Systems and methods for detecting abnormalities during glass ribbon formation

Abstract

Apparatuses and methods are described for detecting anomalies during glass sheet production. In some examples, a glass forming apparatus draws down molten glass to produce a glass sheet. In addition, multiple cameras capture images of at least a portion of an interior of a housing assembly of the glass forming apparatus and the glass sheet. Further, the computing device receives the images from the cameras, and applies a machine vision process to the images. Based on the application of the machine vision process to the images, the computing device determines a position of an object in each of the images. The computing device further determines an anomaly based on the position of the object in each of the images. The computing device then generates anomaly detection data based on the anomaly, and stores the anomaly detection data in a database.

Description

SYSTEMS AND METHODS FOR DETECTING ABNORMALITIES DURINGGLASS RIBBON FORMATIONCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of priority under 35 U.S.C. § 119 of U.S. Provisional Application Serial No. 63 / 531932 filed on August 10, 2023, the content of which is relied upon and incorporated herein by reference in its entirety.FIELD OF THE DISCLOSURE

[0002] The present disclosure relates to the production of glass sheets and, more particularly, to apparatus and methods for detecting abnormalities during glass sheet production.BACKGROUND

[0003] Glass sheets are used in a variety of applications. For example, they may be used in glass display panels such as in mobile devices, laptops, tablets, computer monitors, and television displays. Glass sheets may be manufactured by a slot drawdown process whereby molten glass is drawn through a slot to form a glass sheet. In some instances, however, the slot drawdown process experiences abnormalities. For example, the glass sheet may crack during the drawdown process, or molten glass may spill unexpectedly. These abnormalities may result in the incurrence of unexpected costs. For example, new glass sheets need to be produced to replace the cracked glass sheets, and further production of additional glass sheets can be delayed due to having to produce the replacement glass sheets. In addition, the spilling of molten glass in unexpected places may result in damage to the glass drawdown system, resulting in additional costs. As such, there are opportunities to improve the production of glass sheets.SUMMARY

[0004] The embodiments disclosed herein are directed to apparatus and methods for detecting anomalies (e.g., abnormalities) within glass formation systems during glass sheet production. For instance, the embodiments may detect when molten glass drips unexpectedly during the formation of glass sheets. The embodiments may, additionally or alternatively, detect cracked glass sheets as well as other anomalies during the formation of glass sheets.

[0005] For instance, the embodiments may employ cameras to capture images of a glass sheet as its being formed and drawn down by glass drawing equipment (e.g. , rollers). The cameras may be positioned inside a cavity of a glass forming apparatus such that their field-of-view includes the glass and the glass drawing equipment. In some examples, at least two cameras capture images from respective ends of the glass as its being formed. For instance, a first camera may be positioned within the cavity of the glass forming apparatus at one end to capture images from a first view of the glass drawing equipment and the glass as its being formed. A second camera may be positioned within the cavity of the glass forming apparatus at another end to capture images from a second view of the glass drawing equipment and the glass as its being formed. In some instances, two, three, or more cameras are positioned at each end of the cavity of the glass drawing equipment to capture images from various views (e.g., opposite views) of the glass drawing equipment and the glass as its being formed.

[0006] Further, a computing device, such as a server, may receive the images, and may apply a machine vision process to the images to generate anomaly detection data characterizing whether any anomalies were detected. For instance, the server may receive images from each of the cameras over a corresponding interval of time during the formation of a glass sheet, and may apply the machine vision process described herein to the images. For example, the machine vision process may include generating features based on the images, and inputting the features to a trained machine learning or artificial intelligence model to generate the anomaly detection data. Based on the application of the machine vision process to the images, the computing device may detect a drip (e.g., a drip trajectory of molten glass), and may generate the anomaly detection data characterizing the detected drip. The computing device may perform additional operations based on the detected drip. For instance, the computing device may display an image of the detect drip, may transmit a warning signal (e.g., message), may disengagethe rollers of the glass drawing equipment from the glass being drawn down, or may stop the glass forming apparatus from further drawing down glass, among other examples.

[0007] In some embodiments, a glass forming system includes a glass forming apparatus, at least one camera, and a computing device communicatively coupled to the glass forming apparatus and the at least one camera. The glass forming apparatus is configured to drawdown molten glass to produce a glass sheet. In addition, the at least one camera is configured to capture images of at least a portion of the glass forming apparatus and the glass sheet. Further, the computing device is configured to receive the images from the at least one camera. The computing device is also configured to apply a machine vision process to the images. Based on the application of the machine vision process to the images, the computing device is configured to determine a position of at least one object in each of the images. The computing device is further configured to determine at least one anomaly based on the position of the at least one object in each of the images. The computing device is also configured to generate anomaly detection data based on the at least one anomaly. Further, the computing device is configured to store the anomaly detection data in a database.

[0008] In some embodiments, an apparatus includes a memory storing instructions and a processor communicatively coupled to the memory. The processor is configured to execute the instructions to receive images from at least one camera configured to capture images of at least a portion of a glass forming apparatus and a glass sheet. The computing device is also configured to apply a machine vision process to the images. Based on the application of the machine vision process to the images, the computing device is configured to determine a position of at least one object in each of the images. The computing device is further configured to determine at least one anomaly based on the position of the at least one object in each of the images. The computing device is also configured to generate anomaly detection data based on the at least one anomaly. Further, the computing device is configured to store the anomaly detection data in a database.

[0009] In some embodiments, a method by a processor includes receiving images from at least one camera configured to capture images of at least a portion of a glass forming apparatus and a glass sheet. The method also includes applying a machine vision process to the images. Based on the application of the machine vision process to the images, the method includes determining a position of at least one object in each of the images. The method further includes determining at least one anomaly based on the position of the at least one object in each of the images. The method also includesgenerating anomaly detection data based on the at least one anomaly. Further, the method includes storing the anomaly detection data in a database.

[0010] In some embodiments, a non-transitory computer readable medium stores instructions that, when executed by one or more processors, cause the one or more processors to perform a method that includes receiving images from at least one camera configured to capture images of at least a portion of a glass forming apparatus and a glass sheet. The method also includes applying a machine vision process to the images. Based on the application of the machine vision process to the images, the method includes determining a position of at least one object in each of the images. The method further includes determining at least one anomaly based on the position of the at least one object in each of the images. The method also includes generating anomaly detection data based on the at least one anomaly. Further, the method includes storing the anomaly detection data in a database.BRIEF DESCRIPTION OF DRAWINGS

[0011] The above summary and the below detailed description of illustrative embodiments may be read in conjunction with the appended Figures. The Figures show some of the illustrative embodiments discussed herein. As further explained below, the claims are not limited to the illustrative embodiments. For clarity and ease of reading, Figures may omit views of certain features.

[0012] FIG. 1 illustrates a glass forming anomaly detection system in accordance with some examples.

[0013] FIG. 2 is a block diagram of an anomaly detection computing device in accordance with some examples.

[0014] FIG. 3 illustrates exemplary functions of an anomaly detection computing device in accordance with some examples.

[0015] FIGS. 4A and 4B illustrate cameras that can be employed in a glass forming anomaly detection system in accordance with some examples.

[0016] FIG. 5 illustrates a generated image with a detected drip of molten glass in accordance with some examples.

[0017] FIG. 6 illustrates an exemplary method for detecting anomalies during a glass formation process in accordance with some examples.DETAILED DESCRIPTION

[0018] The present application discloses illustrative (z.e., example) embodiments. The disclosure is not limited to the illustrative embodiments. Therefore, many implementations of the claims will be different than the illustrative embodiments. Various modifications can be made to the claims without departing from the spirit and scope of the disclosure. The claims are intended to cover implementations with such modifications.

[0019] At times, the present application uses directional terms (e.g., front, back, top, bottom, left, right, side, edge, etc.) to give the reader context when viewing the Figures. The claims, however, are not limited to the orientations shown in the Figures. Any absolute term (e.g., high, low, etc.) can be understood as disclosing a corresponding relative term (e.g., higher, lower, etc.). In addition, any singular term (e.g., processor) can be understood as disclosing a corresponding plural term (e.g., processors), and any plural term can be understood as disclosing a corresponding singular term.

[0020] Referring now to the Figures, FIG. 1 illustrates a glass forming anomaly detection system 100 that includes a glass forming apparatus 120, a first camera 140, a second camera 142, a third camera 144, a fourth camera 146, an anomaly detection computing device 160, and a database 170. The glass forming apparatus 120 (e.g., a fusion glass forming apparatus) includes a cavity 121 that is bounded on its longitudinal sides by walls 123 and 124. The walls 123 and 124 are integral with each of opposed and downwardly inclined glass forming surfaces 133 and 134, respectively. The pair of opposed and downwardly inclined glass forming surfaces 133 and 134 terminate and meet at a joining position 135.

[0021] Molten glass (e.g., high-purity molten glass) is delivered into cavity 121 until the molten glass overflows each of the walls 123 and 124, and then flows along the glass forming surfaces 133 and 134 to C. The molten glass flowing over the glass forming surfaces 133 and 134 meets at the joining position 135 and begins to form a glass sheet 150 (e.g., glass ribbon). Pulling rolls 137A, 137B are located downstream of the joining position 135 and engage each of the side edges at both sides of the glass sheet 150 to apply tension to the glass sheet 150. The pulling rolls 137A, 137B may be positioned sufficiently below the joining position 135 that the thickness of the glass sheet 150 is essentially fixed at that location. The pulling rolls 137A, 137B may draw the glass sheet150 downwardly at a prescribed rate that establishes the thickness of the glass sheet 150 as it is drawn from the joining position 135.

[0022] As such, the molten glass draws down from the joining position 135 until a glass sheet 150 is formed. The glass sheet 150 may be formed to any suitable thickness. For example, the glass sheet 150 may be anywhere from 500 micrometers to 3 millimeters (e.g., 2 millimeters) thick in some instances. In other instances, the glass sheet 150 may be thicker than 3 millimeters, or less than 500 micrometers thick.

[0023] As illustrated, the pulling rolls 137A, 137B draw the glass sheet 150 down through a glass sheet housing 102 e.g., a tank). Each of the first camera 140, second camera 142, third camera 144, and fourth camera 146 may be mounted and positioned within the housing 102 to capture images of the glass sheet 150 as its being formed, as well as corresponding inside portions of the housing 102. For instance, the first camera 140 may be mounted at a first end 151 of the housing 102 and be positioned to have its field-of-view include at least a top portion of the glass sheet 150. The second camera 142 may also be mounted at the first end 151 of the housing 102 at a position below the first camera 140, and be positioned to have its field-of-view include at least a bottom portion of the glass sheet 150. The field-of views of the first camera 140 and the second camera 142 may overlap, and may include portions of the inside of the housing, as described herein.

[0024] Likewise, the third camera 144 may be mounted at a second end 153 of the housing 102 and be positioned to have its field-of-view include at least the top portion of the glass sheet 150. The fourth camera 146 may also be mounted at the second end 153 of the housing 102 at a position below the third camera 144, and be positioned to have its field-of-view include at least the bottom portion of the glass sheet 150. The field-of views of the third camera 144 and the fourth camera 146 may overlap, and may include portions of the inside of the housing, as described herein.

[0025] Further, each of the first camera 140, second camera 142, third camera 144, and fourth camera 146 may be any suitable camera, such as a high definition video camera.

[0026] Moreover, the anomaly detection computing device 160 may be communicatively coupled to each of the first camera 140, second camera 142, third camera 144, and fourth camera 146 and, in some examples, glass forming apparatus 120. For example, the anomaly detection computing device 160 may be configured to communicate over a communication network 116 with each of the first camera 140,second camera 142, third camera 144, and fourth camera 146. Communication network 116 may be a wired or wireless communication network. For example, in some examples, communication network 116 may include one or more Universal Serial Bus® (USB) (e.g., USB 2.0) or Ethernet® cables establishing communication channels between the anomaly detection computing device 160 and each of the first camera 140, second camera 142, third camera 144, and fourth camera 146. In some examples, communication network 116 may include a cellular network, a Wi-Fi® network, a Bluetooth® network, or any other suitable wireless network. In some instances, the anomaly detection computing device 160 communicates with one or more of the first camera 140, second camera 142, third camera 144, and fourth camera 146 using one or more protocols, such as the real-time streaming protocol (RTSP). In some examples, the anomaly detection computing device 160 transmits a signal (e.g., a message) to any of the first camera 140, second camera 142, third camera 144, and fourth camera 146 to cause the receiving camera to adjust its sensor’s position (e.g., adjusting yaw, pitch, roll), thereby adjusting the camera’s field of view (e.g., to adjust its field of view up, down, left, or right).

[0027] Moreover, the anomaly detection computing device 160 can receive images from each of the first camera 140, second camera 142, third camera 144, and fourth camera 146. For example, the first camera 140 may capture an image that includes a top portion of a first edge 155 of the glass sheet 150, and may transmit the image to the anomaly detection computing device 160 using communication network 116. Similarly, the second camera 142 may capture an image that includes a bottom portion of the first edge 155 of the glass sheet 150, and may transmit the image to the anomaly detection computing device 160 using communication network 116. Further, the third camera 1444 may capture an image that includes the top portion of a second edge 157 of the glass sheet 150, and may transmit the image to the anomaly detection computing device 160 using communication network 116. Similarly, the fourth camera 146 may capture an image that includes a bottom portion of the second edge 157 of the glass sheet 150, and may transmit the image to the anomaly detection computing device 160 using communication network 116. In some instances, each of the first camera 140, second camera 142, third camera 144, and fourth camera 146 capture images of the glass sheet 150 over a temporal interval (e.g., as the glass sheet 150 is being drawn down), and transmits the images to the anomaly detection computing device 160.

[0028] The anomaly detection computing device 160 receives the images and applies a machine vision process to the images to determine a position of at least oneobject in the images. As described herein, the machine vision process may include generating features based on the received images, and inputting the generated features to one or more trained machine learning or artificial intelligence models. Based on the inputted features, the trained machine learning or artificial intelligence model may generate output data characterizing one or more detected objects and / or a position of the one or more detected objects. For instance, the output data may characterize a detected object and its position, which may be a droplet of molten glass within the housing 102 (e.g., dripping from the walls 123, 124 of the glass forming apparatus 120 to inside the housing 102). In some instances, the object may be a crack of the glass sheet 150.

[0029] Based on the detected objects, the anomaly detection computing device 160 may determine at least one anomaly based on the position of the at least one object in each of the images. For instance, the anomaly detection computing device 160 may compare the positions of the detected object across various images to determine the object is moving (e.g., a falling drop of molten glass, a drip) and, based on its position in each image, may determine that the object is falling through housing 102. Because no objects are expected to fall through housing 102, the anomaly detection computing device 160 determines the object (e.g., the drop of molten glass falling through housing 102) is an anomaly. In addition, the anomaly detection computing device 160 may generate anomaly detection data characterizing the detected anomaly, and may store the generated anomaly detection data within database 170. In some instances, and based on the anomaly detection data, the anomaly detection computing device 160 generate an alert (e.g., display an alert, transmit an alert message to another processing device), generate an anomaly map, or cause the glass to be moved to a different output line, among other examples.

[0030] In some instances, applying the machine vision process to an image, such as an image captured by the first camera 140, includes the following processing steps. The anomaly detection computing device 160 receives an image, and filters the image to generate a differentiated image. For example, the anomaly detection computing device 160 may compare each pixel value of the image to a corresponding threshold value. Pixel values that are less than the threshold value are adjusted to a first value (e.g. , 0). The first value may correspond to a black pixel color, for instance.

[0031] Further, the anomaly detection computing device 160 may determine an intersection and / or union between the differentiated image and a second image, and generates a matched image based on the intersection and / or union. The second imagemay be an image captured from the same or another of the first camera 140, second camera 142, third camera 144, and fourth camera 146. For instance, the second image may be one captured of the glass sheet 150 from a same or different perspective at an earlier point in time. In addition, the anomaly detection computing device 160 may determine a connectivity between the pixels of the matched image to generate a first connected image. For instance, the anomaly detection computing device 160 may determine a “4-connectivity” or “8-connectivity” between the pixels of the matched image to generate the first connected image.

[0032] In addition, the anomaly detection computing device 160 applies a closing process to the first connected image to generate a closed image. For instance, the anomaly detection computing device 160 may dilate the first connected image, and may then erode the dilated image to generate the closed image. The closing process allows for the “filling” of small holes in the first connected image, and preserves the shape and size of large holes and objects in the first connected image.

[0033] The anomaly detection computing device 160 may then determine a connectivity between the pixels of the closed image to generate a second connected image. For instance, the anomaly detection computing device 160 may determine a “4- connectivity” or “8-connectivity” between the pixels of the closed image to generate the second connected image.

[0034] Further, the anomaly detection computing device 160 may determine a region of interest of the second connected image, and may generate an ROI image based on the region of interest of the second connected image. The region of interest may be a predetermined area of the image that may include anomalies. For instance, database 170 may include one or more ROI maps that define ROIs of images. The anomaly detection computing device 160 may obtain an ROI map from the database 170, and may apply the ROI map to the second connected image to generate the ROI image. For example, pixel values for pixels that fall outside of the ROI map may be adjusted to a particular value (e.g., 0), while pixel values for pixels that fall within the ROI map are not adjusted.

[0035] In addition, the anomaly detection computing device 160 may perform operations to scale the ROI image. For instance, the anomaly detection computing device 160 may center the ROI image by subtracting a scaling pixel value e.g., a predetermined pixel value, such as a predetermined mean pixel value) from each pixel value of each pixel of the ROI image to generate a centered image. The anomaly detection computing device 160 may store the centered image in database 170.

[0036] The anomaly detection computing device 160 may perform the above operations for images received from one or more of the first camera 140, second camera 142, third camera 144, and fourth camera 146 and, in some examples, for images received over a time interval. The time interval may correspond to the time during which the glass sheet 150 is being drawn down. Based on the centered images, the anomaly detection computing device 160 may generate anomaly detection data characterizing one or more anomalies within the centered images.

[0037] In some instances, when an anomaly is detected, the anomaly detection computing device 160 performs additional operations. For example, the anomaly detection computing device 160 may provide one or more of the centered images for display when an anomaly has been detected. In some examples, the anomaly detection computing device 160 may transmit a warning message. For instance, the anomaly detection computing device 160 may transmit a warning message to a mobile device (e.g., a cellphone, smartphone, etc.) operated by a technician of a glass forming company. In response to receiving the warning message, the mobile device may display the warning message. In some examples, the anomaly detection computing device 160 may shut down the glass forming apparatus 120, or may disengage the pulling rolls 137A, 137B from drawing down the glass sheet 150 e.g., transmit a signal to the pulling rolls 137A, 137B to cause the pulling rolls 137A, 137B to move away from the glass sheet 150, thereby removing contact to the glass sheet 150.

[0038] FIG. 2 illustrates an exemplary anomaly detection computing device 200, such as the anomaly detection computing device 160 of FIG. 1. As illustrated, anomaly detection computing device 200 can include one or more processors 201, working memory 202, one or more input / output devices 203, instruction memory 207, a transceiver 204, and a display 206, all operatively coupled to one or more data buses 208. Data buses 208 allow for communication among the various devices. Data buses 208 can include wired, or wireless, communication channels.

[0039] Processors 201 can include one or more distinct processors, each having one or more cores. Each of the distinct processors can have the same or different structure. Processors 201 can include one or more central processing units (CPUs), one or more graphics processing units (GPUs), application specific integrated circuits (ASICs), digital signal processors (DSPs), and the like. Processors 201 can be configured to perform a certain function or operation by executing code, stored on instruction memory 207,embodying the function or operation. For example, processors 201 can be configured to perform one or more of any function, method, or operation disclosed herein.

[0040] Instruction memory 207 can store instructions that can be accessed (e.g., read) and executed by processors 201. For example, instruction memory 207 can be a non-transitory, computer-readable storage medium such as a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), flash memory, a removable disk, CD-ROM, any non-volatile memory, or any other suitable memory.

[0041] Further, processors 201 can store data to, and read data from, working memory 202. For example, processors 201 can store a working set of instructions to working memory 202, such as instructions loaded from instruction memory 207. Processors 201 can also use working memory 202 to store dynamic data created during the operation of anomaly detection computing device 200. Working memory 202 can be a random access memory (RAM) such as a static random access memory (SRAM) or dynamic random access memory (DRAM), or any other suitable memory.

[0042] Input-output devices 203 can include any suitable device that allows for data input or output. For example, input-output devices 203 can include one or more of a keyboard, a touchpad, a mouse, a stylus, a touchscreen, a physical button, a speaker, a microphone, or any other suitable input or output device. Input-output device 203 may allow a user to provide input selecting or characterizing preferred operational values, such as camera positions, for instance.

[0043] Display 209 can display user interface 205. User interface 205 can enable user interaction with the anomaly detection computing device 200. For example, user interface 205 can be a user interface for an application that allows a user to enter preferred operational values to update a position of a camera, such as a position of the first camera 140, the second camera 142, the third camera 144, and the fourth camera 146, for example. In some examples, a user can interact with user interface 205 by engaging inputoutput devices 203. In some examples, display 209 can be a touchscreen, where user interface 205 is displayed on the touchscreen.

[0044] Transceiver 204 allows for communication with a network, such as the communication network 116. For example, transceiver 204 may connect to a Wi-Fi®, Bluetooth®, cellular, or any other suitable wireless network, and may send signals (e.g., data) to, and receive signals from, other devices coupled to the wireless network. Processor(s) 201 is operable to receive data from, or send data to, the wireless network via transceiver 204.

[0045] FIG. 3 illustrates exemplary functions, such as machine vision functions, of the anomaly detection computing device 200. As illustrated, anomaly detection computing device 200 includes image differentiator engine 302, image matching engine 304, connection engine 306, closing engine 308, ROI engine 310 and centering engine 312. In some instances, one or more of the image differentiator engine 302, image matching engine 304, connection engine 306, closing engine 308, ROI engine 310 and centering engine 312 include instructions stored in instruction memory 207 that, when executed by the one or more processors 201, cause the one or more processors 201 to carry out the corresponding operations. In some examples, one or more of the image differentiator engine 302, image matching engine 304, connection engine 306, closing engine 308, ROI engine 310 and centering engine 312 can be implemented in logic, such as digital logic, state machines, FPGAs, ASICs, or any other suitable logic.

[0046] Image differentiator engine 302 may receive image data 301. For example, image differentiator engine 302 may receive image data 301 characterizing a captured image from one or more of the first camera 140, second camera 142, third camera 144, and fourth camera 146. The image data 301 may characterize an image of an internal portion of housing 102 and a portion of the glass sheet 150. Further, image differentiator engine 302 may filter the image data 301 to generate differentiated image data 303 characterizing a differentiated image. For example, image differentiator engine 302 may compare each pixel value of the image data 301 to a corresponding threshold value, and may adjust pixel values that are less than the threshold value to a first value (e.g., 0) to generate the differentiated image data 303.

[0047] Further, image matching engine 304 may receive the differentiated image data 303, and may determine an intersection and / or union between the differentiated image and a second image. As described herein, the second image may be an image captured from the same or another of the first camera 140, second camera 142, third camera 144, and fourth camera 146. For instance, the second image may be one captured of the glass sheet 150 from a same or different perspective at an earlier point in time. Based on the determined intersection and / or union, the image matching engine 304 generates matched image data 305 characterizing a matched image.

[0048] Connection engine 306 receives the matched image data 305, and determines a connectivity between the pixels of the matched image to generate first image connected data 307 characterizing a first connected image. For instance, the connectionengine 306 may determine a “4-connectivity” or “8-connectivity” between the pixels of the matched image to generate the first connected image.

[0049] Closing engine 308 receives the first image connected data 307, and applies a closing process to the first image connected data 307 to generate closed image data 309 characterizing a closed image. For instance, the closing engine 308 may dilate the first connected image, and may then erode the dilated image to generate the closed image. The connection engine 306 receives the closed image data 309, and determines a connectivity between the pixels of the closed image to generate second image connected data 311 characterizing a second connected image. For instance, the closing engine 308 may determine a “4-connectivity” or “8-connectivity” between the pixels of the closed image to generate the second connected image.

[0050] ROI engine 310 may receive the second image connected data 311, and may determine a region of interest of the second connected image based on an ROI map, such as an ROI map stored in database 170. The ROI engine 310 may generate ROI image data 313 characterizing an ROI image based on the determined region of interest of the second connected image. For instance, the ROI image may include adjusted pixel values for pixels that fall outside of the ROI map (e.g., adjusted to 0).

[0051] In addition, the centering engine 311 may receive the ROI image data 313, and may scale the ROI image data 313 to generate centered image data 315 characterizing a centered image. For instance, the centering engine 311 may center the ROI image by subtracting a scaling pixel value from each pixel value of each pixel of the ROI image to generate the centered image. The centering engine 311 may store the centered image data 315 within database 170. In some instances, the centering engine 311 provides the centered image data 315 to display 209 for display.

[0052] FIGS. 4 A and 4B illustrate views of an assembly 400 that includes a first camera 402, a second camera 404, and a third camera 406. Assembly 400 may correspond to the first end 151 of the housing 102 that includes the first camera 140 and the second camera 142. Assembly 400 may also correspond to the second end 153 of the housing 102 that includes the third camera 140 and the second camera 142. For example, housing 102 may include a first assembly 400 at the first end 151, and a second assembly 400 at the second end 153. The second camera 404 may be mounted to assembly 400 at a predetermined distance from the first camera 402. Likewise, the third camera 406 may be mounted to assembly 400 at a predetermined distance from the second camera 404.

[0053] FIG. 5 illustrates an example of a generated centered image, such as one characterized by centered image data 315 described with respect to FIG. 3. For example, and as described herein, anomaly detection computing device 160 may receive images from one or more cameras positioned within a glass housing assembly, and applies a machine vision process to the images to determine a position of at least one object in the images. Based on the position of the at least one object in the images, the anomaly detection computing device 160 may detect an anomaly during a glass formation process. For instance, as illustrated in FIG. 5, the anomaly may be dripping molten glass 702, among other examples. In this instance, the drip of molten glass can be seen falling through a housing 701 through which a glass sheet is formed.

[0054] FIG. 6 illustrates an exemplary method 800 that may be performed by one or more computing devices, such as the anomaly detection computing device 160. In some examples, the exemplary method 800 may be carried out by one or more processors, such as one or more processors distributed across one or more networks (e.g., processors within one or more cloud-based servers).

[0055] Beginning at step 802, images are received from cameras. For example, first camera 140 may be positioned within glass sheet housing 102 and configured to capture images of a glass sheet 150 being formed from a first end 151 of the housing 102. In addition, third camera 144 may be positioned within the glass sheet housing 102 and configured to capture images of the glass sheet 150 being formed from a second end 153 of the housing 102. Each of the first camera 140 and the third camera 144 may capture images during the glass formation process, and may transmit the images to the anomaly detection computing device 160.

[0056] Further, at step 804, anomaly detection data is generated based on applying a machine vision process to the images. For example, and as described herein, anomaly detection computing device 160 may filter each image to generate a corresponding differentiated image. For example, the anomaly detection computing device 160 may compare each pixel value of each image to a corresponding threshold value. Pixel values that are less than the threshold value are adjusted to a first value (e.g., 0). Further, the anomaly detection computing device 160 may determine an intersection and / or union between the differentiated image and a second image, and generates a matched image based on the intersection and / or union. In addition, the anomaly detection computing device 160 may determine a connectivity between the pixels of the matched image to generate a first connected image.

[0057] Moreover, the anomaly detection computing device 160 may apply a closing process to the first connected image to generate a closed image. For instance, the anomaly detection computing device 160 may dilate the first connected image, and may then erode the dilated image to generate the closed image. The closing process allows for the “filling” of small holes in the first connected image, and preserves the shape and size of large holes and objects in the first connected image. The anomaly detection computing device 160 may then determine a connectivity between the pixels of the closed image to generate a second connected image.

[0058] Further, the anomaly detection computing device 160 may determine a region of interest of the second connected image, and may generate an ROI image based on the region of interest of the second connected image. In addition, the anomaly detection computing device 160 may perform operations to scale the ROI image. For instance, the anomaly detection computing device 160 may center the ROI image by subtracting a scaling pixel value (e.g., a predetermined pixel value, such as a predetermined mean pixel value) from each pixel value of each pixel of the ROI image to generate a centered image. Based on the centered images, the anomaly detection computing device 160 may generate anomaly detection data characterizing one or more anomalies within the centered images. For instance, and as described herein, based on the position of a detected object in each of the images, the anomaly detection computing device 160 may determine that molten glass is dripping within the glass sheet housing 102.

[0059] Proceeding to step 806, if no anomaly is detected, the method proceeds back to step 802, where additional images are received from the cameras. If, however, an anomaly was detected, the method proceeds to step 808, where one or more operations are performed based on the anomaly detection. For example, at step 810, and image, such as a centering image showing a drip of molten glass (e.g., image 700), is displayed. In some examples, at step 812, a warning message is transmitted. In some examples, at step 814, the rollers (e.g., pulling rolls 137A, 137B) of the glass forming apparatus (e.g., glass forming apparatus 120) may be disengaged from the glass sheet (e.g., glass sheet 150).

[0060] Although the methods described above are with reference to the illustrated flowcharts, it will be appreciated that many other ways of performing the acts associated with the methods can be used. For example, the order of some operations may be changed, and some of the operations described may be optional.

[0061] In addition, the methods and system described herein can be at least partially embodied in the form of computer-implemented processes and apparatus for practicing those processes. The disclosed methods may also be at least partially embodied in the form of tangible, non-transitory machine-readable storage media encoded with computer program code. For example, the steps of the methods can be embodied in hardware, in executable instructions executed by a processor (e.g., software), or a combination of the two. The media may include, for example, RAMs, ROMs, CD-ROMs, DVD-ROMs, BD- ROMs, hard disk drives, flash memories, or any other non-transitory machine-readable storage medium. When the computer program code is loaded into and executed by a computer, the computer becomes an apparatus for practicing the method. The methods may also be at least partially embodied in the form of a computer into which computer program code is loaded or executed, such that, the computer becomes a special purpose computer for practicing the methods. When implemented on a general-purpose processor, the computer program code segments configure the processor to create specific logic circuits. The methods may alternatively be at least partially embodied in application specific integrated circuits for performing the methods.

[0062] The foregoing is provided for purposes of illustrating, explaining, and describing embodiments of this disclosure. Modifications and adaptations to these embodiments will be apparent to those skilled in the art and may be made without departing from the scope or spirit of this disclosure.

Claims

CLAIMSWhat is claimed is:

1. A glass forming system comprising: a glass forming apparatus configured to drawdown molten glass to produce a glass sheet; at least one camera configured to capture images of at least a portion of the glass forming apparatus and the glass sheet; and a computing device communicatively coupled to the at least one camera, the computing device configured to: receive the images from the at least one camera; apply a machine vision process to the images and, based on the application of the machine vision process to the images, determine a position of at least one object in each of the images; determine at least one anomaly based on the position of the at least one object in each of the images; generate anomaly detection data based on the at least one anomaly; and store the anomaly detection data in a database.

2. The glass forming system of claim 1, wherein the computing device is configured to: generate a warning message based on the anomaly detection data; and transmit the warning message to another computing device.

3. The glass forming system of claim 1, wherein the at least one camera comprises a first camera and a second camera, the first camera having a field-of-view from a first end of the glass forming apparatus, and the second camera having a field-of- view from a second end of the glass forming apparatus.

4. The glass forming system of claim 3, wherein the at least one camera comprises a third camera and a fourth camera, the third camera having a field-of-view from the first end of the glass forming apparatus, and the fourth camera having a field-of- view from the second end of the glass forming apparatus.

5. The glass forming system of claim 1, wherein the at least one camera is positioned within a housing of the glass forming apparatus.

6. The glass forming system of claim 1, wherein the at least one anomaly is a drip of molten glass.

7. The glass forming system of claim 1, wherein, to apply the machine vision process to the images, the computing device is configured to: generate features based on the images; input the features to a trained machine learning or artificial intelligence mode; and generate output data characterizing the at least one object and the position of the at least one object.

8. The glass forming system of claim 1, wherein each of the images comprise a plurality of pixels, and wherein, to apply the machine vision process to the images, the computing device is configured to, for each of the images: compare each pixel value of each of the plurality of pixels to a corresponding threshold value; adjust the pixel value of each of the plurality of pixels when the pixel value is less than the corresponding threshold value to generate a differentiated image; and determine the position of the at least one object based on the differentiated image.

9. The glass forming system of claim 8, wherein, to apply the machine vision process to the images, the computing device is configured to: determine at least one of an intersection and a union between the differentiated image and a second image; generate a matched image based on the at least one of the intersection and the union; and determine the position of the at least one object based on the matched image.

10. The glass forming system of claim 9, wherein, to apply the machine vision process to the images, the computing device is configured to:determine a first connectivity between pixels of the matched image to generate a first connected image; and determine the position of the at least one obj ect based on the first connected image.

11. The glass forming system of claim 10, wherein, to apply the machine vision process to the images, the computing device is configured to: apply a closing process to the first connected image to generate a closed image; determine a second connectivity between pixels of the closed image to generate a second connected image; and determine the position of the at least one object based on the second connected image.

12. The glass forming system of claim 11, wherein, to apply the machine vision process to the images, the computing device is configured to: determine a region of interest of the second connected image; generate an ROI image based on the region of interest of the second connected image; and determine the position of the at least one object based on the ROI image.

13. The glass forming system of claim 12, wherein, to apply the machine vision process to the images, the computing device is configured to: subtract a scaling pixel value from each pixel value of each pixel of the ROI image to generate a centered image; and determine the position of the at least one object based on the centered image.

14. The glass forming system of claim 1, wherein the computing device is configured to: compare the position of the at least one object in each of the images; and determine the at least one object is moving.

15. An apparatus comprising: a memory storing instructions; and a processor communicatively coupled to the memory, wherein the processor is configured to execute the instructions to:receive images from at least one camera configured to capture images of at least a portion of a glass forming apparatus and a glass sheet; apply a machine vision process to the images and, based on the application of the machine vision process to the images, determine a position of at least one object in each of the images; determine at least one anomaly based on the position of the at least one object in each of the images; generate anomaly detection data based on the at least one anomaly; and store the anomaly detection data in a database.

16. The apparatus of claim 15, wherein the at least one anomaly is a drip of molten glass.

17. The apparatus of claim 15, wherein, to apply the machine vision process to the images, the processor is configured to execute the instructions to: generate features based on the images; input the features to a trained machine learning or artificial intelligence mode; and generate output data characterizing the at least one object and the position of the at least one object.

18. The apparatus of claim 15, wherein each of the images comprise a plurality of pixels, and wherein, to apply the machine vision process to the images, the processor is configured to execute the instructions to, for each of the images: compare each pixel value of each of the plurality of pixels to a corresponding threshold value; adjust the pixel value of each of the plurality of pixels when the pixel value is less than the corresponding threshold value to generate a differentiated image; and determine the position of the at least one object based on the differentiated image.

19. The apparatus of claim 18, wherein, to apply the machine vision process to the images, the processor is configured to execute the instructions to: determine at least one of an intersection and a union between the differentiated image and a second image;generate a matched image based on the at least one of the intersection and the union; determine a first connectivity between pixels of the matched image to generate a first connected image; apply a closing process to the first connected image to generate a closed image; determine a second connectivity between pixels of the closed image to generate a second connected image; determine a region of interest of the second connected image; generate an ROI image based on the region of interest of the second connected image; subtract a scaling pixel value from each pixel value of each pixel of the ROI image to generate a centered image; and determine the position of the at least one object based on the centered image.

20. A method by at least one processor comprises: receiving images from at least one camera configured to capture images of at least a portion of a glass forming apparatus and a glass sheet; applying a machine vision process to the images and, based on the application of the machine vision process to the images, determine a position of at least one object in each of the images; determining at least one anomaly based on the position of the at least one object in each of the images; generating anomaly detection data based on the at least one anomaly; and storing the anomaly detection data in a database.

Images