Ceramic bathroom intelligent production line multi-view remote monitoring system and method

Abstract

The invention provides a ceramic bathroom intelligent production line multi-view remote monitoring system and method, and the method comprises the steps: carrying out the object calibration of the respective dynamic views of a pouring molding region and a fettling region of a production line, recognizing and comparing the pouring molding actual condition and the fettling actual condition, determining the abnormal condition of pouring molding, carrying out the correlation comparison of pouring molding and fettling, and carrying out the remote monitoring of the pouring molding region and the fettling region. Abnormal conditions existing in the pouring forming procedure are accurately determined according to the implementation condition of the fettling procedure, the pouring forming operation is adjusted in a targeted mode, and the blank appearance problem caused by implementation of the pouring forming procedure is reduced; and the dynamic view of the drying section of the production line is identified to obtain the surface morphology of the green body subjected to the drying process, so that the glaze spraying section of the production line is subjected to visual navigation action adjustment, and the execution accuracy and efficiency of the processes such as pouring forming, fettling and glaze spraying on the production line are improved through multi-view identification monitoring; the overall production efficiency of a production line is improved; and the defective product forming probability is reduced.

Description

Technical Field

[0001] This invention relates to the field of sanitary ware product manufacturing, and more particularly to a multi-view remote monitoring system and method for an intelligent production line of ceramic sanitary ware. Background Technology

[0002] The production of ceramic sanitary ware products mainly includes processes such as casting, trimming, drying, glazing, high-temperature firing, and sintering. To improve production efficiency, these processes are all located on the same production line, ensuring that products are accurately and quickly transported to the next process after completion, achieving efficient seamless operation. The trimming process corrects the shape of the blanks obtained from the casting process, improving their regularity and surface smoothness, effectively reducing the defect rate. Considering the influence of factors such as the water content and viscosity of the ceramic slurry on the casting process, the shape of the blanks obtained from each casting process varies significantly, leading to significant differences in the specific implementation process of the subsequent trimming process, and resulting in longer processing times. If the casting and trimming processes on the production line are not adjusted in a coordinated manner, their efficiency cannot be improved, leading to excessive time consumption and congestion on the production line during these processes, idle time in other downstream processes, and other problems, affecting the overall production efficiency of the production line and increasing the probability of defective products. Summary of the Invention

[0003] The purpose of this invention is to provide a multi-view remote monitoring system and method for an intelligent production line for ceramic sanitary ware. This system calibrates the dynamic views of the casting and molding section and the trimming section of the production line, identifies and compares the actual casting and molding and trimming conditions, determines abnormalities in the casting and molding process, and correlates and compares the casting and molding and trimming processes. It accurately identifies abnormalities in the casting and molding process based on the implementation of the trimming process, and makes targeted adjustments to the casting and molding operation to reduce problems with the shape of the blank caused by the casting and molding process itself. Furthermore, it identifies the dynamic view of the drying section of the production line to obtain the surface morphology of the blank after the drying process, and uses this to visually guide the adjustment of the glazing section of the production line. Through multi-view recognition and monitoring, the system improves the accuracy and efficiency of the casting, molding, trimming, and glazing processes on the production line, thereby increasing the overall production efficiency of the production line and reducing the probability of defective products.

[0004] This invention is achieved through the following technical solution:

[0005] A multi-view remote monitoring system for an intelligent production line of ceramic sanitary ware includes:

[0006] The view acquisition module is used to acquire dynamic views of the casting and forming area and the trimming area of ​​the production line.

[0007] The first view recognition module is used to perform object calibration on the dynamic view and to identify the casting and molding process and the trimming process from the dynamic view.

[0008] The molding anomaly detection module is used to compare the actual casting and molding process with the actual trimming process to determine the abnormal casting and molding situation.

[0009] The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of the casting abnormality.

[0010] The second view recognition module is used to recognize the dynamic view of the drying zone of the production line and obtain the surface morphology of the blank after the drying process is completed.

[0011] The glazing adjustment module is used to perform visual navigation adjustments on the glazing section of the production line based on the surface morphology of the blank.

[0012] Optionally, the view acquisition module is used to acquire dynamic views of the casting and molding section and the trimming section of the production line, including:

[0013] Simultaneously film the casting and molding section and the trimming section of the production line to obtain a first dynamic view and a second dynamic view corresponding to the casting and molding section and the trimming section, respectively, and mark the time axis interval of the same object in the first dynamic view and the second dynamic view; based on the time axis interval of the appearance, the first dynamic view and the second dynamic view are respectively divided into a number of first dynamic sub-views and a number of second dynamic sub-views.

[0014] The first view recognition module is used to perform object calibration on the dynamic view, and to identify the casting and molding process and the trimming process from the dynamic view, including:

[0015] Identify and match the first dynamic sub-view and the second dynamic sub-view of the same object to obtain the casting and molding process and the trimming process of the same object; wherein, the casting and molding process includes the casting and molding shape features of the same object; and the trimming process includes the trimming shape features of the same object.

[0016] Optionally, the first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including:

[0017] Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view;

[0018] Extract the total number of frames from the first dynamic view and the second dynamic view;

[0019] The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view.

[0020] The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula:

[0021]

[0022] Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; J represents the number of feature points contained in the first dynamic view and the second dynamic view (i.e., the number of features that need to be identified from the first dynamic view and the second dynamic view).

[0023] The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view.

[0024] The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula:

[0025]

[0026] Where K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the corresponding casting and shaping interval and the trimming interval in the first dynamic view and the second dynamic view; T p This indicates the frame length corresponding to the first dynamic view and the second dynamic view; x g The index represents the key process stage index corresponding to the normalized first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5;

[0027] The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

[0028] Optionally, the molding anomaly determination module is used to compare the actual casting and molding situation with the actual trimming situation to determine the casting and molding anomaly, including:

[0029] By comparing the actual casting and molding process and the actual trimming process of the same object, the shape change characteristics of the same object during the process from the casting and molding section to the trimming section are determined; wherein, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and molding process to the completion of the trimming process; based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and molding process is determined.

[0030] The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of the casting abnormality, including:

[0031] The spatial distribution of abnormal shape deviations of all objects is compared to obtain the reproduction characteristics of the abnormal spatial distribution of shape deviations after the casting and molding process is completed, thereby identifying the erroneous steps in the casting and molding process; based on the erroneous steps, the operating parameters of the casting and molding process are adjusted.

[0032] Optionally, the second view recognition module is used to recognize dynamic views of the drying zone of the production line to obtain the surface morphology of the blank after the drying process, including:

[0033] Identify the pixel texture features of the dynamic view of the drying section of the production line, transform the pixel texture features, and obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width.

[0034] The glazing adjustment module is used to perform visual navigation adjustments to the glazing area of ​​the production line based on the surface morphology of the blank, including:

[0035] Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This is used to provide visual navigation for the glazing process in the glazing section of the production line, and to adjust the amount and direction of glaze coating on the surface of the blank.

[0036] A multi-view remote monitoring method for an intelligent production line for ceramic sanitary ware includes:

[0037] Obtain dynamic views of the casting and molding section and the trimming section of the production line, perform object calibration on the dynamic views, and identify the actual casting and molding situation and the actual trimming situation from the dynamic views.

[0038] By comparing the actual casting and molding process with the actual trimming process, abnormal casting and molding conditions are identified; based on the recurrence characteristics of the abnormal casting and molding conditions, the casting and molding operation is adjusted.

[0039] The dynamic view of the drying zone of the production line is identified to obtain the surface morphology of the blank after the drying process is completed; based on the surface morphology of the blank, the glazing zone of the production line is adjusted by visual navigation.

[0040] Optionally, dynamic views of the casting and molding section and the trimming section of the production line are obtained, and object calibration is performed on the dynamic views to identify the actual casting and molding situation and the actual trimming situation from the dynamic views, including:

[0041] Simultaneously film the casting and molding section and the trimming section of the production line to obtain a first dynamic view and a second dynamic view corresponding to the casting and molding section and the trimming section, respectively, and mark the time axis interval of the same object in the first dynamic view and the second dynamic view; based on the time axis interval of the appearance, the first dynamic view and the second dynamic view are respectively divided into a number of first dynamic sub-views and a number of second dynamic sub-views.

[0042] Identify and match the first dynamic sub-view and the second dynamic sub-view of the same object to obtain the casting and molding process and the trimming process of the same object; wherein, the casting and molding process includes the casting and molding shape features of the same object; and the trimming process includes the trimming shape features of the same object.

[0043] Optionally, the first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including:

[0044] Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view;

[0045] Extract the total number of frames from the first dynamic view and the second dynamic view;

[0046] The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view.

[0047] The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula:

[0048]

[0049] Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; J represents the number of feature points contained in the first dynamic view and the second dynamic view (i.e., the number of features that need to be identified from the first dynamic view and the second dynamic view).

[0050] The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view.

[0051] The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula:

[0052]

[0053] Where K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the corresponding casting and shaping interval and the trimming interval in the first dynamic view and the second dynamic view; T p This indicates the frame length corresponding to the first dynamic view and the second dynamic view; x g The index represents the key process stage index corresponding to the normalized first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5;

[0054] The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

[0055] Optionally, by comparing the actual casting and molding process with the actual trimming process, abnormal casting conditions can be identified; based on the recurrence characteristics of the abnormal casting conditions, the casting operation can be adjusted, including:

[0056] By comparing the actual casting and molding process and the actual trimming process of the same object, the shape change characteristics of the same object during the process from the casting and molding section to the trimming section are determined; wherein, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and molding process to the completion of the trimming process; based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and molding process is determined.

[0057] The spatial distribution of abnormal shape deviations of all objects is compared to obtain the reproduction characteristics of the abnormal spatial distribution of shape deviations after the casting and molding process is completed, thereby identifying the erroneous steps in the casting and molding process; based on the erroneous steps, the operating parameters of the casting and molding process are adjusted.

[0058] Optionally, a dynamic view of the drying zone of the production line is identified to obtain the surface morphology of the blank after the drying process; based on the surface morphology of the blank, visual navigation adjustments are made to the glazing zone of the production line, including:

[0059] Identify the pixel texture features of the dynamic view of the drying section of the production line, transform the pixel texture features, and obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width.

[0060] Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This is used to provide visual navigation for the glazing process in the glazing section of the production line, and to adjust the amount and direction of glaze coating on the surface of the blank.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] This application provides a multi-view remote monitoring system and method for an intelligent ceramic sanitary ware production line. It performs object labeling on the dynamic views of the casting and molding zone and the trimming zone of the production line, identifies and compares the actual casting and molding and trimming situations, determines abnormalities in the casting and molding process, and correlates and compares the casting and molding and trimming processes. Based on the implementation of the trimming process, it accurately identifies abnormalities in the casting and molding process, and makes targeted adjustments to the casting and molding operation to reduce problems with the shape of the blank caused by the casting and molding process itself. It also identifies the dynamic view of the drying zone of the production line to obtain the surface morphology of the blank after the drying process, and uses this to visually guide the adjustment of the glazing zone of the production line. Through multi-view recognition and monitoring, it improves the accuracy and efficiency of the casting, molding, trimming, and glazing processes on the production line, thereby increasing the overall production efficiency of the production line and reducing the probability of defective products. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention or 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0064] Figure 1This is a structural schematic diagram of a multi-view remote monitoring system for an intelligent production line of ceramic sanitary ware provided by the present invention.

[0065] Figure 2 This is a flowchart illustrating a multi-view remote monitoring method for an intelligent production line of ceramic sanitary ware provided by the present invention. Detailed Implementation

[0066] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, it should be noted that, for ease of description, only the parts relevant to this application are shown in the accompanying drawings, not the entire structure. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this application.

[0067] The terms “comprising” and “having”, and any variations thereof, used in this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0068] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0069] Please see Figure 1 As shown in the figure, an embodiment of this application provides a multi-view remote monitoring system for an intelligent ceramic sanitary ware production line. This multi-view remote monitoring system for an intelligent ceramic sanitary ware production line includes:

[0070] The view acquisition module is used to acquire dynamic views of the casting and forming area and the trimming area of ​​the production line.

[0071] The first view recognition module is used to perform object calibration on the dynamic view and to identify the actual casting and molding process and the actual trimming process from the dynamic view.

[0072] The molding anomaly detection module is used to compare the actual casting and molding process with the actual trimming process to determine the abnormal casting and molding situation.

[0073] The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of abnormal casting conditions.

[0074] The second view recognition module is used to recognize the dynamic view of the drying zone of the production line and obtain the surface morphology of the blank after the drying process is completed.

[0075] The glazing adjustment module is used to visually guide and adjust the glazing area of ​​the production line according to the surface morphology of the blank.

[0076] The beneficial effects of the above embodiments are that the multi-view remote monitoring system for the intelligent ceramic sanitary ware production line calibrates the dynamic views of the casting and molding section and the trimming section of the production line, identifies and compares the actual casting and molding situation and the actual trimming situation, determines abnormal situations in casting and molding, and correlates and compares casting and molding with trimming. It accurately determines the abnormal situations in the casting and molding process from the implementation of the trimming process, and makes targeted adjustments to the casting and molding operation to reduce the shape problems of the blank caused by the implementation of the casting and molding process itself. It also identifies the dynamic view of the drying section of the production line to obtain the surface morphology of the blank after the drying process, and uses this to make visual navigation adjustments to the glazing section of the production line. Through multi-view recognition and monitoring, it improves the accuracy and efficiency of the casting, molding, trimming and glazing processes on the production line, improves the overall production efficiency of the production line and reduces the probability of defective products.

[0077] In another embodiment, the view acquisition module is used to acquire dynamic views of the casting and forming section and the trimming section of the production line, including:

[0078] Simultaneously film the casting and molding section and the trimming section of the production line to obtain the first dynamic view and the second dynamic view of the corresponding casting and molding section and trimming section respectively. Mark the time axis interval of the appearance of the same object in the first dynamic view and the second dynamic view. Based on the time axis interval, the first dynamic view and the second dynamic view are divided into several first dynamic sub-views and several second dynamic sub-views respectively.

[0079] The first view recognition module is used to calibrate objects in the dynamic view, and to identify the actual casting and molding process and the actual trimming process from the dynamic view, including:

[0080] Identify and match the first dynamic subview and the second dynamic subview of the same object to obtain the casting and molding real-time and the trimming real-time of the same object; wherein, the casting and molding real-time includes the casting and molding shape features of the same object; the trimming real-time includes the trimming shape features of the same object.

[0081] The ceramic sanitary ware production line includes, in sequence, the casting and molding process, the trimming process, the drying process, the glazing process, the high-temperature sintering process, and the cooling process. Among these, the casting and molding process and the trimming process are the key processes for constructing the structural shape of ceramic sanitary ware products. Generally speaking, the casting and molding process involves pouring ceramic slurry into a mold for shaping. Considering the influence of factors such as the water content and viscosity of the ceramic slurry, the shape and size of the blank obtained after demolding will deviate from the desired shape and size. In order to avoid the deviation of the blank in shape and size affecting the usability of the final product, the trimming process is needed to correct and change the shape and size of the blank. That is, the actual implementation of the trimming process is related to the actual implementation of the casting and molding process located upstream of the production line. The deviation between the shape characteristics of the billet after the trimming process and the shape characteristics of the billet obtained in the casting process directly reflects the actual trimming position and trimming action range of the trimming process. These actual trimming position and action range directly reflect the deviation in the billet's shape structure caused by the casting process. By calibrating the actual trimming position and action range of the billets obtained after the casting and trimming processes, the current erroneous step in the casting process can be identified. Specifically, the casting and trimming sections (i.e., the sections where the casting process is implemented) and the trimming section (i.e., the sections where the trimming process is implemented) of the production line are simultaneously photographed, obtaining a first dynamic view and a second dynamic view corresponding to the two sections. The time axis intervals of the same billet in the first and second dynamic views are calibrated, thereby dividing the first and second dynamic views into several first dynamic sub-views and several second dynamic sub-views. For example, for each billet, one first dynamic sub-view and one second dynamic sub-view can be determined. These first and second dynamic sub-views record the entire implementation process of the billet in the casting and trimming processes, respectively. Then, the first dynamic sub-view and the second dynamic sub-view of the same blank are identified and matched to obtain the casting and molding process and the blank trimming process of the same object. The shape characteristics of the blank after the casting and molding process and the shape characteristics after the blank trimming process are obtained, which provides a basis for subsequent correlation and comparison of the shape changes of the blank after the above two processes are implemented.

[0082] In another embodiment, the first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including:

[0083] Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view;

[0084] Extract the total number of frames from the first dynamic view and the second dynamic view;

[0085] The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view.

[0086] The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula:

[0087]

[0088] Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; J represents the number of feature points contained in the first dynamic view and the second dynamic view (i.e., the number of features that need to be identified from the first dynamic view and the second dynamic view).

[0089] The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view.

[0090] The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula:

[0091]

[0092] Wherein, K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the casting and molding interval and the trimming interval corresponding to the first dynamic view and the second dynamic view; Tp represents the frame length corresponding to the first dynamic view and the second dynamic view; xg represents the normalized process key stage index corresponding to the first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5;

[0093] The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

[0094] First, the number of keyframes and the total number of frames are extracted from the first and second dynamic views. A key process stage index is calculated using a specific formula. This index reflects the proportion of keyframes in the total number of frames and the impact of the number of feature points on the view. Next, combining video information entropy, the duration of the casting and finishing interval, the duration of a single frame, the normalized key process stage index, and a preset adjustment coefficient, a lower limit for the number of dynamic sub-views is determined using another formula. Finally, this lower limit is used as a constraint to divide the first and second dynamic views, resulting in several first and second dynamic sub-views. Calculating the key process stage index using the number of keyframes, the total number of frames, and the number of feature points quantifies the importance of key process steps in the view, ensuring that the divided dynamic sub-views focus on the core of the process, avoiding the fragmentation of key processes, and making the sub-views more accurately reflect process changes and characteristics. Determining the lower limit for the number of dynamic sub-views based on multiple parameters such as video information entropy and process interval duration balances view complexity with process duration requirements. This approach ensures a sufficient number of subviews to meticulously represent process details while avoiding excessive data redundancy. It provides appropriate granularity for subsequent view analysis and process monitoring, improving analysis efficiency and accuracy. View division is constrained by calculated minimum values, providing clear quantitative basis for the division results and reducing uncertainty caused by manual or random division. Regardless of changes in view content or process duration, a stable and reasonable subview division scheme can be output through formulas, ensuring consistency and reliability of the analysis process and results. Reasonably divided dynamic subviews allow for targeted extraction and processing of process information from each subview, avoiding indiscriminate analysis of the entire view and reducing data processing volume. Simultaneously, it facilitates rapid location of key process stages, providing efficient data support for process optimization and quality inspection, shortening information processing time. Dividing dynamic views into subviews that conform to process characteristics allows for clearer observation of changes in process parameters and feature evolution at each stage, enabling timely detection of process anomalies or quality defects. For example, in the casting and finishing stages, comparative analysis of subviews enables refined monitoring and control of the process.

[0095] In another embodiment, the forming anomaly determination module is used to compare the actual casting and forming situation with the actual blank trimming situation to determine the casting and forming anomaly, including:

[0096] By comparing the actual casting and finishing processes of the same object, the shape change characteristics of the same object during the process from the casting and finishing zone to the finishing zone are determined. Among them, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and finishing process to the completion of the finishing process. Based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and finishing process is determined.

[0097] The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of casting abnormalities, including:

[0098] By comparing the spatial distribution of abnormal shape deviations of all objects, the recurrence characteristics of the spatial distribution of abnormal shape deviations after the casting and molding process are obtained, thereby identifying the erroneous steps in the casting and molding process; and adjusting the operating parameters of the casting and molding process based on the erroneous steps.

[0099] For the same billet, there are changes in its shape after the casting process and after the trimming process. By comparing the actual casting and trimming of the same object, the amount of spatial change in the shape contour of the same billet from the completion of the casting process to the completion of the trimming process is determined. This amount of spatial change in the shape contour is formed by the trimming operation of the billet, that is, the amount of spatial change in the shape contour directly reflects the deviation of the billet's shape structure caused by the casting process. Therefore, based on the amount of spatial change in the shape contour, the abnormal spatial distribution of the shape deviation after the completion of the casting process of the same billet is determined. That is, the distribution location of the area on the billet surface where the actual dimensional contour deviation between the actual shape and the ideal shape of the same billet after the completion of the casting process exceeds the preset deviation threshold, and the actual dimensional contour deviation of the above area are determined. Then, the abnormal spatial distribution of the shape deviation of all billets is compared to obtain the recurrence characteristics of the abnormal spatial distribution location of the shape deviation after the completion of the casting process of all billets. That is, the spatial range in which the abnormal spatial distribution location of the shape deviation of all billets after the completion of the casting process repeatedly appears is obtained, thereby identifying the relevant steps in the casting process of the above spatial range as erroneous steps. Then, based on the original operating parameters in the casting and molding process mentioned above, adjust the operating parameters of the casting and molding process (such as the squeezing force of the ceramic slurry during the casting and molding process) to improve the accuracy of the green body outline after casting and demolding.

[0100] In another embodiment, the second view recognition module is used to recognize dynamic views of the drying zone of the production line to obtain the surface morphology of the blank after the drying process, including:

[0101] The pixel texture features of the dynamic view of the drying zone of the production line are identified, and the pixel texture features are transformed to obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width.

[0102] The glazing adjustment module is used to visually guide and adjust the glazing area of ​​the production line according to the surface morphology of the blank, including:

[0103] Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This allows for visual navigation of the glazing process in the glazing section of the production line, adjusting the amount and direction of glaze coating on the surface of the blank.

[0104] During the drying process of the green body, the internal moisture content decreases, leading to cracks on the surface. If the glaze does not adequately cover these cracks during the glazing process, crack defects will appear on the surface of the finished product after high-temperature sintering. Therefore, we first identify the pixel texture features of the dynamic view of the drying zone of the production line to obtain the location and width of the cracks on the green body surface after the drying process. Then, based on the location and width of the cracks, we determine the areas where the glaze coating thickness changes on the green body surface. This allows for visual navigation of the glazing process in the glazing zone of the production line, adjusting the amount and direction of glaze application on the green body surface to ensure a uniform glaze layer across the entire surface and to ensure that the glaze effectively fills and covers the crack structure on the green body surface.

[0105] Please see Figure 2 As shown in the figure, an embodiment of this application provides a multi-view remote monitoring method for an intelligent ceramic sanitary ware production line. This multi-view remote monitoring method for an intelligent ceramic sanitary ware production line includes:

[0106] Obtain dynamic views of the casting and molding section and the trimming section of the production line, perform object calibration on the dynamic views, and identify the actual casting and molding situation and the actual trimming situation from the dynamic views.

[0107] By comparing the actual casting and molding process with the actual trimming process, abnormal casting and molding conditions can be identified; based on the recurrence characteristics of these abnormalities, the casting and molding operations can be adjusted.

[0108] The dynamic view of the drying zone of the production line is identified to obtain the surface morphology of the blank after the drying process is completed; based on the surface morphology of the blank, the visual navigation action is adjusted in the glazing zone of the production line.

[0109] The beneficial effects of the above embodiments are that the multi-view remote monitoring method for the intelligent ceramic sanitary ware production line calibrates the dynamic views of the casting and molding section and the trimming section of the production line, identifies and compares the actual casting and molding situation and the actual trimming situation, determines abnormal situations in casting and molding, and correlates and compares casting and molding with trimming. It accurately determines the abnormal situations in the casting and molding process from the implementation of the trimming process, and makes targeted adjustments to the casting and molding operation to reduce the shape problems of the blank caused by the implementation of the casting and molding process itself. It also identifies the dynamic view of the drying section of the production line to obtain the surface morphology of the blank after the drying process, and uses this to make visual navigation adjustments to the glazing section of the production line. Through multi-view recognition and monitoring, it improves the accuracy and efficiency of the execution of casting, molding, trimming, glazing and other processes on the production line, improves the overall production efficiency of the production line and reduces the probability of defective products.

[0110] In another embodiment, dynamic views of the casting and molding section and the trimming section of the production line are obtained, and object calibration is performed on the dynamic views. The actual casting and molding situation and the actual trimming situation are identified from the dynamic views, including:

[0111] Simultaneously film the casting and molding section and the trimming section of the production line to obtain the first dynamic view and the second dynamic view of the corresponding casting and molding section and trimming section respectively. Mark the time axis interval of the appearance of the same object in the first dynamic view and the second dynamic view. Based on the time axis interval, the first dynamic view and the second dynamic view are divided into several first dynamic sub-views and several second dynamic sub-views respectively.

[0112] Identify and match the first dynamic subview and the second dynamic subview of the same object to obtain the casting and molding real-time and the trimming real-time of the same object; wherein, the casting and molding real-time includes the casting and molding shape features of the same object; the trimming real-time includes the trimming shape features of the same object.

[0113] The ceramic sanitary ware production line includes, in sequence, the casting and molding process, the trimming process, the drying process, the glazing process, the high-temperature sintering process, and the cooling process. Among these, the casting and molding process and the trimming process are the key processes for constructing the structural shape of ceramic sanitary ware products. Generally speaking, the casting and molding process involves pouring ceramic slurry into a mold for shaping. Considering the influence of factors such as the water content and viscosity of the ceramic slurry, the shape and size of the blank obtained after demolding will deviate from the desired shape and size. In order to avoid the deviation of the blank in shape and size affecting the usability of the final product, the trimming process is needed to correct and change the shape and size of the blank. That is, the actual implementation of the trimming process is related to the actual implementation of the casting and molding process located upstream of the production line. The deviation between the shape characteristics of the billet after the trimming process and the shape characteristics of the billet obtained in the casting process directly reflects the actual trimming position and trimming action range of the trimming process. These actual trimming position and action range directly reflect the deviation in the billet's shape structure caused by the casting process. By calibrating the actual trimming position and action range of the billets obtained after the casting and trimming processes, the current erroneous step in the casting process can be identified. Specifically, the casting and trimming sections (i.e., the sections where the casting process is implemented) and the trimming section (i.e., the sections where the trimming process is implemented) of the production line are simultaneously photographed, obtaining a first dynamic view and a second dynamic view corresponding to the two sections. The time axis intervals of the same billet in the first and second dynamic views are calibrated, thereby dividing the first and second dynamic views into several first dynamic sub-views and several second dynamic sub-views. For example, for each billet, one first dynamic sub-view and one second dynamic sub-view can be determined. These first and second dynamic sub-views record the entire implementation process of the billet in the casting and trimming processes, respectively. Then, the first dynamic sub-view and the second dynamic sub-view of the same blank are identified and matched to obtain the casting and molding process and the blank trimming process of the same object. The shape characteristics of the blank after the casting and molding process and the shape characteristics after the blank trimming process are obtained, which provides a basis for subsequent correlation and comparison of the shape changes of the blank after the above two processes are implemented.

[0114] In another embodiment, the first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including:

[0115] Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view;

[0116] Extract the total number of frames from the first dynamic view and the second dynamic view;

[0117] The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view.

[0118] The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula:

[0119]

[0120] Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; J represents the number of feature points contained in the first dynamic view and the second dynamic view (i.e., the number of features that need to be identified from the first dynamic view and the second dynamic view).

[0121] The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view.

[0122] The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula:

[0123]

[0124] Wherein, K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the casting and molding interval and the trimming interval corresponding to the first dynamic view and the second dynamic view; Tp represents the frame length corresponding to the first dynamic view and the second dynamic view; xg represents the normalized process key stage index corresponding to the first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5;

[0125] The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

[0126] First, the number of keyframes and the total number of frames are extracted from the first and second dynamic views. A key process stage index is calculated using a specific formula. This index reflects the proportion of keyframes in the total number of frames and the impact of the number of feature points on the view. Next, combining video information entropy, the duration of the casting and finishing interval, the duration of a single frame, the normalized key process stage index, and a preset adjustment coefficient, a lower limit for the number of dynamic sub-views is determined using another formula. Finally, this lower limit is used as a constraint to divide the first and second dynamic views, resulting in several first and second dynamic sub-views. Calculating the key process stage index using the number of keyframes, the total number of frames, and the number of feature points quantifies the importance of key process steps in the view, ensuring that the divided dynamic sub-views focus on the core of the process, avoiding the fragmentation of key processes, and making the sub-views more accurately reflect process changes and characteristics. Determining the lower limit for the number of dynamic sub-views based on multiple parameters such as video information entropy and process interval duration balances view complexity with process duration requirements. This approach ensures a sufficient number of subviews to meticulously represent process details while avoiding excessive data redundancy. It provides appropriate granularity for subsequent view analysis and process monitoring, improving analysis efficiency and accuracy. View division is constrained by calculated minimum values, providing clear quantitative basis for the division results and reducing uncertainty caused by manual or random division. Regardless of changes in view content or process duration, a stable and reasonable subview division scheme can be output through formulas, ensuring consistency and reliability of the analysis process and results. Reasonably divided dynamic subviews allow for targeted extraction and processing of process information from each subview, avoiding indiscriminate analysis of the entire view and reducing data processing volume. Simultaneously, it facilitates rapid location of key process stages, providing efficient data support for process optimization and quality inspection, shortening information processing time. Dividing dynamic views into subviews that conform to process characteristics allows for clearer observation of changes in process parameters and feature evolution at each stage, enabling timely detection of process anomalies or quality defects. For example, in the casting and finishing stages, comparative analysis of subviews enables refined monitoring and control of the process.

[0127] In another embodiment, by comparing the actual casting and molding process with the actual trimming process, abnormal casting and molding conditions are identified; based on the recurrence characteristics of the abnormal casting and molding conditions, the casting and molding operation is adjusted, including:

[0128] By comparing the actual casting and finishing processes of the same object, the shape change characteristics of the same object during the process from the casting and finishing zone to the finishing zone are determined. Among them, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and finishing process to the completion of the finishing process. Based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and finishing process is determined.

[0129] By comparing the spatial distribution of abnormal shape deviations of all objects, the recurrence characteristics of the spatial distribution of abnormal shape deviations after the casting and molding process are obtained, thereby identifying the erroneous steps in the casting and molding process; and adjusting the operating parameters of the casting and molding process based on the erroneous steps.

[0130] For the same billet, there are changes in its shape after the casting process and after the trimming process. By comparing the actual casting and trimming of the same object, the amount of spatial change in the shape contour of the same billet from the completion of the casting process to the completion of the trimming process is determined. This amount of spatial change in the shape contour is formed by the trimming operation of the billet, that is, the amount of spatial change in the shape contour directly reflects the deviation of the billet's shape structure caused by the casting process. Therefore, based on the amount of spatial change in the shape contour, the abnormal spatial distribution of the shape deviation after the completion of the casting process of the same billet is determined. That is, the distribution location of the area on the billet surface where the actual dimensional contour deviation between the actual shape and the ideal shape of the same billet after the completion of the casting process exceeds the preset deviation threshold, and the actual dimensional contour deviation of the above area are determined. Then, the abnormal spatial distribution of the shape deviation of all billets is compared to obtain the recurrence characteristics of the abnormal spatial distribution location of the shape deviation after the completion of the casting process of all billets. That is, the spatial range in which the abnormal spatial distribution location of the shape deviation of all billets after the completion of the casting process repeatedly appears is obtained, thereby identifying the relevant steps in the casting process of the above spatial range as erroneous steps. Then, based on the original operating parameters in the casting and molding process mentioned above, adjust the operating parameters of the casting and molding process (such as the squeezing force of the ceramic slurry during the casting and molding process) to improve the accuracy of the green body outline after casting and demolding.

[0131] In another embodiment, a dynamic view of the drying zone of the production line is identified to obtain the surface morphology of the blank after the drying process; based on the surface morphology of the blank, visual navigation adjustments are made to the glazing zone of the production line, including:

[0132] The pixel texture features of the dynamic view of the drying zone of the production line are identified, and the pixel texture features are transformed to obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width.

[0133] Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This allows for visual navigation of the glazing process in the glazing section of the production line, adjusting the amount and direction of glaze coating on the surface of the blank.

[0134] During the drying process of the green body, the internal moisture content decreases, leading to cracks on the surface. If the glaze does not adequately cover these cracks during the glazing process, crack defects will appear on the surface of the finished product after high-temperature sintering. Therefore, we first identify the pixel texture features of the dynamic view of the drying zone of the production line to obtain the location and width of the cracks on the green body surface after the drying process. Then, based on the location and width of the cracks, we determine the areas where the glaze coating thickness changes on the green body surface. This allows for visual navigation of the glazing process in the glazing zone of the production line, adjusting the amount and direction of glaze application on the green body surface to ensure a uniform glaze layer across the entire surface and to ensure that the glaze effectively fills and covers the crack structure on the green body surface.

[0135] Overall, this multi-view remote monitoring system and method for intelligent ceramic sanitary ware production lines calibrates the dynamic views of the casting and molding zone and the trimming zone of the production line, identifies and compares the actual casting and molding and trimming situations, determines abnormalities in the casting and molding process, and correlates and compares the casting and molding and trimming processes to accurately identify abnormalities in the casting and molding process based on the implementation of the trimming process. This allows for targeted adjustments to the casting and molding operations, reducing problems with the shape of the blanks caused by the casting and molding process itself. Furthermore, it identifies the dynamic view of the drying zone of the production line to obtain the surface morphology of the blanks after the drying process, thereby providing visual navigation adjustments to the glazing zone of the production line. Through multi-view recognition and monitoring, the system improves the accuracy and efficiency of the casting, molding, trimming, and glazing processes on the production line, increasing the overall production efficiency of the production line and reducing the probability of defective products.

[0136] The above is only one specific embodiment of the present invention, and any improvements made based on the concept of the present invention shall be considered within the scope of protection of the present invention.

Claims

1. A multi-view remote monitoring system for an intelligent production line of ceramic sanitary ware, characterized in that, include: The view acquisition module is used to acquire dynamic views of the casting and forming area and the trimming area of ​​the production line. The first view recognition module is used to perform object calibration on the dynamic view and to identify the casting and molding process and the trimming process from the dynamic view. The molding anomaly detection module is used to compare the actual casting and molding process with the actual trimming process to determine the abnormal casting and molding situation. The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of the casting abnormality. The second view recognition module is used to recognize the dynamic view of the drying zone of the production line and obtain the surface morphology of the blank after the drying process is completed. The glazing adjustment module is used to perform visual navigation adjustments on the glazing section of the production line based on the surface morphology of the blank.

2. The multi-view remote monitoring system for intelligent production lines of ceramic sanitary ware as described in claim 1, characterized in that: The view acquisition module is used to acquire dynamic views of the casting and molding area and the trimming area of ​​the production line, including: Simultaneously film the casting and molding section and the trimming section of the production line to obtain a first dynamic view and a second dynamic view corresponding to the casting and molding section and the trimming section, respectively, and mark the time axis interval of the same object in the first dynamic view and the second dynamic view; based on the time axis interval of the appearance, the first dynamic view and the second dynamic view are respectively divided into a number of first dynamic sub-views and a number of second dynamic sub-views. The first view recognition module is used to perform object calibration on the dynamic view, and to identify the casting and molding process and the trimming process from the dynamic view, including: Identify and match the first dynamic subview and the second dynamic subview of the same object to obtain the casting and molding process and the trimming process of the same object; wherein, the casting and molding process includes the casting and molding shape features of the same object; and the trimming process includes the trimming shape features of the same object.

3. The multi-view remote monitoring system for intelligent production lines of ceramic sanitary ware as described in claim 1, characterized in that: The first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including: Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view; Extract the total number of frames from the first dynamic view and the second dynamic view; The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view. The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula: Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; and J represents the number of feature points contained in the first dynamic view and the second dynamic view. The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view. The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula: Where K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the corresponding casting and shaping interval and the trimming interval in the first dynamic view and the second dynamic view; T p This indicates the frame length corresponding to the first dynamic view and the second dynamic view; x g The index represents the key process stage index corresponding to the normalized first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5; The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

4. The multi-view remote monitoring system for intelligent production lines of ceramic sanitary ware as described in claim 1, characterized in that: The molding anomaly determination module is used to compare the actual casting and molding process with the actual trimming process to determine abnormal casting and molding conditions, including: By comparing the actual casting and molding process and the actual trimming process of the same object, the shape change characteristics of the same object during the process from the casting and molding section to the trimming section are determined; wherein, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and molding process to the completion of the trimming process; based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and molding process is determined. The casting adjustment module is used to adjust the casting operation based on the recurrence characteristics of the casting abnormality, including: The spatial distribution of abnormal shape deviations of all objects is compared to obtain the reproduction characteristics of the abnormal spatial distribution of shape deviations after the casting and molding process is completed, thereby identifying the erroneous steps in the casting and molding process; based on the erroneous steps, the operating parameters of the casting and molding process are adjusted.

5. The multi-view remote monitoring system for intelligent production lines of ceramic sanitary ware as described in claim 1, characterized in that: The second view recognition module is used to recognize the dynamic view of the drying zone of the production line to obtain the surface morphology of the blank after the drying process, including: Identify the pixel texture features of the dynamic view of the drying section of the production line, transform the pixel texture features, and obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width. The glazing adjustment module is used to perform visual navigation adjustments to the glazing area of ​​the production line based on the surface morphology of the blank, including: Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This is used to provide visual navigation for the glazing process in the glazing section of the production line, and to adjust the amount and direction of glaze coating on the surface of the blank.

6. A multi-view remote monitoring method for an intelligent production line for ceramic sanitary ware, characterized in that, include: Obtain dynamic views of the casting and molding section and the trimming section of the production line, perform object calibration on the dynamic views, identify the actual casting and molding situation and the actual trimming situation from the dynamic views; compare the actual casting and molding situation and the actual trimming situation to determine the abnormal situation of casting and molding; adjust the casting and molding operation according to the recurrence characteristics of the abnormal situation of casting and molding. The dynamic view of the drying zone of the production line is identified to obtain the surface morphology of the blank after the drying process is completed; based on the surface morphology of the blank, the glazing zone of the production line is adjusted by visual navigation.

7. The multi-view remote monitoring method for intelligent production lines of ceramic sanitary ware as described in claim 6, characterized in that: Obtain dynamic views of the casting and molding section and the trimming section of the production line, perform object calibration on the dynamic views, and identify the actual casting and molding situation and the actual trimming situation from the dynamic views, including: Simultaneously film the casting and molding section and the trimming section of the production line to obtain a first dynamic view and a second dynamic view corresponding to the casting and molding section and the trimming section, respectively, and mark the time axis interval of the same object in the first dynamic view and the second dynamic view; based on the time axis interval of the appearance, the first dynamic view and the second dynamic view are respectively divided into a number of first dynamic sub-views and a number of second dynamic sub-views. Identify and match the first dynamic subview and the second dynamic subview of the same object to obtain the casting and molding process and the trimming process of the same object; wherein, the casting and molding process includes the casting and molding shape features of the same object; and the trimming process includes the trimming shape features of the same object.

8. The multi-view remote monitoring method for intelligent production lines of ceramic sanitary ware as described in claim 6, characterized in that: The first dynamic view and the second dynamic view are respectively divided into a plurality of first dynamic subviews and a plurality of second dynamic subviews, including: Extract the frame numbers corresponding to the keyframes in the first dynamic view and the second dynamic view; Extract the total number of frames from the first dynamic view and the second dynamic view; The process key stage index corresponding to the first dynamic view and the second dynamic view is obtained by using the number of frames corresponding to the key frames of the first dynamic view and the second dynamic view and the total number of frames in the first dynamic view and the second dynamic view. The key process stage indices corresponding to the first dynamic view and the second dynamic view are obtained using the following formula: Where x represents the process key stage index corresponding to the first dynamic view and the second dynamic view; N represents the number of frames corresponding to the key frames in the first dynamic view and the second dynamic view; M represents the total number of frames in the first dynamic view and the second dynamic view; and J represents the number of feature points contained in the first dynamic view and the second dynamic view. The lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is determined by using the process key stage index corresponding to the first dynamic view and the second dynamic view. The minimum number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view is obtained by the following formula: Where K represents the lower limit of the number of dynamic subviews that need to be divided for the first dynamic view and the second dynamic view; H represents the video information entropy corresponding to the first dynamic view and the second dynamic view; ΔT represents the existence duration of the corresponding casting and shaping interval and the trimming interval in the first dynamic view and the second dynamic view; T p This indicates the frame length corresponding to the first dynamic view and the second dynamic view; x g The index represents the key process stage index corresponding to the normalized first dynamic view and the second dynamic view; n represents the preset adjustment coefficient, and the value range of the adjustment coefficient is 3.2-4.5; The first dynamic view and the second dynamic view are divided using the lower limit of the number of dynamic subviews that need to be divided into corresponding first dynamic subviews and second dynamic subviews as constraints.

9. The multi-view remote monitoring method for intelligent production lines of ceramic sanitary ware as described in claim 6, characterized in that: By comparing the actual casting and molding process with the actual trimming process, abnormal casting conditions are identified; based on the recurrence characteristics of these abnormalities, the casting operation is adjusted, including: By comparing the actual casting and molding process and the actual trimming process of the same object, the shape change characteristics of the same object during the process from the casting and molding section to the trimming section are determined; wherein, the shape change characteristics include the amount of spatial change in the shape contour of the same object from the completion of the casting and molding process to the completion of the trimming process; based on the shape change characteristics, the abnormal spatial distribution of the shape deviation of the same object after the completion of the casting and molding process is determined. The spatial distribution of abnormal shape deviations of all objects is compared to obtain the reproduction characteristics of the abnormal spatial distribution of shape deviations after the casting and molding process is completed, thereby identifying the erroneous steps in the casting and molding process; based on the erroneous steps, the operating parameters of the casting and molding process are adjusted.

10. The multi-view remote monitoring method for intelligent production lines of ceramic sanitary ware as described in claim 6, characterized in that: Identify the dynamic view of the drying zone of the production line to obtain the surface morphology of the blank after the drying process; based on the surface morphology of the blank, perform visual navigation adjustments to the glazing zone of the production line, including: Identify the pixel texture features of the dynamic view of the drying section of the production line, transform the pixel texture features, and obtain the surface crack distribution pattern of the billet after the drying process is completed; wherein, the surface crack distribution pattern of the billet includes the crack location and crack width. Based on the distribution pattern of cracks on the surface of the blank, the area where the glaze coating thickness changes on the surface of the blank is determined. This is used to provide visual navigation for the glazing process in the glazing section of the production line, and to adjust the amount and direction of glaze coating on the surface of the blank.

Images