in relation to a flaw identification method according to an embodiment. Fig. 5A is a first explanatory diagram. Fig. 5B is a second explanatory diagram. Fig. 5C is a third explanatory diagram. [Figure 6] Figures 6A to 6C are diagrams illustrating a continuation of the flow of Figure 5C. Figure 6A is a fourth explanatory diagram. Figure 6B is a fifth explanatory diagram. Figure 6C is a sixth explanatory diagram. [Figure 7] Figures 7A to 7C are diagrams illustrating a continuation of the flow of Figure 6C. Figure 7A is a seventh explanatory diagram. Figure 7B is an eighth explanatory diagram. Figure 7C is a ninth explanatory diagram. [Figure 8] FIG. 8 is a second flowchart showing a procedure that generalizes and supplements the flows shown in FIGS. 3A to 4B and the flows shown in FIGS. 5A to 7C. [Figure 9] FIG. 9 is a diagram illustrating a first method for preventing a product passing through a damaged transport roll from coming into contact with a scratch based on the position of the scratch on the damaged transport roll, in relation to the scratch identification method according to the embodiment. [Figure 10] FIG. 10 is a diagram illustrating a second method for preventing a product passing through a damaged transport roll from coming into contact with a scratch based on the position of the scratch on the damaged transport roll, in relation to the scratch identification method according to the embodiment. [Figure 11] FIG. 11 is a third flowchart showing a procedure that generalizes the flow shown in FIGS. DETAILED DESCRIPTION OF THE INVENTION

[0029] [Application example] An outline of an application example of the present invention will be explained below with reference to some of the drawings. The present disclosure can be applied to a flaw identification system 1 as shown in Fig. 1. Furthermore, by using the flaw identification system 1, the present disclosure can be applied to the process as shown in the flowchart of Fig. 2.

[0030] FIG. 1 is a schematic diagram showing an example of the configuration of a flaw identification system 1 to which the present disclosure is applicable. The flaw identification system 1 in this application example is broadly composed of a conveying line 10 consisting of multiple conveying rolls 100 (only one conveying roll 100 is labeled with a reference number in FIG. 1 for simplicity; the same applies to the following figures), multiple cameras 11a and 11b, multiple sensors 12a-12d, a controller 13, a monitor 14, and a storage 15. The flaw identification system 1 in this application example is generally used in production sites such as factories, where, as part of a mass-produced product production process, plate-shaped products 2a-2c, such as printed circuit boards, are conveyed in a predetermined direction (the direction indicated by the arrow in FIG. 1; the same applies to the following figures) on the conveying line 10 with a fixed spacing between the front and back of each product. It is assumed that each of the products 2a-2c has a specific surface, either the front or back, in contact with the conveying roll 100. Note that, hereinafter, the terms "front" and "rear" will be used to indicate the positional relationship. When the expression "ro" is used, this predetermined direction is used as the reference. For example, in FIG. 1, on the conveyor line 10, the product 2a is located at the front and the product 2c is located at the rear.

[0031] The conveying line 10 may include a conveying roll 100 having scratches that may cause damage to the products 2a-2c when the products 2a-2c pass through it. Hereinafter, this conveying roll 100 is referred to as a damaged conveying roll 100a. Damage that may cause damage to the products 2a-2c includes, for example, depressions on the surface of the conveying roll 100 due to aging, or protrusions formed by hardened foreign matter adhering to the surface of the conveying roll 100. When the products 2a-2c come into contact with such scratches, scratches such as cuts or scrapes are caused on the surface in contact with the conveying roll 100 (hereinafter, the "surface of the products 2a-2c in contact with the conveying roll 100" will be simply referred to as the "contact surface of the products 2a-2c"). In this case, scratches may be caused at regular intervals on the contact surface of the products 2a-2c due to the rotation of the conveying roll 100. Hereinafter, the products 2a-2c damaged by the damaged conveying roll 100a will be referred to as the damaged products 2a-2c. In FIG. 1, the products 2a and 2b are exemplified as damaged products 2a and 2b.

[0032] The cameras 11a and 11b are installed at intervals below the conveyor line 10 along a predetermined direction. The lenses of the cameras 11a and 11b face the conveyor line 10, and when the products 2a-2c pass by, the cameras 11a and 11b capture images of the contact surfaces of the products 2a-2c through the gap between the two adjacent conveyor rolls 100. In FIG. 1, the damaged conveyor roll 100a is located behind the imaging range of the camera 11a and in front of the imaging range of the camera 11b. Therefore, the camera 11a can capture images of the damaged products 2a and 2b, but the camera 11b cannot capture images of the damaged products 2a and 2b. In this application example, although a video is exemplified as the imaging result, continuous still images may also be used.

[0033] The sensors 12a-12d detect the positions of the products 2a-2c as they pass by, for example, by reading the barcodes attached to the products 2a-2c. Although the sensors 12a-12d are installed below the conveyor line 10 in FIG. 1, they may also be installed above the conveyor line 10.

[0034] The controller 13 determines whether damaged products 2a and 2b have been detected based on the images of the products 2a-2c acquired from the cameras 11a and 11b and the position information of the products 2a-2c acquired from the sensors 12a-12d. For example, if the time when the sensors 12a and 12b detected the position of the product 2b (the time when the product 2b passed the sensors 12a and 12b) is calculated, and the distance from the sensors 12a and 12b to the camera 11a and the conveying speed of the conveying line 10 (the rotational speed of the conveying roll 100) are known, the time when the camera 11a captured the image of the product 2b (the time when the product 2b passed through the imaging range of the camera 11a) can be calculated based on these information. By calculating this imaging time, the product 2b can be identified from among the products 2a-2c in the image captured by the camera 11a. If it is further determined that a scratch has been detected on the contact surface of the product 2b, the product 2b can be identified as a damaged product 2b. That is, by acquiring images and position information of the products 2a-2c from both the cameras 11a, 11b and the sensors 12a-12d, the controller 13 can identify each of the products 2a-2c in the images, eliminating the need to compare the images with the times at which the sensors 12a-12d detected the positions of the products 2a-2c. This makes it easier and more accurate to detect damaged products 2a, 2b than when only images of the products 2a-2c are acquired from the cameras 11a, 11b. Furthermore, scratches can be detected from the images at all times. Note that, as a method for detecting damaged products 2a, 2b, for example, a method of detecting scratches by optically analyzing the images of the damaged products 2a, 2b may also be used. Furthermore, the controller 13 is equipped with an IoT Gateway (not shown) internally, and can use the drive recorder function of the IoT Gateway to detect scratches from the images. The products 2a and 2b may be detected from the controller 13. Here, the controller 13 corresponds to the control section in this disclosure.

[0035] Furthermore, at the time when camera 11a captures the images of products 2a-2c, controller 13 has already acquired information on whether or not the products 2a-2c captured by camera 11b are damaged products 2a-2c (in FIG. 1, the products 2a-2c captured by camera 11b are not, as a result, damaged products 2a-2c). Thus, if controller 13 acquires information that the products 2a-2c captured by camera 11a are damaged products 2a-2c, it can be determined that the damaged transport roll 100a is included among the transport rolls 100 arranged between the imaging ranges of cameras 11a and 11b (hereinafter, these transport rolls 100 are referred to as "transport rolls 100 in a specific range"). Here, camera 11a corresponds to the "damage-capturing camera" in this disclosure, and camera 11b corresponds to the "non-damage-capturing camera" in this disclosure.

[0036] When the controller 13 determines that a damaged product 2b has been detected (for simplicity, only the damaged product 2b is illustrated here, excluding the damaged product 2a), it acquires the following four pieces of information from the cameras 11a and 11b and the sensors 12a-12d. The first piece of information is the ID of the damaged product 2b. In this application example, the product name is illustrated as the ID, but any information that can identify the damaged product 2b may be used. The second piece of information is the result of the damage detection based on all images of the damaged product 2b (including the product 2b before the damage was detected). Figure 1 specifically shows the result that a damage was detected in the image of the product 2b acquired from the camera 11a, and the result that no damage was detected in the image of the product 2b acquired from the camera 11b. The third piece of information is the ID of the camera 11a. The fourth piece of information is the ID of the camera 11b.

[0037] The monitor 14 is mutually connected to the controller 13 via a wireless or wired communication line. The monitor 14 further acquires the above four pieces of information acquired by the controller 13, and displays and notifies the user of the name of the damaged product 2b and the fact that the damaged transport roll 100a is included in the transport rolls 100 in the specific range. The monitor 14 is preferably placed in a location that is easy for a user or manager of the scratch identification system 1 to check, and may be placed in a room separate from the controller 13. Here, the monitor 14 corresponds to the notification unit in this disclosure.

[0038] FIG. 2 is a flowchart showing the steps of a flaw identification method using a flaw identification system 1 to which the present disclosure can be applied. In this flowchart, cameras 11a and 11b first capture images of products 2a-2c, and controller 13 determines whether damaged products 2a-2c have been detected based on the images of products 2a-2c and the positional information of products 2a-2c acquired from sensors 12a-12d (S101). At this time, when camera 11a captures products 2a-2c, controller 13 has already acquired information that the products 2a-2c captured by camera 11b are not damaged products 2a-2c. In other words, when any one camera captures products 2a-2c, controller 13 has already acquired information as to whether the products 2a-2c captured by another camera installed one behind the camera are damaged products 2a-2c. Here, S101 corresponds to the determination step in the present disclosure. If the controller 13 determines that the damaged products 2a-2c have been detected (S101: yes), the controller 13 acquires the above-mentioned four pieces of information (for the damaged product 2b, the damaged product 2a-2c is broadly defined as the damaged product 2a-2c) from the cameras 11a and 11b and the sensors 12a-12d (S102). Here, S102 corresponds to the information acquisition step in this disclosure. The controller 13 identifies the transport rolls 100 in the specific range based on the ID of the camera 11a and the ID of the camera 11b (S103). Note that S103 may be considered as part of the information acquisition step S102. The monitor 14 further acquires the above-mentioned four pieces of information acquired by the controller 13, and notifies by displaying the product names of the damaged products 2a-2c and the fact that the damaged transport roll 100a is included in the transport rolls 100 in the specific range (S104). Here, S104 corresponds to the notification step in this disclosure. The steps after the notification step S104 The process is described in the examples below.

[0039] [Example] The flaw identification method and flaw identification system 1 according to the embodiments of the present invention will be described in more detail below with reference to the drawings (including the drawings that have been explained in the above application examples). Note that the flaw identification method and flaw identification system 1 according to the present invention are not intended to be limited to the following configurations.

[0040] <System configuration> Returning now to the explanation of Fig. 1, the flaw identification system 1 according to this embodiment has the same configuration as the flaw identification system 1 described in the application example, and therefore detailed explanation of the contents described in the application example will be omitted. Furthermore, in this specification, the same components will be described using the same reference numerals.

[0041] The controller 13 determines whether to stop the operation of the conveying line 10 based on the damage on the damaged products 2a-2c notified by the monitor 14 and a predetermined indicator of the severity of the damage. For example, if the type of damage is a cut or void, the damage is considered to be relatively serious and an indicator for stopping the operation of the conveying line 10 is predetermined. If the type of damage is a scratch, the damage is considered to be relatively less serious and an indicator for not stopping the operation of the conveying line 10 is predetermined. Furthermore, depending on the severity of the damage on the damaged products 2a-2c, various problems may occur in the damaged conveying roll 100a. For example, if the damage is relatively serious, the damaged conveying roll 100a may suddenly crack, whereas if the damage is relatively less serious, the damage on the damaged conveying roll 100a may gradually become larger due to aging.

[0042] If the operation of the conveying line 10 is to be stopped, the monitor 14 displays a message to that effect. If the operation of the conveying line 10 is not to be stopped, the monitor 14 displays a message to the effect that all or part of the conveying roll 100 will be replaced after a certain date and time has passed, for example, on the next predetermined scheduled date for maintenance of the conveying line 10. This makes it easier for a user or manager of the flaw identification system 1 in this embodiment to take appropriate measures for the damaged conveying roll 100a. The monitor 14 may also have a function to notify by sound.

[0043] Furthermore, the storage 15 stores images acquired by the cameras 11a and 11b for each of the products 2a-2c. This makes it possible to check for defects that occurred in products that were previously transported on the transport line 10. Furthermore, the number of cameras 11a and 11b installed may be increased if the number is small compared to the number of transport rolls 100. In this embodiment, two cameras 11a and 11b are sufficient to identify a specific range of transport rolls 100, and by keeping the number of cameras to a minimum, the communication cost from the cameras 11a and 11b to the controller 13 can be reduced as much as possible. Here, the storage 15 corresponds to the storage unit in this disclosure.

[0044] <Flowchart> Returning now to the explanation of FIG. 2, as described above, the controller 13 determines whether or not to stop the operation of the conveying line 10 based on the damage on the damaged products 2a-2c notified by the monitor 14 and a predetermined index of the severity of the damage (S105). Here, S105 corresponds to the second determination step in the present disclosure. If the operation of the conveying line 10 is to be stopped (S105: yes), the monitor 14 notifies the user by displaying a message to that effect (S106). If the operation of the conveying line 10 is not to be stopped (S105: no), the monitor 14 notifies the user by displaying a message to the effect that all or part of the conveying rolls 100 will be replaced after a certain date and time has elapsed (S107). If the controller 13 does not determine that damaged products 2a-2c have been detected (S101: no), or if the controller 13 acquires the above-described four pieces of information in the information acquisition step S102, After acquiring the images, the storage 15 stores the images acquired by the cameras 11a and 11b for each of the products 2a to 2c (S108). Here, S108 corresponds to the storing step in this disclosure.

[0045] <How to identify scratches> 3A to 4B, a flow of a method for finding a damaged transport roll 100a from among the transport rolls 100 in a specific range after the monitor 14 displays and notifies that the operation of the transport line 10 will be stopped in S106 shown in Fig. 2. In the flaw identification system 1 shown in Fig. 3A to 4B, the distance between two adjacent transport rolls 100 (which can be considered as the distance x1 between the transport rolls 100 in the specific range and the damaged transport roll 100a as shown in Fig. 3A; hereinafter, simply referred to as "distance x1") is greater than the circumference of the transport roll 100.

[0046] 3A to 4B, a sample 3 is transported on a specific range of transport rolls 100 to find a damaged transport roll 100a among the transport rolls 100 in the specific range. The sample 3 is a dummy substrate used in place of the products 2a-2c and has approximately the same material and thickness as the products 2a-2c. In FIGS. 3A to 4B, the sample 3 is transported a certain distance in a specific direction on the transport rolls 100 in the specific range, and then transported in the opposite direction from the specific direction until the leading edge of the sample 3 is positioned in front of the imaging range of the camera 11b. This procedure is repeated to check whether the surface of the sample 3 is damaged. It is assumed that the procedures shown in FIGS. 3A to 4B are all performed offline, the sample 3 is transported without slipping on the transport rolls 100 (i.e., idling of the transport rolls 100 is not taken into account), the leading edge position of the sample 3 is calculable, and all of the transport rolls 100 have the same rotation speed, rotation direction, and rotation start and end timings. As a method for transporting the sample 3, the sample 3 may be transported by being held by hand, or by using a device such as an operation panel to operate the rotation of the transport rolls 100 without a person approaching the transport rolls 100. For simplicity, in Figures 3A to 4B, the number of transport rolls 100 in the specific range is set to three, including the damaged transport roll 100a, and hereinafter, these transport rolls 100 in the specific range will be referred to as the "first transport roll 100," the "second transport roll 100a," and the "third transport roll 100," in order from the one closest to the imaging range of the camera 11b.

[0047] The method for finding a damaged transport roll 100a among transport rolls 100 in a specific range is as follows: First, place the sample 3 on the transport roll 100 so that its leading edge is positioned in front of the imaging range of the camera 11b (FIG. 3A). Next, transport the sample 3 a distance x1 in a predetermined direction (FIG. 3B). At this time, the portion of the sample 3 indicated by the thin, thick line corresponds to the circumference of the transport roll 100. Even if the first transport roll 100 rotates once after the sample 3 contacts the first transport roll 100, the sample 3 does not contact the second transport roll 100a. In other words, if the distance x1 is greater than the circumference of the transport roll 100, the rear transport roll 100 rotates more than once between the time the sample 3 contacts the rear transport roll 100 and the time the sample 3 contacts the front transport roll 100, and therefore the sample 3 is in contact with the entire surface of the rear transport roll 100.

[0048] Next, the sample 3 is transported in the opposite direction to the predetermined direction so that the front end of the sample 3 is again positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (Fig. 3C). If the controller 13 does not detect any scratches, it is understood that there are no scratches on the first transport roll 100. Next, the sample 3 is transported in the predetermined direction by twice the distance x1 (Fig. 4A). At this time, considering it in the same way as Fig. 3B, the sample 3 is not in contact with the third transport roll 100, so if there are scratches on the surface of the sample 3, it can be said that they are caused by the second transport roll 100a. Finally, the front end of the sample 3 is again positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (Fig. 3C). If the controller 13 does not detect any scratches, it is understood that there are no scratches on the first transport roll 100. Next, the sample 3 is transported in the predetermined direction by twice the distance x1 (Fig. 4A). At this time, as in Fig. 3B, the sample 3 is not in contact with the third transport roll 100, so if there are scratches on the surface of the sample 3, it can be said that they are caused by the second transport roll 100a. The sample 3 is transported in the opposite direction to the predetermined direction so that it is positioned in front of the imaging range of b, and the camera 11b captures an image of the surface of the sample 3 (FIG. 4B). If the controller 13 detects a scratch, it is determined that the second transport roll 100a is scratched.

[0049] 5A to 7C, the flow of a method for finding a damaged transport roll 100a among the transport rolls 100 in a specific range after the monitor 14 displays and notifies that the operation of the transport line 10 will be stopped in S106 shown in FIG. 2 is described below in the same manner as the flow shown in FIGS. 3A to 4B. The difference from the flow shown in FIGS. 3A to 4B is that in the flaw identification system 1 shown in FIGS. 5A to 7C, the distance between two adjacent transport rolls 100 (including the distance between the transport rolls 100 in the specific range and the damaged transport roll 100a; hereinafter, simply referred to as "distance x2") is equal to or less than the circumference of the transport roll 100. Furthermore, in FIGS. 5A to 7C, the number of transport rolls 100 in the specific range is set to five, including the damaged transport roll 100a. Hereinafter, the transport rolls 100 in the specific range, including the damaged transport roll 100a, will be referred to as the "first transport roll 100" to the "fourth transport roll 100" in order of their proximity to the imaging range of the camera 11b. The above-mentioned use of Sample 3 and the premise are the same as those in the flow shown in FIGS. 3A to 4B.

[0050] The method for finding a damaged transport roll 100a among transport rolls 100 within a specific range is as follows: First, place the sample 3 on the transport roll 100 so that its leading edge is positioned in front of the imaging range of the camera 11b (FIG. 5A). Next, transport the sample 3 in a predetermined direction by a distance x2 (FIG. 5B). At this time, the portion of the sample 3 indicated by the thin, thick line is shorter than the circumference of the transport roll 100, and the number of rotations of the first transport roll 100 from when the sample 3 contacts the first transport roll 100 until when it contacts the second transport roll 100 is less than one. In other words, if the distance x2 is equal to or shorter than the circumference of the transport roll 100, the number of rotations of the rear transport roll 100 from when the sample 3 contacts the rear transport roll 100 of the two adjacent transport rolls 100 until when it contacts the front transport roll 100 is less than one, and therefore it can be seen that the sample 3 is in contact with only a portion of the surface of the rear transport roll 100. Furthermore, if the distance x2 is 1 / a times the circumference of the transport roll 100 (where a is a natural number), the sample 3 in contact with the transport roll 100 will come into contact with the transport roll 100 that is located a roll ahead of the transport roll 100 after the transport roll 100 makes one rotation. For example, if the distance x2 is 1 / 2 times the circumference of the first transport roll 100, the sample 3 in contact with the first transport roll 100 will come into contact with the third transport roll 100a after the first transport roll 100 makes one rotation.

[0051] Next, the sample 3 is transported in the opposite direction to the predetermined direction so that the leading edge of the sample 3 is again positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (FIG. 5C). However, at the time of FIG. 5B, the first transport roll 100 has not yet rotated once, so the sample 3 has not yet come into contact with the entire surface of the first transport roll 100. Therefore, even if the controller 13 does not detect any scratches at the time of FIG. 5C, it cannot be concluded that the first transport roll 100 is free of scratches. Next, the sample 3 is transported in the predetermined direction by twice the distance x2 (FIG. 6A), and then transported in the opposite direction to the predetermined direction so that the leading edge of the sample 3 is again positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (FIG. 6B). If the controller 13 does not detect any scratches at the time of FIG. 6B, it can be determined that the first transport roll 100 is free of scratches. However, considering the same as FIG. 5C, it cannot be concluded that the second transport roll 100 is free of scratches.

[0052] Next, similarly, the sample 3 is transported in the predetermined direction by three times the distance x2 (FIG. 6C), and then the sample 3 is transported in the opposite direction to the predetermined direction so that the front end of the sample 3 is positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (FIG. 7A). If the controller 13 does not detect any scratches at this point, it can be concluded that the second transport roll 100 is not scratched. However, considering the same as in FIG. 5C, it cannot be said with certainty that the third transport roll 100a is not scratched. Here, it should be noted that at the point in time shown in FIG. 6C, the sample 3 is in contact with the third transport roll 100a and the third transport roll 100 has rotated to a certain extent, but the scratches on the third transport roll 100a are not in contact with the sample 3, so the scratches are not detected by the controller 13 at the point in time shown in FIG. 7A. In other words, it should be noted that even if the sample 3 comes into contact with the damaged transport roll 100a, the scratches may not be detected by the controller 13 unless the damaged transport roll 100a rotates at least once in that state.

[0053] Next, similarly, the sample 3 is transported in the predetermined direction by four times the distance x2 (FIG. 7B), and then transported in the opposite direction so that the front end of the sample 3 is positioned in front of the imaging range of the camera 11b, and the camera 11b captures an image of the surface of the sample 3 (FIG. 7C). If the controller 13 detects a scratch, it is determined that the scratch is on the third transport roll 100a or the fourth transport roll 100. After that, these two transport rolls 100, 100a can be compared visually, or the fourth transport roll 100 can be rotated once while the sample 3 is in contact with it, and if no new scratches are found on the sample 3, it can be determined that the scratch is on the third transport roll 100a.

[0054] As described above, by transporting the sample 3 on the transport rolls 100 in a specific range, a damaged transport roll 100a can be accurately found among the transport rolls 100 in the specific range. When the distance between two adjacent transport rolls 100 (hereinafter referred to as the "transport roll distance") is equal to or less than the circumference of the transport roll 100 (as shown in FIGS. 5A to 7C), the flow of the method for finding the damaged transport roll 100a is somewhat more complicated than when the transport roll distance is greater than the circumference of the transport roll 100 (as shown in FIGS. 3A to 4B). However, even in the former case (as shown in FIGS. 5A to 7C), the damaged transport roll 100a can be accurately found. Furthermore, if the rotation of the transport roll 100 can be controlled using a device such as a control panel as described above, the damaged transport roll 100a can be accurately found even if it is in a place where it is difficult to visually check (for example, a high place or a place where it is difficult for people to enter due to obstacles, etc.).

[0055] <Flowchart> FIG. 8 is a second flowchart showing a generalized and supplementary procedure of the procedures shown in FIGS. 3A to 4B and 5A to 7C. The procedure shown in FIG. 8 starts from S106 shown in FIG. 2, in which the transport line 10 is stopped, a notification of this fact is sent, and the procedure is executed offline. In this flowchart, first, the upper limit of n is set to the number of transport rolls 100 in a specific range, that is, n=1 (S109). Next, the sample 3 is placed on the transport roll 100 so that the leading edge of the sample 3 is positioned in front of the imaging range of the camera 11b (S110). Next, the sample 3 is transported in a predetermined direction by n times the distance between the transport rolls (currently, the distance between the transport rolls) (S111). Next, the sample 3 is transported again in the opposite direction to the predetermined direction so that the leading edge of the sample 3 is positioned in front of the imaging range of the camera 11b (S112), and the camera 11b captures an image of the surface of the sample 3 (S113). If the controller 13 detects a scratch (S113: yes), it notifies the system 13 that a damaged transport roll 100a has been found (S114). Note that, as described above, if the inter-transport roll distance is equal to or less than the circumference of the transport roll 100 (as shown in FIGS. 5A to 7C), the notified damaged transport roll 100a may not actually be scratched, and the transport roll 100 immediately preceding the notified damaged transport roll 100a may be the actual damaged transport roll 100a. Furthermore, since the series of procedures is performed offline, in S114, it may be other components of the scratch identification system 1, rather than the monitor 14, that notify the system 1 of the discovery of the damaged transport roll 100a. If the damaged transport roll 100a is found, the series of procedures ends.

[0056] If the controller 13 does not detect a scratch (S113: no), the value of n is incremented by 1 (S115), and the controller 13 or another component of the scratch identification system 1 determines whether the newly obtained value of n (currently n=2) is greater than the upper limit, i.e., the number of transport rolls 100 in the specific range (S116). If the value of n is equal to or less than the upper limit (S116: no), the process returns to S111. The process from S111 to S116 is repeated until a scratch is detected by the controller 13 in S113. If the value of n is greater than the upper limit (S116: yes), an error is output (S117), and the process ends. Here, the error refers to an error indicating that the damaged transport roll 100a was not included in the transport rolls 100 in the specific range. In other words, if such an error is output, it is possible that one of the components in the scratch identification system 1 has mistakenly determined that a damaged conveying roll 100a is included in the conveying rolls 100 in a specific range, and measures must be taken, such as maintaining the component in question or replacing it with a new component.

[0057] <How to identify scratches> FIG. 9 is a diagram illustrating a first method for preventing a product 2d passing through a damaged transport roll 100a from coming into contact with a scratch based on the location of the scratch on the damaged transport roll 100a, in relation to a scratch identification method according to an embodiment. In FIG. 9, the scratch identification system 1 may further include a second transport line 10a installed substantially parallel to the transport line 10, and a pusher 4 that pushes the product 2d transported on the second transport line 10a onto the transport line 10 along the longitudinal direction of the transport roll 100. FIG. 9 also shows a schematic top view of the scratch identification system 1. The same scratch identification system 1 as in FIG. 1 is used in FIG.

[0058] In FIG. 1, if the controller 13 determines that damaged products 2a, 2b have been detected in the images of the products 2a, 2b acquired from the camera 11a, the controller 13 acquires the product names of the damaged products 2a, 2b and the locations of the scratches on the damaged products 2a, 2b in FIG. 9. The position on the conveyor line 10 from which the pusher 4 pushes out the products 2d is predetermined. For example, as shown in FIG. 9, the pusher 4 pushes out the products 2d so that the longitudinal direction of the product 2d closer to the second conveyor line 10a is aligned with the position of a two-dot chain line X perpendicular to the longitudinal direction of the conveyor roll 100 (hereinafter referred to as the "reference position X"). The pusher 4 may be designed to automatically change the reference position X depending on the product name. The controller 13 also acquires this reference position X.

[0059] Once the damaged products 2a, 2b are detected, the damaged transport roll 100a can be found using the procedure shown in FIG. 8, and the location of the scratch on the damaged transport roll 100a can also be found. Alternatively, the approximate location of the scratch on the damaged transport roll 100a can be estimated before the damaged transport roll 100a is found, taking into account the location of the scratch on the damaged products 2a, 2b and the orientation of the damaged products 2a, 2b during transport. Let D be the difference in the distance along the longitudinal direction of the transport roll 100 between the scratch location on the damaged transport roll 100a and the reference position X. If the width W of the product 2d, which is parallel to the longitudinal direction of the transport roll 100 during transport, is equal to or less than the distance D, the product 2d can be prevented from coming into contact with the scratch even when passing through the damaged transport roll 100a. The controller 13 calculates the width W and determines that the product 2d is producible if the width W is equal to or less than the distance D, and determines that the product 2d is unproducible if the width W is greater than the distance D. Furthermore, as long as it is within the range of the predetermined standard values, the design value of the product 2d may be changed so that the width W is equal to or less than the distance D. The monitor (not shown in FIG. 9) notifies the user that the transport rolls 100 in the specific range include a damaged transport roll 100a, and if the product 2d is damaged, displays the name of the product and the width W that can prevent contact with the damage.

[0060] FIG. 10 illustrates a second method for preventing a product 2d passing through a damaged transport roll 100a from coming into contact with a scratch based on the location of the scratch on the damaged transport roll 100a, as part of the scratch identification method according to the embodiment. The width W′, which is changed from the width W shown in FIG. 9, is equal to or less than the distance D. Therefore, the product 2d will not come into contact with the scratch location shown in FIG. 9 (shown as a white circle in FIG. 10; hereinafter, this location will be referred to as the “original scratch location”) even when transported on the transport line 10. However, as shown in FIG. 10, if the scratch on the damaged transport roll 100a is located below the original scratch location in the orientation shown in FIG. 10 (hereinafter, the lower side in the orientation shown in FIG. 10 will be referred to as the “front side” and the upper side as the “rear side”) and overlaps with the reference position X (or is located slightly rearward of the reference position X), the product 2d will come into contact with the scratch when transported on the transport line 10, even if the width W′ is equal to or less than the distance D. In other words, the closer the scratch is to the front, the higher the risk of the product 2d coming into contact with the scratch, even if the width is reduced.

[0061] In such a case, by changing the position at which the product 2d is pushed out by the pusher 4 from the reference position X to the reference position X', the product 2d is pushed further back than the position of the scratch, and even if the product 2d passes through the damaged transport roll 100a, it is possible to prevent the product 2d from coming into contact with the scratch. Furthermore, after changing to the reference position X', the width W' may be increased as long as the product 2d does not come into contact with the scratch even if it passes through the damaged transport roll 100a. Basically, if the scratch is located on the back side, the reference position should be set on the front side, and if the scratch is located on the front side, the reference position should be set on the back side.

[0062] When the reference position is changed, the controller 13 calculates a new width based on the changed reference position that will prevent contact with the damaged transport roll 100a even when the transport roll passes through the damaged roll 100a. The monitor 14 displays and notifies the user of the changed reference position and the new width.

[0063] <Flowchart> FIG. 11 is a third flowchart showing a generalized procedure of the flows shown in FIGS. 9 and 10. The flow shown in FIGS. 9 and 10 is the flow after the controller 13 determines that damaged products 2a and 2b have been detected in the determination step S101 shown in FIG. 2 (the flow after S101: yes). In this flowchart, the controller 13 first acquires the position of the scratch on the damaged product and the reference position from the image of the damaged product (which may be at least one of the damaged products 2a-2c) acquired from the camera 11a (S118). Here, S118 corresponds to the second information acquisition step in this disclosure. Next, the controller 13 discovers or estimates the position of the scratch on the damaged transport roll 100a from the position of the scratch on the damaged product. Based on the position of the scratch on the damaged transport roll 100a and the reference position, the controller 13 calculates the width of the product 2d that will prevent the product 2d from coming into contact with the scratch even when passing through the damaged transport roll 100a (S119). Here, S119 corresponds to the calculation step in this disclosure. If the reference position is changed, the controller 13 calculates a new width based on the changed reference position that will prevent contact with the damaged transport roll 100a even when passing through it. Next, the monitor 14 notifies the user by displaying a message indicating that the damaged transport roll 100a is included in the specific range of transport rolls 100, and if a product 2d is damaged, the name of the product and the width that will prevent contact with the scratch (S120). Here, S120 corresponds to the second notification step in this disclosure. If the reference position is changed, the monitor 14 notifies the user by displaying the changed reference position and the new width obtained in the calculation step S119.

[0064] <Appendix 1> A method for identifying a damaged conveying roll (100a), which is a conveying roll having a damage that may cause damage to a product (2a-2c) when the product (2a-2c) passes through the damaged conveying roll (100a), when the damaged conveying roll is included in a conveying line (10) consisting of a plurality of conveying rolls (100), comprising the steps of: The conveyor line is provided with gaps along a predetermined direction in which the product is conveyed. a judgment step (S101) of capturing an image of the underside of the product using a plurality of cameras (11a, 11b) placed on the conveyor belt, and judging whether or not a damaged product (2a-2c) that has been damaged by the damaged conveying roll has been detected based on the image capturing result and position information of the product; an information acquisition step (S102) for acquiring, when it is determined that the damaged product has been detected in the determination step, the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera (11a) among the plurality of cameras that first imaged the damaged product, and the ID of the non-damaged imaging camera (11b) installed one behind the damaged imaging camera with reference to the predetermined direction; a notification step (S104) of notifying the user of the ID of the damaged product and the fact that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, A method for identifying damage, characterized in that, in the information acquisition process, when the product passes through one of the multiple cameras, it is known whether the product that passes through another camera located one camera behind the one camera with the specified direction as the reference is the damaged product.

[0065] <Appendix 2> A conveying line (10) consisting of a plurality of conveying rolls (100); a plurality of cameras (11a, 11b) installed at intervals along a predetermined direction in which the products (2a-2c) are conveyed on the conveying line; a plurality of sensors (12a-12d) for detecting the position of the product when the product being conveyed on the conveyor line passes by; A damage identification system (1) for identifying a damaged conveying roll (100a), which is a conveying roll having a damage that may cause damage to the product when the product passes through the damaged conveying roll (100a), when the damaged conveying roll is included in the conveying line, comprising: determining whether or not a damaged product (2a-2c) has been detected, the damaged product being the product that has been damaged by the damaged conveying roll, based on the image capturing results of the product acquired from the plurality of cameras and the position information of the product acquired from the plurality of sensors; a control unit (13) that, when it is determined that the damaged product has been detected, acquires the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera (11a) among the plurality of cameras that first captured an image of the damaged product, and the ID (11b) of the non-damaged imaging camera installed one behind the damaged imaging camera with reference to the predetermined direction; a notification unit (14) that notifies the user of the ID of the damaged product and that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, A damage identification system (1), characterized in that when the product passes through one of the multiple cameras, the control unit has already acquired information as to whether the product that has passed through another camera located just before the one camera based on the specified direction is the damaged product. [Explanation of symbols]

[0066] 1: Scratch Identification System 10: Conveyor line 10a: Second conveying line 100: Transport roll 100a: Transport roll (damaged transport roll) 11a, 11b: Camera 12a-12d: Sensors 13: Controller 14: Monitor 15: Storage 2a-2d: Product (damaged product) 3: Sample 4: Pusher

Claims

1. A method for identifying a damaged conveying roll, which is a conveying roll having a scratch that may cause a scratch on a product when the product passes through the damaged conveying roll, when the damaged conveying roll is included in a conveying line consisting of a plurality of conveying rolls, comprising the steps of: a determination step of capturing an image of the underside of the product using a plurality of cameras installed at intervals along a predetermined direction in which the product is transported on the transport line, and determining whether or not a damaged product, which is the product that has been damaged by the damaged transport roll, has been detected based on the image capturing results and position information of the product; an information acquisition process for acquiring, when it is determined in the determination process that the damaged product has been detected, the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera among the multiple cameras that first captured an image of the damaged product, and the ID of the non-damaged imaging camera installed one behind the damaged imaging camera with reference to the predetermined direction; a notification step of notifying the user of the ID of the damaged product and the fact that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, A method for identifying damage, characterized in that, in the information acquisition process, when the product passes through one of the multiple cameras, it is known whether the product that passes through another camera located one camera behind the one camera with the specified direction as the reference is the damaged product.

2. a second determination step of determining whether or not to stop operation of the conveying line based on the damage in the damaged product notified in the notification step and a predetermined index of severity of the damage; The flaw identification method described in claim 1, characterized in that if the operation of the conveying line is to be stopped, a notification is given to that effect, and if the operation of the conveying line is not to be stopped, a notification is given to the effect that all or part of the plurality of conveying rolls will be replaced after a certain date and time has passed.

3. A storage device for storing the image capturing results acquired from each of the plurality of cameras for each of the product IDs. The flaw identification method according to claim 1 or 2, further comprising a preserving step.

4. The flaw identification method according to claim 1 , wherein the number of the plurality of cameras is smaller than the number of the plurality of transport rolls.

5. 3. The flaw identification method described in claim 2, characterized in that, when it is determined in the second judgment process that the operation of the conveying line should be stopped, the sample is conveyed on the conveying rolls in the specific range in the specified direction a specified distance, and then, using the specified direction as a reference, the sample is conveyed in the opposite direction to the specified direction until the front of the sample is positioned in front of the imaging range of the flaw-free imaging camera, this procedure being repeated while changing the specified distance until the flaw is found.

6. The sample is transported on the transport roll without slipping, 6. The flaw identifying method according to claim 5, wherein the transport rolls have the same rotation speed, rotation direction, and rotation start and end timings.

7. With n=1, the sample is placed on the transport roll so that the leading edge of the sample is positioned in front of the imaging range of the flaw-free imaging camera based on the predetermined direction, the sample is transported in the predetermined direction by n times the distance between two adjacent transport rolls, and the sample is transported again in the opposite direction to the predetermined direction so that the leading edge is positioned in front of the imaging range of the flaw-free imaging camera; If the flaw is not found, add 1 to the value of n. If the value of n is equal to or less than the number of transport rolls in the specific range, the sample is transported again in the predetermined direction by n times the distance between two adjacent transport rolls, and the procedure of transporting the sample in the opposite direction to the predetermined direction so that the leading edge is again positioned in front of the imaging range of the flawless imaging camera is repeated; The method for identifying flaws according to claim 5 or 6, characterized in that if the value of n is greater than the number of transport rolls in the specific range, an output is made indicating that the damaged transport roll is not included in the transport rolls in the specific range.

8. When the pusher pushes the product onto the conveying line along the longitudinal direction of the plurality of conveying rolls and it is determined that the damaged product has been detected in the determining step, a second information acquisition step of acquiring the position of the damage in the damaged product obtained from the imaging result and a predetermined reference position where the pusher pushes out the product; a calculation step of calculating a width of the product in a direction parallel to the longitudinal direction of the plurality of conveying rolls when the product is conveyed on the plurality of conveying rolls, based on the position of the scratch on the damaged product obtained from the imaging result and the reference position; The method for identifying a defect according to any one of claims 1 to 7, further comprising a second notification step of notifying that the damaged transport roll is included in the transport rolls in the specific range, and of the ID and width of the damaged product.

9. In the calculation step, the reference position can be changed, and when changed, a new width is calculated based on the position of the scratch on the damaged product obtained from the imaging result and the changed reference position; 9. The flaw identifying method according to claim 8, wherein the second notification step notifies the reference position after the change and the new width.

10. a conveying line consisting of a plurality of conveying rolls; a plurality of cameras installed at intervals along a predetermined direction in which the products are conveyed on the conveying line; a plurality of sensors that detect the position of the product when the product being conveyed on the conveyor line passes by; A damage identification system for identifying a damaged transport roll, which is a transport roll having a damage that may cause damage to the product when the product passes through the damaged transport roll, when the damaged transport roll is included in the transport line, determining whether or not a damaged product, which is the product that has been damaged by the damaged transport roll, has been detected based on the image capture results of the product acquired from the plurality of cameras and the position information of the product acquired from the plurality of sensors; a control unit that, when it is determined that the damaged product has been detected, acquires the ID of the damaged product, the damage detection result based on all of the imaging results of the damaged product, the ID of the damaged imaging camera among the multiple cameras that first captured an image of the damaged product, and the ID of the non-damaged imaging camera installed one behind the damaged imaging camera with reference to the predetermined direction; a notification unit that notifies the user of the ID of the damaged product and that the damaged transport roll is included in a specific range of transport rolls arranged between the imaging range of the damaged product imaging camera and the imaging range of the non-damaged product imaging camera, A damage identification system characterized in that, at the time the product passes through one of the multiple cameras, the control unit has already acquired information as to whether the product that has passed through another camera located just before the one camera based on the specified direction is the damaged product.

11. the control unit determines whether to stop operation of the conveying line based on the damage in the damaged product notified by the notification unit and a predetermined index of severity of the damage; The flaw identification system described in claim 10, characterized in that the notification unit notifies the user if the operation of the conveying line is to be stopped, and notifies the user that all or part of the plurality of conveying rolls will be replaced after a certain date and time has passed if the operation of the conveying line is not to be stopped.

12. 12. The flaw identification system according to claim 10, further comprising a storage unit that stores the image capture results acquired from each of the plurality of cameras for each product ID.

13. The flaw identification system according to claim 10 , wherein the number of the plurality of cameras is smaller than the number of the plurality of transport rolls.

14. The device further includes a pusher that pushes the product onto the conveying line along the longitudinal direction of the plurality of conveying rolls, and when the control unit determines that the damaged product has been detected, The control unit acquires the position of the damage in the damaged product obtained from the imaging result and a predetermined reference position where the pusher pushes out the product, Calculating a width of the product in a direction parallel to the longitudinal direction of the plurality of conveying rolls when the product is conveyed on the plurality of conveying rolls based on the position of the scratch on the damaged product obtained from the imaging result and the reference position; The flaw identification system according to any one of claims 10 to 13, characterized in that the notification unit notifies that the damaged transport roll is included in the transport rolls in the specific range, and notifies the ID and width of the damaged product.

15. The reference position can be changed, and when changed, the control unit calculates a new width based on the position of the scratch in the damaged product obtained from the imaging result and the changed reference position, The flaw identification system according to claim 14 , wherein the notification unit notifies the reference position after the change and the new width.

Images