Traditional ceramic handicraft display system based on ai technology

Abstract

The application discloses a traditional ceramic handicraft display system based on AI technology and relates to the field of virtual reality, comprising: an analysis module, which is used for receiving curved workpiece processing and reconstruction requirements, constructing a semantic requirement model, and converting user input information into quantitative instructions; a processing module, which is used for receiving output instructions of the analysis module, synchronously starting a collection device, acquiring global point cloud data of the workpiece, and completing noise reduction, registration, and establishment of an association index between data and processing requirements; the application can accurately receive text, graphics and parameter type user input and convert them into quantitative instructions, fit traditional ceramic processing core links, dynamically adjust the angle and position during data collection, ensure that the data is comprehensive and complete, filter effective features according to process limits, strength requirements and aesthetic standards, and generate a three-dimensional model with continuous curved surfaces and consistent topology.

Description

Technical Field

[0001] This invention relates to the field of virtual reality technology, specifically to a traditional ceramic handicraft display system based on AI technology. Background Technology

[0002] Ceramic handicrafts use natural clay as raw material and are made through processes such as material selection, kneading, shaping (throwing / sculpting / printing), drying, trimming, glazing, and firing. By controlling the proportion of raw materials, shaping techniques, and kiln temperature and atmosphere, unique shapes, patterns, and textures are given to the products, giving them both practical and artistic value.

[0003] The invention patent application with application number 202510750207.9 discloses a method, system and storage medium for three-dimensional dynamic display of handicrafts. This application solves the problem that "traditional three-dimensional display technology mainly relies on static scanning and texture mapping, which can only realize the restoration of the basic geometric shape and surface color of handicrafts, and cannot simulate the key characteristics of flexible handicrafts. This results in digital exhibits showing obvious 'hardness' and losing the soft drape, natural wrinkle changes and dynamic gloss changes under light of the original handicrafts".

[0004] However, there is no existing technology for constructing 3D models of handicrafts through semantic parsing, and the existing technology has a poor level of intelligence in the process of constructing 3D models of handicrafts.

[0005] To address this, we have proposed a traditional ceramic handicraft display system based on AI technology. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a traditional ceramic handicraft display system based on AI technology, which can effectively solve the problems of the existing technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions; This invention discloses a traditional ceramic handicraft display system based on AI technology, comprising: The system comprises the following modules: a parsing module, a processing module, and a display module. The parsing module receives the machining and reconstruction requirements for curved workpieces, constructs a semantic requirement model, and converts user input information into quantitative instructions. The processing module receives the output instructions from the parsing module, synchronously starts the acquisition equipment, obtains the full-domain point cloud data of the workpiece, performs noise reduction and registration, and establishes an association index between the data and the machining requirements. The construction module extracts surface geometric features based on the preprocessed point cloud data and generates an initial 3D model of the workpiece using a built-in topology reconstruction algorithm. The fusion module calls the parsed machining parameters and embeds the cutting path, accuracy requirements, feed rate, and cutting depth information into the initial 3D model of the workpiece. The verification module detects accuracy deviations in the initial 3D model carrying the information. If the accuracy deviation exceeds a preset threshold, the initial 3D model is calibrated and the verification module is refreshed; otherwise, it is transmitted to the display module. The display module receives the model output from the verification module and forwards it to the preset machining control system and the 3D visualization platform. The parsing module is interactively connected to the processing module and the construction module via a wireless network. The construction module is interactively connected to the fusion module and the verification module via a wireless network. The fusion module and the verification module are interactively connected to the display module via a wireless network.

[0008] Furthermore, the noise reduction operation in the processing module is triggered based on the density distribution difference of the point cloud data, and noise reduction is performed by removing outliers whose dispersion exceeds a preset range. The registration operation uses the reference feature surface of the ceramic workpiece as the registration reference, and the association index is established by label mapping and hierarchical association between the preprocessed point cloud data and the process parameters in the quantization instruction.

[0009] Furthermore, when the processing module starts the acquisition device, it controls the preset acquisition device to perform collaborative acquisition along the circumference and axial direction of the ceramic workpiece. During the acquisition process, the coverage range of the point cloud data is checked in real time. When the coverage range does not meet the preset requirements, the position and acquisition angle of the acquisition device are automatically adjusted and supplementary acquisition is performed to ensure that the acquired full-domain point cloud data has no acquisition blind spots.

[0010] Furthermore, when extracting the geometric features of the curved surface, the construction module collects the contour curvature, texture concavity and convexity parameters, body wall thickness distribution, and interface transition curvature of the ceramic workpiece. The validity screening of feature parameters is performed based on the preset ceramic surface process adaptation logic, which includes the reasonable range of surface curvature, the matching ratio of the height of the decorative protrusions to the thickness of the body, and the smoothness requirements of the interface transition, thereby eliminating invalid feature parameters that exceed the process adaptation range. The topology reconstruction algorithm generates an initial 3D model of the workpiece based on the filtered effective feature parameters through topological connection of feature points, surface patch division, and smooth splicing. After the model is generated, the surface continuity of the model is checked by detecting the deviation of the included angle between adjacent surface patches and verifying the consistency of the surface tangent direction. The geometric topology consistency is checked by verifying the integrity of node connections and the absence of redundancy in the topology. After both checks pass, the initial 3D model is pushed to the fusion module. If the checks fail, the algorithm returns to the feature extraction stage to re-collect and filter geometric features.

[0011] Furthermore, during the operation phase of the fusion module, the cutting path, accuracy requirements, feed rate, and cutting depth information are constructed into a multi-level embedded structure according to the process flow of ceramic processing; Among them, the accuracy requirement serves as the basic level for embedding core constraint information into the model. The basic level transforms the accuracy standard into geometric constraints that the model can recognize through parameterized constraint formulas, providing a unified accuracy benchmark for each processing step. The cutting path, feed rate, and depth of cut information are constructed based on the accuracy constraints of the basic level to form the corresponding process level. The information in each process level establishes a mapping relationship. That is, after the cutting path is dynamically planned according to the surface geometry of the ceramic workpiece, the feed rate and depth of cut are adaptively matched according to the curvature change of the cutting path and the material removal requirements to ensure that the three work together to meet the accuracy requirements. After the information is embedded, the compatibility of information at each level is verified, including the adaptability of the cutting path and surface geometry, the rationality of the linkage between feed rate and cutting depth, and the consistency between the information at each process level and the accuracy requirements of the basic level.

[0012] Furthermore, the constraint formula is: ; in, This represents the actual combined geometric deviation under the combined effects of the cutting path, feed rate, and depth of cut. This represents the weighting coefficients indicating the influence of cutting path, feed rate, and depth of cut on accuracy deviation. This indicates the cutting path length of the actual embedded model in the fusion module. This represents the theoretically optimal cutting path length obtained based on the surface geometry, pattern distribution, and machining process optimization of the ceramic workpiece. This indicates the feed rate of the actual embedded model in the fusion module. This indicates the reference feed rate that matches the preset accuracy requirements. This represents the cutting depth of the actual embedded model in the fusion module. This represents the reference cutting depth to match the preset accuracy requirements. This indicates the preset allowable deviation threshold for accuracy; in, All are greater than zero, and their sum is 1.

[0013] Furthermore, the accuracy deviation detection of the verification module includes geometric dimension deviation, process parameter matching deviation, and surface topology deviation: Geometric dimensional deviation detection includes calculating the difference between the actual model value and the theoretical requirement value of the main body size of the ceramic workpiece, and determining the degree to which the dimensional tolerance range is met; Process parameter matching deviation detection includes the fit analysis of the cutting path, feed rate, and depth of cut embedded in the model with the process parameters in the semantic requirements model, and the rationality assessment of the collaborative matching between various process parameters; Surface topology deviation detection includes verifying the integrity of the model topology, checking the correctness of the connection relationships between topology nodes, and detecting the smoothness of the surface patch joints; When the deviation in any dimension exceeds the corresponding preset threshold, the verification module automatically calls the corresponding preset calibration algorithm based on the deviation type. Specifically, the geometric dimension deviation is corrected by using a dimension compensation calibration algorithm, which adjusts the key geometric parameters of the model to correct the deviation; the process parameter matching deviation is corrected by using a parameter adaptation adjustment algorithm, which optimizes the process parameters embedded in the model based on the standard parameters in the demand model; and the surface topology deviation is corrected by using a topology reconstruction optimization algorithm, which reorganizes and optimizes the model topology. After calibration, restart the full-dimensional accuracy deviation detection until all dimensional deviations are within the corresponding preset threshold range, obtain a qualified model, and transmit it to the display module; If the deviation requirement is still not met after the preset limit of calibration attempts, the calibration process will be terminated, and a prompt message containing the deviation type, the degree of exceeding the limit, and the calibration attempt record will be output.

[0014] Furthermore, the display module converts the verified model according to a preset standardized data format. The converted model data is then forwarded to the machining control system and the 3D visualization platform via a two-way communication link. During the forwarding process, the data transmission status is monitored in real time. When an abnormality occurs during transmission, the breakpoint resume mechanism is automatically activated, enabling the machining control system and the 3D visualization platform to obtain complete model data.

[0015] Furthermore, the parsing module is interconnected with the processing module and the construction module via a wireless network. The construction module is interconnected with the fusion module and the verification module via a wireless network. The fusion module and the verification module are interconnected with the display module via a wireless network.

[0016] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects: This invention accurately accepts user input in the form of text, graphics, and parameters and converts it into quantitative instructions. It aligns with the core processes of traditional ceramic processing, dynamically adjusting the angle and position of data acquisition to ensure comprehensive and complete data. Based on process limits, strength requirements, and aesthetic standards, it selects effective features to generate a three-dimensional model with continuous curved surfaces and consistent topology. Key processing information is embedded according to process levels, and parameter coordination constraints ensure accuracy and adaptability. It detects deviations in multiple dimensions and performs targeted calibration until preset requirements are met. Data is stably transmitted to the processing control and visualization platform in a standardized format, accurately restoring the shape, decoration, and craftsmanship essence of traditional ceramic handicrafts while improving processing adaptability and intuitiveness of display. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is a schematic diagram of a traditional ceramic handicraft display system based on AI technology. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] The present invention will be further described below with reference to embodiments.

[0021] Example: This embodiment presents a traditional ceramic handicraft display system based on AI technology, such as... Figure 1 As shown, it includes: The parsing module is used to receive the requirements for machining and reconstructing curved workpieces, build a semantic requirement model, and convert user input information into quantitative instructions. The user input information received by the parsing module includes a text description of process requirements, a graphic sketch of ceramic design, and key process parameter limitations. In the semantic requirements model construction phase, the text process description is segmented into process terms, size parameters are extracted and shape requirements are broken down. The graphic design sketch is subjected to feature contour recognition, pattern details are extracted and proportions are analyzed. Then, the extracted information in the text and graphics is mapped across dimensions through preset process feature association rules to form a semantic requirements model with core dimensions including process type, shape parameters, pattern specifications and precision standards. The generation of quantitative instructions is based on the core dimensions of the semantic demand model. Through parameter standardization transformation, natural language and graphic information are transformed into machine-recognizable quantitative data. Furthermore, each data item in the quantitative instructions is associated with a specific processing step in traditional ceramic handicrafts. The specific processing steps include throwing, trimming, glazing, and carving. The processing module is used to receive the output instructions from the parsing module to synchronously start the acquisition device, acquire the full-domain point cloud data of the workpiece, and complete noise reduction, registration, and establish an association index between the data and processing requirements. The noise reduction operation in the processing module is triggered based on the density distribution difference of point cloud data, and noise reduction is performed by removing outliers whose dispersion exceeds the preset range. The registration operation uses the reference feature surface of the ceramic workpiece as the registration reference, and the association index is established by mapping the preprocessed point cloud data with the process parameters in the quantization command and establishing hierarchical association. When the processing module starts the acquisition device, it controls the preset acquisition device to perform collaborative acquisition along the circumference and axial direction of the ceramic workpiece. During the acquisition process, the coverage of the point cloud data is checked in real time. When the coverage does not meet the preset requirements, the position and acquisition angle of the acquisition device are automatically adjusted and supplementary acquisition is performed to ensure that the acquired full-area point cloud data has no acquisition blind spots. The module is used to extract surface geometric features based on preprocessed point cloud data and generate an initial 3D model of the workpiece through a built-in topology reconstruction algorithm. When extracting surface geometric features, the construction module collects the contour curvature, texture concavity and convexity parameters, body wall thickness distribution, and interface transition curvature of the ceramic workpiece. The feature parameter validity screening is performed based on the preset ceramic surface process adaptation logic. This adaptation rule is based on the process limits, structural strength requirements and aesthetic standards of traditional ceramic hand-forming. Specifically, it includes the reasonable range of surface curvature, the matching ratio of the height of the decorative protrusions to the thickness of the body, and the smoothness requirements of the interface transition, etc. Invalid feature parameters that exceed the process adaptation range are eliminated accordingly. The topology reconstruction algorithm generates an initial 3D model of the workpiece based on the filtered effective feature parameters through topological connection of feature points, surface patch division, and smooth stitching. After the model is generated, the surface continuity of the model is checked by detecting the deviation of the included angle between adjacent surface patches and verifying the consistency of the surface tangent direction. The geometric topology consistency is checked by verifying the integrity of node connections and the absence of redundancy in the topology. After both checks pass, the initial 3D model is pushed to the fusion module. If the checks fail, the algorithm returns to the feature extraction stage to re-collect and filter geometric features. The fusion module is used to call the parsed machining parameters and embed the cutting path, accuracy requirements, feed rate, and cutting depth information into the initial 3D model of the workpiece. During the operation phase of the fusion module, the cutting path, accuracy requirements, feed rate, and cutting depth information are constructed into a multi-level embedded structure according to the ceramic processing procedure. The procedure includes at least body forming, decoration carving, and edge trimming. Among them, the accuracy requirement serves as the basic level for embedding core constraint information into the model. The basic level transforms the accuracy standard into geometric constraints that the model can recognize through parameterized constraint formulas, providing a unified accuracy benchmark for each processing step. The cutting path, feed rate, and depth of cut information are constructed based on the accuracy constraints of the basic level to form the corresponding process level. The information in each process level establishes a mapping relationship. That is, after the cutting path is dynamically planned according to the surface geometry of the ceramic workpiece, the feed rate and depth of cut are adaptively matched according to the curvature change of the cutting path and the material removal requirements to ensure that the three work together to meet the accuracy requirements. After the information is embedded, the compatibility of information at each level is verified, including the adaptability of cutting path and surface geometry, the rationality of feed rate and cutting depth linkage, the consistency of information at each process level with the accuracy requirements of the basic level, and the stability and ease of data storage and retrieval of information in the model. After all verification items pass, the model carrying complete information is output to the verification module. The constraint formula is: ; in, This represents the actual combined geometric deviation under the combined effects of the cutting path, feed rate, and depth of cut. This represents the weighting coefficients indicating the influence of cutting path, feed rate, and depth of cut on accuracy deviation. This indicates the cutting path length of the actual embedded model in the fusion module. This represents the theoretically optimal cutting path length obtained based on the surface geometry, pattern distribution, and machining process optimization of the ceramic workpiece. This indicates the feed rate of the actual embedded model in the fusion module. This indicates the reference feed rate that matches the preset accuracy requirements. This represents the cutting depth of the actual embedded model in the fusion module. This represents the reference cutting depth to match the preset accuracy requirements. This indicates the preset allowable deviation threshold for accuracy; in, All are greater than zero, and their sum is 1; The verification module is used to detect the accuracy deviation of the initial 3D model carrying information. When the accuracy deviation exceeds the preset threshold, the initial 3D model is calibrated and the verification module is refreshed. Otherwise, it is transmitted to the display module. The accuracy deviation detection of the verification module includes geometric dimension deviation, process parameter matching deviation, and surface topology deviation: Geometric dimensional deviation detection includes calculating the difference between the actual model value and the theoretical requirement value of the main body size of the ceramic workpiece, and determining the degree to which the dimensional tolerance range is met; Process parameter matching deviation detection includes the fit analysis of the cutting path, feed rate, and depth of cut embedded in the model with the process parameters in the semantic requirements model, and the rationality assessment of the collaborative matching between various process parameters; Surface topology deviation detection includes verifying the integrity of the model topology, checking the correctness of the connection relationships between topology nodes, and detecting the smoothness of the surface patch joints; When the deviation in any dimension exceeds the corresponding preset threshold, the verification module automatically calls the corresponding preset calibration algorithm based on the deviation type. Specifically, the geometric dimension deviation is corrected by using a dimension compensation calibration algorithm, which adjusts the key geometric parameters of the model to correct the deviation; the process parameter matching deviation is corrected by using a parameter adaptation adjustment algorithm, which optimizes the process parameters embedded in the model based on the standard parameters in the demand model; and the surface topology deviation is corrected by using a topology reconstruction optimization algorithm, which reorganizes and optimizes the model topology. After calibration, restart the full-dimensional accuracy deviation detection until all dimensional deviations are within the corresponding preset threshold range, obtain a qualified model, and transmit it to the display module; If the deviation requirement is still not met after the number of calibration attempts reaches the preset limit, the calibration process will be terminated and a prompt message containing the deviation type, the degree of exceeding the limit, and the calibration attempt record will be output. The display module is used to receive the model output by the verification module and forward the model to the preset processing control system and the 3D visualization platform. The display module converts the verified model according to a preset standardized data format. The converted model data is then forwarded to the machining control system and the 3D visualization platform via a two-way communication link. During the forwarding process, the data transmission status is monitored in real time. When an abnormality occurs during transmission, the breakpoint resume mechanism is automatically activated, enabling the machining control system and the 3D visualization platform to obtain complete model data. The parsing module interacts with the processing module and the building module via a wireless network. The building module interacts with the fusion module and the verification module via a wireless network. The fusion module and the verification module interact with each other via a wireless network and there is a display module.

[0022] In this embodiment, the parsing module receives the requirements for machining and reconstructing the curved workpiece, constructs a semantic requirement model, and converts the user input information into quantitative instructions. The processing module, running after the parsing module, receives the output instructions and simultaneously starts the acquisition device to obtain the full-domain point cloud data of the workpiece and completes noise reduction and registration, establishing an association index between the data and the machining requirements. The construction module further extracts the surface geometric features based on the preprocessed point cloud data and generates an initial 3D model of the workpiece through a built-in topology reconstruction algorithm. Then, the fusion module calls the parsed machining parameters to embed the cutting path, accuracy requirements, feed rate, and cutting depth information into the initial 3D model of the workpiece. The verification module performs accuracy deviation detection on the initial 3D model carrying the information. When the accuracy deviation exceeds a preset threshold, the initial 3D model is calibrated and the verification module is refreshed. Otherwise, it is transmitted to the display module. Finally, the display module receives the model output by the verification module and forwards the model to the preset machining control system and the 3D visualization platform.

[0023] In the above embodiments, the system can accurately convert ceramic processing requirements, comprehensively collect and optimize workpiece data, generate a three-dimensional model that fits traditional processes, incorporate suitable processing parameters and ensure that the accuracy meets the standards, and stably transmit data to the processing and visualization end. It not only restores the details and process standards of traditional ceramic handicrafts, but also improves the processing adaptability and display effect.

[0024] It should be noted that: The process feature association rules are constructed based on the standardized processing flow of traditional ceramic handicrafts. The correspondence between process type and outline features in graphics, the conversion formula between size parameters and graphic proportions, and the matching standards between decorative specification descriptions and graphic decorative details are established. By accurately mapping the process requirements in the text dimension to the shape information in the graphic dimension according to the above correspondence, the integrity and consistency of information in each core dimension of the semantic requirement model are ensured, providing a clear basis for cross-dimensional information fusion.

[0025] The acquisition equipment selected is a laser 3D scanner, which must meet the scanning accuracy requirements of the ceramic workpiece (scanning resolution not less than 0.01mm), the adaptability of the acquisition range (covering ceramic workpieces with diameters of 5cm-50cm and heights of 3cm-60cm), and the acquisition frequency of not less than 30 frames / second. When the equipment acquires data along the circumference and axis of the ceramic workpiece, the circumferential acquisition interval is set to 15°-30°, and the axial acquisition step size is 2cm-5cm.

[0026] The quantitative indicators for ceramic curved surface process adaptation logic are determined based on the actual technological limits, structural strength requirements, and aesthetic standards of traditional ceramic hand-forming. Specifically, the reasonable range for surface curvature is -0.5 rad to 0.5 rad; the height of decorative protrusions does not exceed 30% of the body thickness; the angle between adjacent curved surface pieces at the interface transition does not exceed 5°; and the difference in body wall thickness distribution does not exceed 0.3 mm. During feature parameter screening, if any parameter exceeds the above quantitative thresholds, it is determined to be an invalid feature parameter and discarded.

[0027] In the multi-level embedded structure, the precision constraint information of the base level is stored in the form of a parameterized dataset, and the process level reads the precision standard of the base level in real time through a data call interface. Each process level is constructed sequentially in the order of blank forming → pattern carving → edge trimming. The processing parameter output results of the preceding process level serve as the input basis for the subsequent process level. For example, the cutting path planning results of the blank forming process level provide a basis for the adaptation of feed rate and cutting depth of the pattern carving process level. At the same time, the information of each level adopts a unified data encoding format.

[0028] The size compensation calibration algorithm adopts a linear adjustment model based on deviation values. According to the ratio of the difference between the actual deviation of the geometric dimensions and the theoretical requirement value, it makes targeted adjustments to key geometric parameters such as the length, diameter, and height of the model. The parameter adaptation adjustment algorithm is implemented through iterative optimization. Taking the standard parameters in the semantic requirement model as the target values, each iteration sets an adjustment step size of 0.1-0.3 according to the degree of deviation, gradually optimizing the process parameters embedded in the model until the fit reaches more than 95%. The topology reconstruction optimization algorithm first disassembles the abnormal topology structure, re-establishes the node connection relationship according to the surface geometric features, and smooths the splicing of surface patches to ensure the integrity of the topology structure, the correctness of the node connection and the absence of redundancy. The specific implementation logic of various calibration algorithms is clearly defined.

[0029] Weighting coefficients in the accuracy constraint formula Based on the influence of cutting path, feed rate, and depth of cut on machining accuracy in traditional ceramic processing, the value ranges were determined by fitting machining experimental data from over 100 sets of different ceramic workpieces. Specifically, k1 (cutting path influence weight) ranged from 0.4 to 0.6, k2 (feed rate influence weight) from 0.2 to 0.3, and k3 (depth of cut influence weight) from 0.1 to 0.3. In practical applications, fine adjustments can be made within these ranges based on the material of the ceramic workpiece (e.g., kaolin, porcelain stone) and the complexity of its shape (e.g., simple shapes, complex patterns) to ensure the accuracy of the calculated deviation.

[0030] In summary, the system in the above embodiments can accurately accept user input in the form of text, graphics, and parameters and convert it into quantitative instructions. It aligns with the core processes of traditional ceramic processing, dynamically adjusts the angle and position of data acquisition to ensure comprehensive and complete data. Based on process limits, strength requirements, and aesthetic standards, it selects effective features to generate a three-dimensional model with continuous curved surfaces and consistent topology. It embeds key processing information according to process levels, ensures accuracy through parameter coordination constraints, detects deviations in multiple dimensions and performs targeted calibration until preset requirements are met, and transmits data stably to the processing control and visualization platform in a standardized format. This not only accurately restores the shape, decoration, and essence of traditional ceramic craftsmanship but also improves processing adaptability and intuitiveness of display. It achieves efficient digital presentation and high-quality inheritance of traditional techniques, balancing processing accuracy and the dissemination of skills, enabling traditional ceramic craftsmanship to be better continued and promoted in digital scenarios.

[0031] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A traditional ceramic handicraft display system based on AI technology, characterized in that: include: The parsing module is used to receive the requirements for machining and reconstructing curved workpieces, build a semantic requirement model, and convert user input information into quantitative instructions. The processing module is used to receive the output instructions from the parsing module to synchronously start the acquisition device, acquire the full-domain point cloud data of the workpiece, and complete noise reduction, registration, and establish an association index between the data and processing requirements. The module is used to extract surface geometric features based on preprocessed point cloud data and generate an initial 3D model of the workpiece through a built-in topology reconstruction algorithm. The fusion module is used to call the parsed machining parameters and embed the cutting path, accuracy requirements, feed rate, and cutting depth information into the initial 3D model of the workpiece. The verification module is used to detect the accuracy deviation of the initial 3D model carrying information. When the accuracy deviation exceeds the preset threshold, the initial 3D model is calibrated and the verification module is refreshed. Otherwise, it is transmitted to the display module. The display module is used to receive the model output by the verification module and forward the model to the preset processing control system and the 3D visualization platform.

2. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, The user input information received by the parsing module includes a text description of process requirements, a graphic sketch of ceramic design, and key process parameter limitations. The semantic requirements model construction stage performs process term segmentation, size parameter extraction and shape requirement decomposition on the text process description, feature contour recognition, pattern detail extraction and proportional relationship analysis on the graphic design sketch, and then cross-dimensional mapping of the extracted information in the text and graphics through preset process feature association rules to form a semantic requirements model with core dimensions including process type, shape parameters, pattern specifications and precision standards. The generation of the quantitative instructions is based on the core dimensions of the semantic demand model. Natural language and graphic information are transformed into machine-recognizable quantitative data through parameter standardization. Each data item in the quantitative instructions is associated with a specific processing step in traditional ceramic handicrafts. The specific processing steps include throwing, trimming, glazing, and carving.

3. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, The noise reduction operation in the processing module is triggered based on the density distribution difference of point cloud data, and noise reduction is performed by removing outlier points whose dispersion exceeds a preset range. The registration operation uses the reference feature surface of the ceramic workpiece as the registration reference, and the association index is established by label mapping and hierarchical association between the preprocessed point cloud data and the process parameters in the quantization instruction.

4. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, When the processing module starts the acquisition device, it controls the preset acquisition device to perform collaborative acquisition along the circumference and axial direction of the ceramic workpiece. During the acquisition process, the coverage of the point cloud data is checked in real time. When the coverage does not meet the preset requirements, the position and acquisition angle of the acquisition device are automatically adjusted and supplementary acquisition is performed to ensure that the acquired full-domain point cloud data has no acquisition blind spots.

5. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, When extracting surface geometric features, the construction module collects the contour curvature, texture concavity and convexity parameters, body wall thickness distribution, and interface transition curvature of the ceramic workpiece. The validity screening of feature parameters is performed based on the preset ceramic surface process adaptation logic, which includes the reasonable range of surface curvature, the matching ratio of the height of the decorative protrusions to the thickness of the body, and the smoothness requirements of the interface transition, thereby eliminating invalid feature parameters that exceed the process adaptation range. The topology reconstruction algorithm generates an initial 3D model of the workpiece based on the filtered effective feature parameters through topological connection of feature points, surface patch division, and smooth splicing. After the model is generated, the surface continuity of the model is checked by detecting the deviation of the included angle between adjacent surface patches and verifying the consistency of the surface tangent direction. The geometric topology consistency is checked by verifying the integrity of node connections and the absence of redundancy in the topology. After both checks pass, the initial 3D model is pushed to the fusion module. If the checks fail, the algorithm returns to the feature extraction stage to re-collect and filter geometric features.

6. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, During the operation phase of the fusion module, the cutting path, accuracy requirements, feed rate, and cutting depth information are constructed into a multi-level embedded structure according to the process flow of ceramic processing. Among them, the accuracy requirement serves as the basic level for embedding core constraint information into the model. The basic level transforms the accuracy standard into geometric constraints that the model can recognize through parameterized constraint formulas, providing a unified accuracy benchmark for each processing step. The cutting path, feed rate, and depth of cut information are constructed based on the accuracy constraints of the basic level to form the corresponding process level. The information in each process level establishes a mapping relationship. That is, after the cutting path is dynamically planned according to the surface geometry of the ceramic workpiece, the feed rate and depth of cut are adaptively matched according to the curvature change of the cutting path and the material removal requirements to ensure that the three work together to meet the accuracy requirements. After the information is embedded, the compatibility of information at each level is verified, including the adaptability of the cutting path and surface geometry, the rationality of the linkage between feed rate and cutting depth, and the consistency between the information at each process level and the accuracy requirements of the basic level.

7. The traditional ceramic handicraft display system based on AI technology according to claim 6, characterized in that, The constraint formula is: ； in, This represents the actual combined geometric deviation resulting from the combined effects of the cutting path, feed rate, and depth of cut. This represents the weighting coefficients indicating the influence of cutting path, feed rate, and depth of cut on accuracy deviation. This indicates the cutting path length of the actual embedded model in the fusion module. This represents the theoretically optimal cutting path length obtained based on the surface geometry, texture distribution, and machining process optimization of the ceramic workpiece. This indicates the feed rate of the actual embedded model in the fusion module. This indicates the reference feed rate that matches the preset accuracy requirements. This represents the cutting depth of the actual embedded model in the fusion module. This represents the reference cutting depth to match the preset accuracy requirements. This indicates the preset allowable deviation threshold for accuracy; in, All are greater than zero, and their sum is 1.

8. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, The accuracy deviation detection of the verification module includes geometric dimension deviation, process parameter matching deviation, and surface topology deviation: Geometric dimensional deviation detection includes calculating the difference between the actual model value and the theoretical requirement value of the main body size of the ceramic workpiece, and determining the degree to which the dimensional tolerance range is met; Process parameter matching deviation detection includes the fit analysis of the cutting path, feed rate, and depth of cut embedded in the model with the process parameters in the semantic requirements model, and the rationality assessment of the collaborative matching between various process parameters; Surface topology deviation detection includes verifying the integrity of the model topology, checking the correctness of the connection relationships between topology nodes, and detecting the smoothness of the surface patch joints; When the deviation in any dimension exceeds the corresponding preset threshold, the verification module automatically calls the corresponding preset calibration algorithm based on the deviation type. Specifically, the geometric dimension deviation is corrected by using a dimension compensation calibration algorithm, which adjusts the key geometric parameters of the model to correct the deviation; the process parameter matching deviation is corrected by using a parameter adaptation adjustment algorithm, which optimizes the process parameters embedded in the model based on the standard parameters in the demand model; and the surface topology deviation is corrected by using a topology reconstruction optimization algorithm, which reorganizes and optimizes the model topology. After calibration, restart the full-dimensional accuracy deviation detection until all dimensional deviations are within the corresponding preset threshold range, obtain a qualified model, and transmit it to the display module; If the deviation requirement is still not met after the preset limit of calibration attempts, the calibration process will be terminated, and a prompt message containing the deviation type, the degree of exceeding the limit, and the calibration attempt record will be output.

9. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, The display module converts the verified model according to a preset standardized data format. The converted model data is then forwarded to the machining control system and the 3D visualization platform via a two-way communication link. During the forwarding process, the data transmission status is monitored in real time. When an abnormality occurs during transmission, the breakpoint resume mechanism is automatically activated, enabling the machining control system and the 3D visualization platform to obtain complete model data.

10. The traditional ceramic handicraft display system based on AI technology according to claim 1, characterized in that, The parsing module is interactively connected to the processing module and the construction module via a wireless network. The construction module is interactively connected to the fusion module and the verification module via a wireless network. The fusion module and the verification module are interactively connected to the display module via a wireless network.

Images