Automatic monitoring system and method for blockage of gypsum slurry spraying port

Abstract

The invention relates to the technical field of gypsum board production, in particular to an automatic monitoring system and method for blockage of a gypsum slurry spraying port, and the system comprises an industrial camera used for collecting a spraying image formed by spraying of the gypsum slurry spraying port on a hard-edge hard-surface roller, and an image processing unit used for carrying out image processing on the spraying image, the material spraying area formed by material spraying of the material spraying opening on the hard-edge hard-surface roller is obtained; and the blockage early warning unit is used for judging the material spraying area based on a preset threshold value, determining the blockage state of the material spraying opening and carrying out early warning according to the blockage state. Area detection can be formed by utilizing the industrial camera, the uniformity condition of the spraying material sprayed to the hard edge and the hard surface is automatically judged, the blockage condition of the spraying opening is further judged, automatic alarm is synchronously performed, a worker is reminded of blockage of the spraying opening, and automatic and accurate blockage monitoring is achieved.

Description

Technical Field

[0001] This invention relates to the field of gypsum board production technology, specifically to an automatic monitoring system and method for clogging of gypsum slurry spray nozzles. Background Technology

[0002] In gypsum board production, the spray nozzles for gypsum slurry can become clogged over time, leading to uneven application of the slurry to the hard-edge and hard-surface rollers, or even a complete lack of slurry. This results in uneven gypsum board surfaces, which operators cannot monitor in real time, ultimately causing uneven board surfaces and insufficient slurry, leading to defective products.

[0003] Existing technologies typically use flow alarm valves to monitor nozzle blockage, but since flow alarm valves can also become blocked, this can lead to monitoring failure or ineffective monitoring. Summary of the Invention

[0004] The purpose of this invention is to provide an automatic monitoring system and method for clogging of gypsum slurry spray nozzles, in order to solve the technical problem in the prior art where the flow alarm valve can also be clogged, leading to monitoring failure.

[0005] To solve the above-mentioned technical problems, the present invention specifically provides the following technical solution:

[0006] An automatic monitoring system for clogging of gypsum slurry spray nozzles includes:

[0007] An industrial camera is used to capture images of gypsum slurry being sprayed onto a hard-edged, hard-faced roller.

[0008] The image processing unit is used to perform image processing on the spray image to obtain the spray area formed by the spray nozzle spraying material on the hard-edged hard-faced roller.

[0009] The blockage warning unit is used to judge the spray area based on a preset threshold, determine the blockage status of the spray nozzle, and issue a warning based on the blockage status.

[0010] As a preferred embodiment of the present invention, the optimal values ​​of the exposure time and frame rate of the industrial camera acquiring the sprayed image are determined by a pre-established optimal parameter setting model;

[0011] The optimal parameter setting model is as follows:

[0012] KPI = BP(ET, FPS);

[0013] In the formula, ET is the exposure time, FPS is the frame rate, KPI is the accuracy of clogging identification based on the sprayed material images acquired by the industrial camera with ET and FPS, and BP is the BP neural network.

[0014] In this experiment, the spray nozzles of the gypsum slurry were artificially blocked to serve as the test spray nozzles.

[0015] Multiple exposure times and frame rates were selected as experimental exposure times and frame rates.

[0016] The industrial camera was controlled to continuously acquire images of the material spraying from the experimental nozzle onto the hard-edged hard-faced roller at each experimental exposure time and experimental frame rate, which were used as experimental image sequences.

[0017] The spray area is extracted from each spray image in the experimental image sequence, and the spray area is judged based on a preset threshold to determine the clogging state of the spray nozzle in the spray image, thus obtaining the experimental image spray nozzle clogging state sequence.

[0018] The proportion of images that are consistent with artificial blockage states in the sequence of blockage states at the nozzle of the experimental images is statistically analyzed, and this proportion is used as the accuracy of blockage identification.

[0019] Using each experimental exposure time and frame rate, along with the corresponding blockage identification accuracy, as training samples, a BP neural network is trained based on these training samples to construct an optimal parameter setting model that maps experimental exposure time and frame rate to the blockage identification accuracy.

[0020] The optimal parameter setting model is used to calculate each selectable value of the exposure time and frame rate of the industrial camera, and the accuracy of blockage identification corresponding to each selectable value of the exposure time and frame rate of the industrial camera is obtained.

[0021] The exposure time and frame rate values ​​corresponding to the highest accuracy of blockage recognition are selected as the optimal values ​​for the exposure time and frame rate of the industrial camera to acquire the sprayed material image.

[0022] As a preferred embodiment of the present invention, the step of the image processing unit determining the spray area includes:

[0023] Discreteness calculations are performed on the spray images captured in real time by the industrial camera and the spray images captured in history by the industrial camera in order to filter out the images of interest for blockage identification from the spray images captured in real time.

[0024] The spray area is extracted from the image of interest to obtain the spray area used for clogging identification.

[0025] As a preferred embodiment of the present invention, the step of the image processing unit selecting images of interest for blockage identification from the real-time acquired spray images includes:

[0026] The discretization between the real-time spray images captured by the industrial camera and the historical spray images captured by the industrial camera is calculated, and the discretization is as follows:

[0027]

[0028] In the formula, L represents discreteness, G is the pixel matrix of the real-time acquired sprayed material image, and G k Let |GG| be the pixel matrix of the k-th historically acquired spray image. k |For G and G k The Euclidean distance between them, where m is the total number of historically acquired spray images, and k is the counting variable;

[0029] The discreteness is compared with a discreteness threshold, where,

[0030] When the discreteness is greater than or equal to the discreteness threshold, the real-time acquired spray image is marked as the image of interest for clogging identification.

[0031] When the discreteness is less than the discreteness threshold, the real-time acquired spray images are marked as non-interest images for clogging identification.

[0032] As a preferred embodiment of the present invention, the step of the image processing unit extracting the spray area from the image of interest includes:

[0033] The image of interest is segmented using a threshold to obtain the image region in the image of interest used to calculate the spray area.

[0034] The number of pixels in the image region used to measure the spray area in the image of interest is counted to obtain the spray area used for clogging identification.

[0035] As a preferred embodiment of the present invention, the step of the congestion warning unit in determining the congestion status includes:

[0036] The spray area used for clogging identification is compared with a preset threshold, where,

[0037] If the spray area used for blockage identification is within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be blocked.

[0038] If the spray area used for clogging identification is not within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be in a non-clogging state.

[0039] As a preferred embodiment of the present invention, the blockage warning unit performs a warning operation when the gypsum slurry spray nozzle is blocked.

[0040] In a preferred embodiment of the present invention, the industrial camera, the image processing unit, and the congestion warning unit are sequentially and communicatively connected, wherein...

[0041] The industrial camera transmits the captured images of the sprayed material to the image processing unit;

[0042] The image processing unit transmits the processed spray area to the blockage warning unit.

[0043] As a preferred embodiment of the present invention, the preset threshold is the spray area in the spray image formed when the spray nozzle of the gypsum slurry is in a non-blocked state.

[0044] As a preferred embodiment of the present invention, the present invention provides an automatic monitoring method for clogging of gypsum slurry spray nozzles, applied to the aforementioned automatic monitoring system for clogging of gypsum slurry spray nozzles, the method comprising the following steps:

[0045] Images of gypsum slurry being sprayed onto a hard-edged, hard-faced roller are captured.

[0046] The spray image is processed to obtain the spray area formed by the spray nozzle spraying material onto the hard-edged hard-faced roller;

[0047] The spray area is judged based on a preset threshold to determine the blockage status of the spray nozzle, and an early warning is issued based on the blockage status.

[0048] Compared with the prior art, the present invention has the following advantages:

[0049] This invention utilizes an industrial camera to detect the area of ​​the sprayed material, automatically judging the uniformity of the sprayed material on hard edges and surfaces, thereby determining the blockage status of the spray nozzle, and simultaneously triggering an automatic alarm to alert the operator of the nozzle blockage, thus achieving automatic and accurate blockage monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0051] Figure 1 A block diagram of an automatic monitoring system for clogging of gypsum slurry spray nozzles provided in an embodiment of the present invention;

[0052] Figure 2 A flowchart of an automatic monitoring method for clogging of gypsum slurry spray nozzles provided in an embodiment of the present invention;

[0053] Figure 3 This is a schematic diagram of the installation structure of an industrial camera provided in an embodiment of the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] like Figure 1 and Figure 3 As shown, the present invention provides an automatic monitoring system for clogging of gypsum slurry spray nozzles, comprising:

[0056] An industrial camera is used to capture images of gypsum slurry being sprayed onto a hard-edged, hard-faced roller.

[0057] The image processing unit is used to process the spray image to obtain the spray area formed by the spray nozzle spraying material onto the hard-edge hard-face roller.

[0058] The blockage warning unit is used to judge the spraying area based on a preset threshold, determine the blockage status of the spraying nozzle, and issue a warning based on the blockage status.

[0059] This invention uses an industrial camera to monitor nozzle blockage by capturing images of the material being sprayed from the nozzle. This enables real-time monitoring, improving the timeliness of nozzle blockage monitoring, eliminating the delay in blockage identification, and reducing the likelihood of uneven gypsum board surfaces and material shortages leading to defective products caused by nozzle blockage.

[0060] In this invention, the real-time acquisition of industrial cameras generates a large amount of image data. To improve the targeting of blockage monitoring and avoid redundant image data processing, this invention first performs discrete calculations on the spray images acquired in real-time by the industrial camera and the spray images acquired in the past by the industrial camera. This determines whether the spray image acquired in real-time (at the current moment) by the industrial camera has changed compared to the previous spray images. If the discreteness is high, it indicates that the spray image acquired in real-time has changed compared to the spray image at the previous moment. Correspondingly, it indicates that the current spray nozzle is more likely to be blocked. This means that the spray image acquired in real-time (at the current moment) by the industrial camera is of research significance (or has image processing value) for judging the blockage of the spray nozzle and is an image of interest for monitoring the blockage of the spray nozzle.

[0061] Therefore, this invention first uses discrete computation to filter the real-time acquired spray images, selecting images of interest for monitoring the clogging of the spray nozzles for image recognition processing. Real-time acquired spray images corresponding to those with a low probability of clogging, or those that are not currently clogged, are selected. In practical terms, the image processing process for these images is redundant, avoiding the need for image processing for all real-time acquired spray images. This reduces the data processing required for real-time judgment of spray nozzle clogging, capturing only valuable images of interest, saving computational resources and improving system efficiency.

[0062] Furthermore, after acquiring valuable images of interest, the present invention extracts the spray area through image recognition and combines it with threshold judgment to accurately grasp the clogging status of the spray nozzle, thereby improving the monitoring accuracy on the basis of the above-mentioned high-efficiency monitoring.

[0063] Furthermore, to further avoid redundancy in image acquisition and adapt to the application scenario of nozzle blockage determination, the present invention optimizes the parameters of the industrial camera. This allows the exposure time and frame rate of the industrial camera to acquire nozzle images with the minimum amount of nozzle images while achieving optimal accuracy in blockage identification. Specifically, as follows:

[0064] The optimal values ​​for exposure time and frame rate for industrial cameras to capture sprayed material images are determined by a pre-established optimal parameter setting model;

[0065] The optimal parameter setting model is:

[0066] KPI = BP(ET, FPS);

[0067] In the formula, ET is the exposure time, FPS is the frame rate, KPI is the accuracy of clogging identification based on the sprayed material images acquired by the industrial camera with ET and FPS, and BP is the BP neural network.

[0068] In this experiment, the spray nozzles of the gypsum slurry were artificially blocked to serve as the test spray nozzles.

[0069] Multiple exposure times and frame rates were selected as experimental exposure times and frame rates.

[0070] The industrial camera was controlled to continuously acquire images of the material spraying from the experimental nozzle onto the hard-edged hard-faced roller at each experimental exposure time and experimental frame rate, which were used as experimental image sequences.

[0071] The spray area is extracted from each spray image in the experimental image sequence, and the spray area is judged based on a preset threshold to determine the clogging status of the spray nozzle in the spray image, thus obtaining the experimental image spray nozzle clogging status sequence.

[0072] The proportion of images that are consistent with artificial blockage states in the sequence of blockage states at the nozzle of the experimental images is statistically analyzed, and this proportion is used as the accuracy of blockage identification.

[0073] Using each experimental exposure time and frame rate, along with the corresponding blockage identification accuracy, as training samples, a BP neural network is trained based on these training samples to construct an optimal parameter setting model that maps experimental exposure time and frame rate to the blockage identification accuracy.

[0074] The optimal parameter setting model is used to calculate each selectable value of the exposure time and frame rate of the industrial camera, and the accuracy of blockage identification corresponding to each selectable value of the exposure time and frame rate of the industrial camera is obtained.

[0075] The optimal exposure time and frame rate for acquiring sprayed material images by the industrial camera are selected based on the exposure time and frame rate corresponding to the highest accuracy of blockage identification.

[0076] This invention provides an example of optimal parameter settings for an industrial camera, as follows:

[0077] First, prepare a set of experimental exposure times (ET1, ET2, ET3) and experimental frame rates (FPS1, FPS2, FPS3). For the sake of simplicity, a small amount of data is selected here. In actual use, the selection will be based on the actual scenario.

[0078] The industrial camera is controlled to continuously acquire images G1 of the material spraying from the experimental nozzle onto the hard-edged, hard-faced roller at (ET1,FPS1), (ET1,FPS2), (ET1,FPS3), (ET2,FPS1), ..., (ET3,FPS3). (ET1,FPS1) G2 (ET1,FPS1) G3 (ET1,FPS1) G3 (ET1,FPS1) , ..., G10 (ET1,FPS1) G1 (ET1,FPS1) G2 (ET1,FPS2) G3 (ET1,FPS2) G3 (ET1,FPS2) , ..., G10 (ET1,FPS2) ; as the experimental image sequence {G1 (ET1,FPS1) G2 (ET1,FPS1) G3 (ET1,FPS1) G3 (ET1,FPS1) , ..., G10 (ET1,FPS1)}, {G1 (ET1,FPS1) G2 (ET1,FPS2) G3 (ET1,FPS2) G3 (ET1,FPS2) , ..., G10 (ET1,FPS2)} where, {G1 (ET1,FPS1) G2 (ET1,FPS1) G3 (ET1,FPS1)G3 (ET1,FPS1) , ..., G10 (ET1,FPS1) To control the industrial camera, 10 images of material being sprayed from the experimental nozzle onto the hard-edged, hard-faced roller are captured at (ET1, FPS1). {G1 (ET1,FPS1) G2 (ET1,FPS2) G3 (ET1,FPS2) G3 (ET1,FPS2) , ..., G10 (ET1,FPS2) To control the industrial camera, 10 spray images were formed by the experimental nozzle spraying material onto the hard-edged hard-faced roller at (ET1, FPS2), and so on, 10 spray images were formed at (ET1, FPS3), (ET2, FPS1), ..., (ET3, FPS3).

[0079] Extract the spray area {S1} from each spray image in the experimental image sequence. (ET1,FPS1) S2 (ET1,FPS1) S3 (ET1,FPS1) S3 (ET1,FPS1) S10 (ET1,FPS1)}, and the spray area {G1 (ET1,FPS1) G2 (ET1,FPS1) G3 (ET1,FPS1) G3 (ET1,FPS1) , ..., G10 (ET1,FPS1) The values ​​are compared one by one with the preset threshold S0. When the value is less than S0, it is determined that the nozzle in the spray image is blocked. When the value is greater than or equal to S0, it is determined that the nozzle in the spray image is not blocked. The resulting sequence of nozzle blockage states in the experimental image is {blocked state, not blocked state, blocked state, blocked state, ..., not blocked state}. The sequence of nozzle blockage states in the experimental image is formed by (ET1,FPS2), (ET1,FPS3), (ET2,FPS1), ..., (ET3,FPS3).

[0080] In the experimental image sequence of nozzle blockage state {blockage state, non-blockage state, blockage state, blockage state, ..., non-blockage state} corresponding to (ET1,FPS1), the proportion of images that are consistent with the artificial blockage state is counted as the blockage recognition accuracy. For example, if there are 7 blockage states, the proportion of the total 10 blockage states is 7 / 10, then the recognition accuracy is 7 / 10. Similarly, the blockage recognition accuracy of (ET1,FPS2), (ET1,FPS3), (ET2,FPS1), ..., (ET3,FPS3) is, for example, 5 / 10, 7 / 10, 9 / 10, ..., 3 / 10 respectively.

[0081] At this point, (ET1,FPS1), (ET1,FPS2), (ET1,FPS3), (ET2,FPS1), ..., (ET3,FPS3) are combined with 7 / 10, 5 / 10, 7 / 10, 9 / 10, ..., 3 / 10 to form training samples to train the BP neural network and construct an optimal parameter setting model that maps experimental exposure time and experimental frame rate to the accuracy of blockage recognition.

[0082] Finally, using the optimal parameter setting model KPI=BP(ET,FPS), the blockage identification accuracy corresponding to all selectable values ​​in the exposure time (ET1,ET2,ET3,ET4,…,ETn) and frame rate (FPS1,FPS2,FPS3,FPS4,…,FPSn) is calculated one by one. When the highest blockage identification accuracy data is unique, the exposure time and frame rate selectable values ​​corresponding to the highest blockage identification accuracy are taken as the optimal values ​​of exposure time and frame rate for industrial camera to acquire sprayed material images. When the highest blockage identification accuracy data is not unique, the exposure time and lowest frame rate selectable values ​​corresponding to the highest blockage identification accuracy are taken as the optimal values ​​of exposure time and frame rate for industrial camera to acquire sprayed material images.

[0083] The steps by which the image processing unit determines the spray area include:

[0084] Discreteness calculations are performed on the spray images captured in real time by the industrial camera and the spray images captured in history by the industrial camera in order to filter out the images of interest for blockage identification from the spray images captured in real time.

[0085] Extract the spray area from the image of interest to obtain the spray area used for clogging identification.

[0086] The steps by which the image processing unit selects images of interest for clogging identification from the real-time acquired spray images include:

[0087] The discretization of the real-time spray images captured by the industrial camera and the historical spray images captured by the industrial camera is calculated. The discretization is as follows:

[0088]

[0089] In the formula, L represents discreteness, G is the pixel matrix of the real-time acquired sprayed material image, and G k Let |GG| be the pixel matrix of the k-th historically acquired spray image. k |For G and G k The Euclidean distance between them, where m is the total number of historically acquired spray images, and k is the counting variable;

[0090] The discreteness is compared with a discreteness threshold, where,

[0091] When the discreteness is greater than or equal to the discreteness threshold, the real-time acquired spray image is marked as the image of interest for clogging identification.

[0092] When the discreteness is less than the discreteness threshold, the real-time acquired spray images are marked as non-interest images for clogging identification.

[0093] In this invention, the real-time acquisition of industrial cameras generates a large amount of image data. To improve the targeting of blockage monitoring and avoid redundant image data processing, this invention first performs discrete calculations on the spray images acquired in real-time by the industrial camera and the spray images acquired in the past by the industrial camera. This determines whether the spray image acquired in real-time (at the current moment) by the industrial camera has changed compared to the previous spray images. If the discreteness is high, it indicates that the spray image acquired in real-time has changed compared to the spray image at the previous moment. Correspondingly, it indicates that the current spray nozzle is more likely to be blocked. This means that the spray image acquired in real-time (at the current moment) by the industrial camera is of research significance (or has image processing value) for judging the blockage of the spray nozzle and is an image of interest for monitoring the blockage of the spray nozzle.

[0094] Therefore, this invention first uses discrete computation to filter the real-time acquired spray images, selecting images of interest for monitoring the clogging of the spray nozzles for image recognition processing. Real-time acquired spray images corresponding to those with a low probability of clogging, or those that are not currently clogged, are selected. In practical terms, the image processing process for these images is redundant, avoiding the need for image processing for all real-time acquired spray images. This reduces the data processing required for real-time judgment of spray nozzle clogging, capturing only valuable images of interest, saving computational resources and improving system efficiency.

[0095] Furthermore, after acquiring valuable images of interest, this invention extracts the spray area through image recognition and combines it with threshold judgment to accurately determine the clogging status of the spray nozzle. Building upon the aforementioned high-efficiency monitoring, this invention improves monitoring accuracy, as detailed below:

[0096] The steps by which the image processing unit extracts the spray area from the image of interest include:

[0097] The image of interest is segmented by thresholding to obtain the image region in the image of interest used to calculate the spray area;

[0098] The number of pixels in the image region of interest used to measure the spray area is counted to obtain the spray area used for clogging identification.

[0099] The steps of extracting the spray area in each spray image of the experimental image sequence, as well as the spray area in the spray image formed when the spray nozzle of the gypsum slurry is in a non-blocked state, are consistent with the steps of extracting the spray area in the image of interest. Only the topic of the steps needs to be changed from the image of interest to the spray image.

[0100] The steps taken by the congestion warning unit to determine the congestion status include:

[0101] The spray area used for clogging identification is compared with a preset threshold, where,

[0102] If the spray area used for blockage identification is within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be blocked.

[0103] If the spray area used for clogging identification is not within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be in a non-clogging state.

[0104] The blockage warning unit will issue a warning when the gypsum slurry spray nozzle is blocked.

[0105] The industrial camera, image processing unit, and congestion warning unit are sequentially connected for communication.

[0106] The industrial camera transmits the captured images of the sprayed material to the image processing unit;

[0107] The image processing unit transmits the processed spray area to the blockage warning unit.

[0108] The preset threshold is the spray area in the spray image formed when the gypsum slurry nozzle is in a non-blocked state.

[0109] This invention provides an embodiment for determining blockage status:

[0110] When there are no previously acquired spray images, i.e., the industrial camera captures the first spray image, the spray area is directly extracted from the spray image and compared with a preset threshold to determine whether the current spray nozzle is blocked. When the current spray nozzle is determined to be blocked, an early warning is issued to remind the operator to carry out maintenance.

[0111] Then, after the operator clears the blockage in the spray nozzle, the industrial camera continues to acquire spray images. At this time, the acquired spray images can be used as historical spray images. The spray images acquired from the next moment will be discretely calculated with the historical spray images to obtain the image of interest, and the spray area will be extracted from the image of interest.

[0112] If the spray area used for blockage identification is within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be blocked.

[0113] If the spray area used for clogging identification is not within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be in a non-clogging state.

[0114] By continuing the above process, real-time monitoring of the gypsum slurry spray nozzle can be achieved.

[0115] This invention provides another embodiment for determining blockage status:

[0116] When there are no previously acquired spray images, i.e., the industrial camera captures the first spray image, the spray area is directly extracted from the spray image and compared with a preset threshold to determine whether the current spray nozzle is blocked. If the current spray nozzle is not blocked, the first spray image is used as the previously acquired spray image.

[0117] Then, the industrial camera continues to acquire spray images. At this time, the spray images acquired in real time by the industrial camera are compared with the historically acquired spray images to perform discrete calculations to obtain the image of interest, and the spray area is extracted from the image of interest.

[0118] If the spray area used for blockage identification is within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be blocked.

[0119] If the spray area used for clogging identification is not within the preset threshold range, the spray nozzle of the gypsum slurry is determined to be in a non-clogging state.

[0120] By continuing the above process, real-time monitoring of the gypsum slurry spray nozzle can be achieved.

[0121] like Figure 2 As shown, this invention provides an automatic monitoring method for gypsum slurry nozzle blockage, applied to an automatic monitoring system for gypsum slurry nozzle blockage. The method includes the following steps:

[0122] Images of gypsum slurry being sprayed onto a hard-edged, hard-faced roller are captured.

[0123] Image processing is performed on the spray image to obtain the spray area formed by the spray nozzle spraying material onto the hard-edged hard-faced roller;

[0124] The spray area is judged based on a preset threshold to determine the blockage status of the spray nozzle, and an early warning is issued based on the blockage status.

[0125] This invention utilizes an industrial camera to detect the area of ​​the sprayed material, automatically judging the uniformity of the sprayed material on hard edges and surfaces, thereby determining the blockage status of the spray nozzle, and simultaneously triggering an automatic alarm to alert the operator of the nozzle blockage, thus achieving automatic and accurate blockage monitoring.

[0126] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. An automatic monitoring system for gypsum slurry spray port blockage, characterized in that: include: Industrial camera, used to capture the spraying image formed by the spraying port of gypsum slurry spraying on the hard-edged and hard-surfaced roller; An image processing unit is used to perform image processing on the spraying image to obtain a spraying area formed by the spraying port on the hard-edged and hard-surface roller; The blockage warning unit is used to judge the spraying area based on a preset threshold value, determine the blockage state of the spraying port, and issue a warning according to the blockage state.

2. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 1, characterized in that: The optimal values ​​of the exposure time and frame rate of the industrial camera for collecting the spraying image are determined by a pre-established optimal parameter setting model; The optimal parameter setting model is: KPI = BP (ET, FPS); In the formula, ET is the exposure time, FPS is the frame rate, KPI is the accuracy of blockage recognition based on the spraying image collected by the industrial camera with ET and FPS, and BP is the BP neural network; Among them, the injection port of gypsum slurry was artificially blocked to serve as the experimental injection port; Select multiple exposure times and frame rates as experimental exposure times and experimental frame rates; The industrial camera is controlled to continuously collect spray images formed by the spray nozzle spraying on the hard-edged and hard-surface roller at each experimental exposure time and experimental frame rate as an experimental image sequence; Extracting a spraying area from each spraying image in the experimental image sequence, and judging the spraying area based on a preset threshold, determining a blocking state of the spraying port in the spraying image, and obtaining a spraying port blocking state sequence of the experimental images; In the experimental image sequence of the nozzle blockage state, the proportion of images consistent with the artificial blockage state is counted as the accuracy of blockage recognition; Each experimental exposure time and experimental frame rate and the blockage recognition accuracy corresponding to each experimental exposure time and experimental frame rate are used as training samples, and the BP neural network is used for training based on the training samples to construct an optimal parameter setting model mapping between the experimental exposure time and the experimental frame rate and the blockage recognition accuracy; The optimal parameter setting model is used to measure each optional value of the exposure time and frame rate of the industrial camera one by one, and the accuracy of blockage recognition corresponding to each optional value of the exposure time and frame rate of the industrial camera is obtained; The optional values ​​of exposure time and frame rate corresponding to the highest blockage recognition accuracy are selected as the optimal values ​​of exposure time and frame rate for the industrial camera to collect the spraying image.

3. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 2, characterized in that: The step of determining the spraying area by the image processing unit comprises: The material spraying images acquired in real time by the industrial camera and the material spraying images acquired in history by the industrial camera are discretely calculated, so as to screen out images of interest for blockage identification from the material spraying images acquired in real time; The spraying area is extracted from the image of interest to obtain the spraying area for blockage identification.

4. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 3, characterized in that: The step of the image processing unit screening out an image of interest for blockage identification from the material spraying image collected in real time comprises: The discreteness of the spraying image collected by the industrial camera in real time and the spraying image collected by the industrial camera in history is calculated, and the discreteness is: In the formula, L is discreteness, G is the pixel matrix of the spray image collected in real time, and G k is the pixel matrix of the kth historically collected spraying image, |GG k | for G and G k The Euclidean distance between them, m is the total number of historically collected spray images, and k is a counting variable; The discreteness is compared to a discreteness threshold, where When the discreteness is greater than or equal to the discreteness threshold, the material spraying image acquired in real time is marked as an image of interest for blockage identification; When the discreteness is less than the discreteness threshold, the material spraying image acquired in real time is marked as an image of no interest for blockage identification.

5. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 4, characterized in that: The step of extracting the spraying area from the image of interest by the image processing unit comprises: Performing threshold segmentation on the image of interest to obtain an image region in the image of interest for measuring the spraying area; The number of pixels in the image region used for measuring the spraying area in the image of interest is counted to obtain the spraying area used for clogging identification.

6. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 5, characterized in that: The step of determining the congestion status by the congestion warning unit includes: The spray area used for blockage identification is compared with a preset threshold, where: When the spraying area used for blockage identification is within a preset threshold range, it is determined that the spraying port of the gypsum slurry is in a blocked state; When the spraying area used for blockage identification is not within the preset threshold range, it is determined that the spraying port of the gypsum slurry is in a non-blocked state.

7. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 6, characterized in that: The blockage warning unit performs a warning operation when the spraying port of the gypsum slurry is in a blocked state.

8. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 7, characterized in that: The industrial camera, the image processing unit and the jam warning unit are sequentially connected in communication, wherein: The industrial camera transmits the collected spraying image to the image processing unit; The image processing unit transmits the spraying area obtained through processing to the clogging warning unit.

9. The automatic monitoring system for clogging of a gypsum slurry spraying port according to claim 8, characterized in that: The preset threshold is a spraying area in the spraying image formed when the spraying port of the gypsum slurry is in a non-blocked state.

10. An automatic monitoring method for clogging of a gypsum slurry spraying port, characterized in that: An automatic monitoring system for clogging of a gypsum slurry spraying port as described in any one of claims 1 to 9, the method comprising the following steps: Collecting the spraying image formed by the spraying port of the gypsum slurry spraying on the hard-edged and hard-surface roller; Performing image processing on the spraying image to obtain a spraying area formed by the spraying port on the hard-edged and hard-surface roller; The spraying area is judged based on a preset threshold value, the blockage state of the spraying port is determined, and an early warning is issued according to the blockage state.

Images