Virtual restoration system for damaged cultural relics

Abstract

The invention discloses a virtual restoration system for damaged cultural relics, relates to the field of virtual restoration of the damaged cultural relics, and solves the problems that the restoration effects of textures and geometric shapes of damaged areas are difficult to be completely consistent, the recognition capability for complex damage modes and disease types is weak, and the restoration effect is poor. The technical problems of difficulty in realizing sharing and cooperative processing of cultural relic data among different mechanisms and lack of digital display and propagation of models in the prior art are solved in the invention. The method comprises the following modules: a cultural relic fragment data import and export module, which is used for obtaining three-dimensional information of cultural relic fragments through a three-dimensional scanning device and then carrying out preprocessing operation; the cultural relic fragment data feature enhancement and classification module is used for extracting geometric features and texture features of cultural relic fragments by using corresponding technologies, and constructing a three-dimensional digital model after classification and splicing by combining with a machine algorithm, and the cultural relic fragment interactive splicing module is used for developing a data sharing platform by using a cloud computing technology. And constructing a virtual cultural relic restoration environment by using a virtual reality technology.

Description

Technical Field

[0001] The invention belongs to the field of virtual restoration of damaged cultural relics, and in particular relates to a virtual restoration system for damaged cultural relics. Background Art

[0002] In the field of cultural heritage protection, with the rapid development of science and technology, digital technology has become an important means of repairing and protecting damaged cultural relics. Therefore, a virtual restoration system for damaged cultural relics has come into being. It uses 3D scanning equipment and 3D reconstruction technology to improve scanning accuracy and modeling speed. It uses advanced image processing technology and learning algorithms to improve the restoration accuracy of textures and geometric shapes in the virtual restoration process. It develops intelligent recognition and analysis modules to improve recognition accuracy and efficiency, and develops cross-platform and cross-system data exchange and collaborative work mechanisms to promote the widespread application of cultural relics restoration work. It uses virtual reality and augmented reality technologies to realize the digital display and interactive experience of the cultural relics restoration process, so that the public can intuitively understand the restoration process and effects.

[0003] Although existing technologies for the digital restoration of damaged cultural relics have achieved certain results, existing restoration methods often rely on the experience of experts and manual operations, which are not only time-consuming and labor-intensive, but also pose the risk of causing secondary damage to the cultural relics. It is difficult to restore the texture and geometry of the damaged area to be completely consistent through three-dimensional scanning and modeling, and the ability to identify complex damage patterns and disease types is weak, making it difficult to share and coordinate cultural relic data between different institutions, resulting in a lack of unified standards and specifications for restoration work. The application of digital display and dissemination technology in cultural relic restoration is not yet popular, and users cannot intuitively understand the restoration process and effects. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a virtual restoration system for damaged cultural relics, which is used to solve the following technical problems:

[0005] Although existing technologies for the digital restoration of damaged cultural relics have achieved certain results, existing restoration methods often rely on the experience of experts and manual operations, which are not only time-consuming and labor-intensive, but also pose the risk of causing secondary damage to the cultural relics. It is difficult to restore the texture and geometry of the damaged area to be completely consistent through three-dimensional scanning and modeling, and the ability to identify complex damage patterns and disease types is weak, making it difficult to share and coordinate cultural relic data between different institutions, resulting in a lack of unified standards and specifications for restoration work. The application of digital display and dissemination technology in cultural relic restoration is not yet popular, and users cannot intuitively understand the restoration process and effects.

[0006] To solve the above problems, the first aspect of the present invention provides a virtual restoration system for damaged cultural relics, comprising the following modules:

[0007] Cultural relic fragment data import and export module: After obtaining the three-dimensional information of cultural relic fragments through three-dimensional scanning equipment, pre-processing operations are performed and digital storage is performed to form complete fragment model information;

[0008] Cultural relic fragment data feature enhancement and classification module: The geometric features of cultural relic fragments are extracted from point cloud data using computational geometry technology, and the texture features of cultural relic fragments are extracted using image processing technology. The cultural relic fragments are classified and spliced ​​in combination with machine learning algorithms, and merged into a three-dimensional digital model. The ResNet framework in the convolutional neural network is used to build a model for identifying damaged areas, and the degree of damage to cultural relics is divided by the comprehensive damage degree value;

[0009] Interactive splicing module for cultural relics fragments: By analyzing the existing data format and structure, defining data elements, using cloud computing technology to develop a data sharing platform, and using virtual reality technology and augmented reality technology to build a virtual cultural relic restoration environment, the system provides real-time feedback on the splicing effect based on the user's selection and operation.

[0010] As a further solution of the present invention, the preprocessing operation after obtaining the three-dimensional information of the cultural relic fragments by the three-dimensional scanning device includes the following steps:

[0011] Use 3D laser scanners and structured light scanners to conduct all-round scanning of the cultural relic fragments to obtain 3D point cloud information. Preprocess the collected raw data, including denoising, filtering and registration, and mark and record the damaged areas of the cultural relics.

[0012] As a further solution of the present invention, the method of extracting geometric features of cultural relic fragments from point cloud data by computational geometry technology, extracting texture features of cultural relic fragments by image processing technology, classifying and splicing cultural relic fragments in combination with machine learning algorithms, and merging them into a three-dimensional digital model includes the following steps:

[0013] The geometric features of the cultural relic fragments, including shape, size and curvature, are extracted from the point cloud data using computational geometry technology. The texture features of the cultural relic fragments are extracted using image processing technology. The extracted geometric features and texture features are fused to form a comprehensive feature vector. The support vector machine is used as a classifier to train the comprehensive feature vector and build a classification model for the cultural relic fragments.

[0014] The four-point algorithm is used to roughly align the fragments of the cultural relic. After determining the positional relationship between the fragments, the iterative closest point algorithm is used for precise alignment. The fragments of the cultural relic are assigned to multiple nodes for simultaneous feature extraction and splicing. The MPI message passing mechanism is used for communication and data synchronization between nodes. When all the fragments of the cultural relic are spliced ​​together, they are merged into a three-dimensional digital model of the entire cultural relic. The synthesized three-dimensional digital model is then verified to check whether there are any unspliced ​​or spliced ​​areas.

[0015] As a further solution of the present invention, the method of extracting texture features of cultural relic fragments using image processing technology comprises the following steps:

[0016] Use a high-resolution camera to capture images of cultural relics fragments at different angles, read and load the images of cultural relics fragments to be processed, convert the original images into grayscale images, apply filters to remove noise in the images, smooth the images and highlight edge features, convert the grayscale images into binary images, use edge detection algorithms to extract edge information in the images, and then perform morphological operations;

[0017] The gray-level co-occurrence matrix is ​​used to extract the texture features of the image. The parameters for calculating the gray-level co-occurrence matrix, including direction and interval, are determined. Each pixel in the image is traversed. According to the selected direction and interval, the probability of occurrence of pixel pairs that meet the conditions is counted. The gray-level co-occurrence matrix is ​​normalized. The texture features of the processed cultural relic fragment image are extracted, including contrast, correlation and energy uniformity eigenvalues. Each pixel in the image is traversed to obtain its corresponding texture eigenvalue. The eigenvalue is converted into a corresponding color value using gray-level mapping. A texture feature map is constructed by assigning the converted color value to the position of the corresponding pixel in the original image.

[0018] As a further solution of the present invention, the ResNet framework in the convolutional neural network is used to construct a model for identifying damaged areas, and the degree of damage to cultural relics is divided according to the comprehensive damage degree value, including the following steps:

[0019] Collect image data of cultural relics fragments containing various damaged areas, annotate the collected images, perform data enhancement on the annotated images, select the ResNet framework in the convolutional neural network, and use the annotated data to train the neural network model. During the training process, adjust the model parameters through the back propagation algorithm to minimize the loss function value, deploy the trained model to the actual application environment, and use the deployed model to perform real-time recognition of newly collected cultural relics images to automatically detect damaged areas;

[0020] Use a camera to record images of damaged areas, observe the edges of damaged areas, surface textures, distribution and width of cracks, extract area and density features of the identified damaged areas, and combine the ratio of the damaged area area to the normal area area and the ratio of the damaged area density to the normal area density to obtain a comprehensive damage degree value;

[0021] The calculation formula of comprehensive disease degree value is:

[0022]

[0023] Among them, C is the comprehensive disease severity value, S damage is the area of ​​the damaged area, S normal is the area of ​​the normal region, D damage is the density of the damaged area, D normal is the density of the normal area;

[0024] The degree of damage to cultural relics is divided into three levels according to the comprehensive damage degree value;

[0025] If 0≤C<30%, it is defined as a mild disease degree;

[0026] If 30% ≤ C < 70%, it is defined as a moderate disease degree;

[0027] If 70% ≤ C, it is defined as a severe disease degree.

[0028] As a further solution of the present invention, the data sharing platform is developed by analyzing the existing data format and structure, defining data elements, and using cloud computing technology, including the following steps:

[0029] Identify the institutions and departments that will participate in sharing, determine the data needs and usage scenarios of each institution, analyze the existing data format and structure, define data elements, including data name, type, length and format, and specify the data exchange format, use cloud computing technology to develop a data sharing platform, build a data warehouse to store and manage data from different institutions, encrypt data during transmission, convert data from different sources into a unified format, establish a correspondence between the new and old data standards, and import the cleaned data into the sharing platform.

[0030] As a further solution of the present invention, the virtual reality technology and augmented reality technology are used to construct a virtual cultural relic restoration environment, and the system provides real-time feedback splicing effects according to the user's selection and operation, including the following steps:

[0031] A virtual cultural relic restoration environment is constructed using virtual reality technology and augmented reality technology, including a virtual workbench, tools and cultural relic fragment elements, providing a user interactive interface for splicing cultural relic fragments. The system provides real-time feedback on the splicing effect based on the user's selection and operation. After the splicing is completed, the user can export the virtually restored cultural relics in the form of a three-dimensional model and share it on social media.

[0032] Beneficial effects of the present invention:

[0033] The present invention improves scanning accuracy and modeling speed through 3D scanning and modeling algorithms, reduces manual intervention, and realizes rapid and accurate identification and repair of damaged areas. It uses image processing technology and machine learning algorithms to improve the repair accuracy of textures and geometric shapes during virtual repair, making the repaired cultural relics more realistic and natural.

[0034] The present invention establishes unified data standards and specifications to achieve efficient sharing and collaborative processing of cultural relics data among different institutions, develops cross-platform and cross-system data exchange and collaborative work systems, promotes the widespread promotion and application of cultural relics restoration work, and uses virtual reality and augmented reality technologies to achieve digitization and interactive experience of the cultural relics restoration process, so that the public can intuitively understand the restoration process and effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 It is a schematic diagram of the module framework of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] See also Figure 1 As shown, the present invention is a virtual restoration system for damaged cultural relics, comprising the following modules:

[0039] Cultural relic fragment data import and export module: After obtaining the three-dimensional information of cultural relic fragments through three-dimensional scanning equipment, pre-processing operations are performed and digital storage is performed to form complete fragment model information;

[0040] Cultural relic fragment data feature enhancement and classification module: The geometric features of cultural relic fragments are extracted from point cloud data using computational geometry technology, and the texture features of cultural relic fragments are extracted using image processing technology. The cultural relic fragments are classified and spliced ​​in combination with machine learning algorithms, and merged into a three-dimensional digital model. The ResNet framework in the convolutional neural network is used to build a model for identifying damaged areas, and the degree of damage to cultural relics is divided by the comprehensive damage degree value;

[0041] Interactive splicing module for cultural relics fragments: By analyzing the existing data format and structure, defining data elements, using cloud computing technology to develop a data sharing platform, and using virtual reality technology and augmented reality technology to build a virtual cultural relic restoration environment, the system provides real-time feedback on the splicing effect based on the user's selection and operation.

[0042] Specifically, the cultural relic fragment data import and export module is used to obtain the three-dimensional information of the cultural relic fragments through the three-dimensional scanning equipment and then perform preprocessing operations. The cultural relic fragment data feature enhancement and classification module is used to use the corresponding technology to extract the geometric features and texture features of the cultural relic fragments, and combine the machine algorithm to classify and splice them to construct a three-dimensional digital model. The cultural relic fragment interactive splicing module is used to use cloud computing technology to develop a data sharing platform and use virtual reality technology to build a virtual cultural relic restoration environment.

[0043] In one embodiment of the present invention, the preprocessing operation after acquiring the three-dimensional information of the cultural relic fragments by a three-dimensional scanning device includes the following steps:

[0044] Use 3D laser scanners and structured light scanners to conduct all-round scanning of the cultural relic fragments to obtain 3D point cloud information. Preprocess the collected raw data, including denoising, filtering and registration, and mark and record the damaged areas of the cultural relics.

[0045] Specifically, select appropriate 3D laser scanners and structured light scanners according to the size, shape and material of the cultural relic fragments, adjust the position and angle of the scanner to ensure full coverage of the surface of the cultural relic fragments, use the scanner to project and photograph the surface of the cultural relic fragments, obtain the 3D coordinate information of the cultural relic surface, and scan the cultural relic fragments multiple times from multiple angles to obtain geometric data, so that the scanning data from each perspective generates a set of point cloud data; perform denoising, filtering and alignment preprocessing operations on the collected raw data, and convert the point cloud data from multiple perspectives into the same reference coordinate system to form a complete point cloud data set.

[0046] In one embodiment of the present invention, the method of extracting geometric features of cultural relic fragments from point cloud data by computational geometry technology, extracting texture features of cultural relic fragments by image processing technology, classifying and splicing cultural relic fragments in combination with machine learning algorithms, and merging them into a three-dimensional digital model includes the following steps:

[0047] The geometric features of the cultural relic fragments, including shape, size and curvature, are extracted from the point cloud data using computational geometry technology. The texture features of the cultural relic fragments are extracted using image processing technology. The extracted geometric features and texture features are fused to form a comprehensive feature vector. The support vector machine is used as a classifier to train the comprehensive feature vector and build a classification model for the cultural relic fragments.

[0048] The four-point algorithm is used to roughly align the fragments of the cultural relic. After determining the positional relationship between the fragments, the iterative closest point algorithm is used for precise alignment. The fragments of the cultural relic are assigned to multiple nodes for simultaneous feature extraction and splicing. The MPI message passing mechanism is used for communication and data synchronization between nodes. When all the fragments of the cultural relic are spliced ​​together, they are merged into a three-dimensional digital model of the entire cultural relic. The synthesized three-dimensional digital model is then verified to check whether there are any unspliced ​​or spliced ​​areas.

[0049] Specifically, geometric features including shape, size and curvature are extracted from the preprocessed point cloud data, the three-dimensional point cloud data is converted into two-dimensional image data, and image processing algorithms are applied to extract texture features. The extracted geometric features and texture features are fused through multimodal feature fusion to form a comprehensive feature vector. A support vector machine is used as a classifier to train the comprehensive feature vector, and the trained classification model is used to classify the cultural relic fragments. The cultural relic fragments are roughly aligned using a four-point algorithm to determine the positional relationship between the fragments. Based on the rough alignment, an iterative closest point algorithm is used for precise alignment to obtain higher alignment accuracy. The cultural relic fragments are assigned to multiple nodes, each node simultaneously performs feature extraction and splicing, and an MPI message passing mechanism is used for communication and data synchronization between nodes. When all the cultural relic fragments are spliced, they are merged into an overall three-dimensional digital model.

[0050] In one embodiment of the present invention, the method of extracting texture features of cultural relic fragments using image processing technology includes the following steps:

[0051] Use a high-resolution camera to capture images of cultural relics fragments at different angles, read and load the images of cultural relics fragments to be processed, convert the original images into grayscale images, apply filters to remove noise in the images, smooth the images and highlight edge features, convert the grayscale images into binary images, use edge detection algorithms to extract edge information in the images, and then perform morphological operations;

[0052] The gray-level co-occurrence matrix is ​​used to extract the texture features of the image. The parameters for calculating the gray-level co-occurrence matrix, including direction and interval, are determined. Each pixel in the image is traversed. According to the selected direction and interval, the probability of occurrence of pixel pairs that meet the conditions is counted. The gray-level co-occurrence matrix is ​​normalized. The texture features of the processed cultural relic fragment image are extracted, including contrast, correlation and energy uniformity eigenvalues. Each pixel in the image is traversed to obtain its corresponding texture eigenvalue. The eigenvalue is converted into a corresponding color value using gray-level mapping. A texture feature map is constructed by assigning the converted color value to the position of the corresponding pixel in the original image.

[0053] Specifically, after converting the original image into a grayscale image and then converting the grayscale image into a binary image, an edge detection algorithm is used to extract edge information in the image, and a gray-level co-occurrence matrix is ​​used to extract texture features of the image. The parameters for calculating the gray-level co-occurrence matrix are determined, including direction and interval, where the direction is horizontal 0°, vertical 90°, and diagonal 45° or 135°; the interval is the physical distance between pixels. Each pixel in the image is traversed, and the probability of occurrence of pixel pairs that meet the conditions is counted according to the selected direction and interval. If the horizontal direction and interval are selected as 1, all adjacent pixels with a horizontal distance of 1 will be counted. The number of occurrences of pixel pairs with a grayscale co-occurrence distance of 1; each element in the grayscale co-occurrence matrix is ​​divided by the sum of all elements in the matrix for normalization, and texture features are extracted from the processed cultural relic fragment image, including contrast, correlation and energy uniformity eigenvalues. Each pixel in the image is traversed, and the texture eigenvalue corresponding to the pixel is obtained according to its grayscale value and the grayscale values ​​of the surrounding pixels. The extracted texture eigenvalue is converted into a corresponding color value through grayscale mapping, and the converted color value is assigned to the position of the corresponding pixel in the original image. A texture feature map is constructed to show the texture distribution and characteristics of the cultural relic fragment.

[0054] In one embodiment of the present invention, the ResNet framework in the convolutional neural network is used to build a model for identifying damaged areas, and the degree of damage to cultural relics is divided according to the comprehensive damage degree value, including the following steps:

[0055] Collect image data of cultural relics fragments containing various damaged areas, annotate the collected images, perform data enhancement on the annotated images, select the ResNet framework in the convolutional neural network, and use the annotated data to train the neural network model. During the training process, adjust the model parameters through the back propagation algorithm to minimize the loss function value, deploy the trained model to the actual application environment, and use the deployed model to perform real-time recognition of newly collected cultural relics images to automatically detect damaged areas;

[0056] Use a camera to record images of damaged areas, observe the edges of damaged areas, surface textures, distribution and width of cracks, extract area and density features of the identified damaged areas, and combine the ratio of the damaged area area to the normal area area and the ratio of the damaged area density to the normal area density to obtain a comprehensive damage degree value;

[0057] The calculation formula of comprehensive disease degree value is:

[0058]

[0059] Among them, C is the comprehensive disease severity value, S damage is the area of ​​the damaged area, S normal is the area of ​​the normal region, D damage is the density of the damaged area, D normal is the density of the normal area;

[0060] The degree of damage to cultural relics is divided into three levels according to the comprehensive damage degree value;

[0061] If 0≤C<30%, it is defined as a mild disease degree;

[0062] If 30% ≤ C < 70%, it is defined as a moderate disease degree;

[0063] If 70% ≤ C, it is defined as a severe disease degree.

[0064] Specifically, image data of cultural relics fragments containing various damaged areas are collected, and data enhancement processing is performed after the collected images are annotated. The ResNet framework in the convolutional neural network is selected as the basic network structure, and the neural network model is trained using the annotated data. During the training process, the model parameters are adjusted through the back propagation algorithm to minimize the loss function value, and the trained model is deployed in the actual environment. The comprehensive disease degree value is obtained by jointly calculating the ratio of the area of ​​the damaged area to the area of ​​the normal area and the ratio of the density of the damaged area to the density of the normal area. According to the comprehensive disease degree value, the degree of cultural relics disease is divided into three levels, including mild, moderate and severe.

[0065] In one embodiment of the present invention, the method of analyzing the existing data format and structure, defining data elements, and developing a data sharing platform using cloud computing technology includes the following steps:

[0066] Identify the institutions and departments that will participate in sharing, determine the data needs and usage scenarios of each institution, analyze the existing data format and structure, define data elements, including data name, type, length and format, and specify the data exchange format, use cloud computing technology to develop a data sharing platform, build a data warehouse to store and manage data from different institutions, encrypt data during transmission, convert data from different sources into a unified format, establish a correspondence between the new and old data standards, and import the cleaned data into the sharing platform.

[0067] Specifically, determine the institutions and departments that will participate in data sharing, define their respective responsibilities and roles, and communicate with each participating institution and department to understand their data needs and usage scenarios. Conduct a comprehensive analysis of existing data, including data types, sources, formats, and storage methods, to lay the foundation for subsequent data integration and standardization. Based on the results of the analysis, define data elements, including data name, type, length, and format. Develop unified data exchange standards and formats to enable smooth data exchange between different systems. Use cloud computing technology to develop a data sharing platform, establish a data warehouse, and centrally store and manage data from different institutions. Use encryption technology during data transmission to ensure data security during transmission. While introducing new data standards, establish a correspondence between old and new data standards to ensure data compatibility and import the cleaned data into the sharing platform.

[0068] In one embodiment of the present invention, the virtual reality technology and augmented reality technology are used to construct a virtual cultural relic restoration environment, and the system provides real-time feedback splicing effects according to the user's selection and operation, including the following steps:

[0069] A virtual cultural relic restoration environment is constructed using virtual reality technology and augmented reality technology, including a virtual workbench, tools and cultural relic fragment elements, providing a user interactive interface for splicing cultural relic fragments. The system provides real-time feedback on the splicing effect based on the user's selection and operation. After the splicing is completed, the user can export the virtually restored cultural relics in the form of a three-dimensional model and share it on social media.

[0070] Specifically, use 3D modeling software to create elements of virtual workbenches, tools, and cultural relic fragments, design user interfaces, including menus, instructions, and feedback mechanisms, implement user hand tracking so that users can grab, move, and rotate cultural relic fragments, write scripts to process user input and update status based on operations, develop a system that provides real-time feedback based on user selections and operations, allow users to save and export results after completing virtual restoration, and integrate social media sharing functions so that users can show their results to friends and family.

[0071] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A virtual restoration system for damaged cultural relics, characterized in that: Includes the following modules: Cultural relic fragment data import and export module: After obtaining the three-dimensional information of cultural relic fragments through three-dimensional scanning equipment, pre-processing operations are performed and digital storage is performed to form complete fragment model information; Cultural relic fragment data feature enhancement and classification module: The geometric features of cultural relic fragments are extracted from point cloud data using computational geometry technology, and the texture features of cultural relic fragments are extracted using image processing technology. The cultural relic fragments are classified and spliced ​​in combination with machine learning algorithms, and merged into a three-dimensional digital model. The ResNet framework in the convolutional neural network is used to build a model for identifying damaged areas, and the degree of damage to cultural relics is divided by the comprehensive damage degree value; Interactive splicing module for cultural relics fragments: By analyzing the existing data format and structure, defining data elements, using cloud computing technology to develop a data sharing platform, and using virtual reality technology and augmented reality technology to build a virtual cultural relic restoration environment, the system provides real-time feedback on the splicing effect based on the user's selection and operation.

2. The virtual restoration system for damaged cultural relics according to claim 1 is characterized in that: The three-dimensional information of the cultural relic fragments is obtained by a three-dimensional scanning device and then preprocessed, including the following steps: Use 3D laser scanners and structured light scanners to conduct all-round scanning of the cultural relic fragments to obtain 3D point cloud information. Preprocess the collected raw data, including denoising, filtering and registration, and mark and record the damaged areas of the cultural relics.

3. The virtual restoration system for damaged cultural relics according to claim 1 is characterized in that: The method of extracting geometric features of cultural relic fragments from point cloud data by computational geometry technology, extracting texture features of cultural relic fragments by image processing technology, and classifying and splicing cultural relic fragments by combining machine learning algorithms to form a three-dimensional digital model includes the following steps: The geometric features of the cultural relic fragments, including shape, size and curvature, are extracted from the point cloud data using computational geometry technology. The texture features of the cultural relic fragments are extracted using image processing technology. The extracted geometric features and texture features are fused to form a comprehensive feature vector. The support vector machine is used as a classifier to train the comprehensive feature vector and build a classification model for the cultural relic fragments. The four-point algorithm is used to roughly align the fragments of the cultural relic. After determining the positional relationship between the fragments, the iterative closest point algorithm is used for precise alignment. The fragments of the cultural relic are assigned to multiple nodes for simultaneous feature extraction and splicing. The MPI message passing mechanism is used for communication and data synchronization between nodes. When all the fragments of the cultural relic are spliced ​​together, they are merged into a three-dimensional digital model of the entire cultural relic. The synthesized three-dimensional digital model is then verified to check whether there are any unspliced ​​or spliced ​​areas.

4. The virtual restoration system for damaged cultural relics according to claim 3 is characterized in that: The method of extracting texture features of cultural relic fragments by using image processing technology comprises the following steps: Use a high-resolution camera to capture images of cultural relics fragments at different angles, read and load the images of cultural relics fragments to be processed, convert the original images into grayscale images, apply filters to remove noise in the images, smooth the images and highlight edge features, convert the grayscale images into binary images, use edge detection algorithms to extract edge information in the images, and then perform morphological operations; The gray-level co-occurrence matrix is ​​used to extract the texture features of the image. The parameters for calculating the gray-level co-occurrence matrix, including direction and interval, are determined. Each pixel in the image is traversed. According to the selected direction and interval, the probability of occurrence of pixel pairs that meet the conditions is counted. The gray-level co-occurrence matrix is ​​normalized. The texture features of the processed cultural relic fragment image are extracted, including contrast, correlation and energy uniformity eigenvalues. Each pixel in the image is traversed to obtain its corresponding texture eigenvalue. The eigenvalue is converted into a corresponding color value using gray-level mapping. A texture feature map is constructed by assigning the converted color value to the position of the corresponding pixel in the original image.

5. The virtual restoration system for damaged cultural relics according to claim 1 is characterized in that: The method of using the ResNet framework in the convolutional neural network to build a model for identifying damaged areas and classifying the degree of damage to cultural relics by comprehensive damage degree values ​​includes the following steps: Collect image data of cultural relics fragments containing various damaged areas, annotate the collected images, perform data enhancement on the annotated images, select the ResNet framework in the convolutional neural network, and use the annotated data to train the neural network model. During the training process, adjust the model parameters through the back propagation algorithm to minimize the loss function value, deploy the trained model to the actual application environment, and use the deployed model to perform real-time recognition of newly collected cultural relics images to automatically detect damaged areas; Use a camera to record images of damaged areas, observe the edges of damaged areas, surface textures, distribution and width of cracks, extract area and density features of the identified damaged areas, and combine the ratio of the damaged area area to the normal area area and the ratio of the damaged area density to the normal area density to obtain a comprehensive damage degree value; The calculation formula of comprehensive disease severity value is: Among them, C is the comprehensive disease severity value, S damage is the area of ​​the damaged area, S normal is the area of ​​the normal region, D damage is the density of the damaged area, D normal is the density of the normal area; The degree of damage to cultural relics is divided into three levels according to the comprehensive damage degree value; If 0≤C<30%, it is defined as a mild disease degree; If 30% ≤ C < 70%, it is defined as a moderate disease degree; If 70% ≤ C, it is defined as a severe disease degree.

6. The virtual restoration system for damaged cultural relics according to claim 1 is characterized in that: The method of analyzing the existing data format and structure, defining data elements, and developing a data sharing platform using cloud computing technology includes the following steps: Identify the institutions and departments that will participate in sharing, determine the data needs and usage scenarios of each institution, analyze the existing data format and structure, define data elements, including data name, type, length and format, and specify the data exchange format, use cloud computing technology to develop a data sharing platform, build a data warehouse to store and manage data from different institutions, encrypt data during transmission, convert data from different sources into a unified format, establish a correspondence between the new and old data standards, and import the cleaned data into the sharing platform.

7. The virtual restoration system for damaged cultural relics according to claim 1 is characterized in that: The method of using virtual reality technology and augmented reality technology to build a virtual cultural relic restoration environment, and the system provides real-time feedback stitching effects according to the user's selection and operation, includes the following steps: A virtual cultural relic restoration environment is constructed using virtual reality technology and augmented reality technology, including a virtual workbench, tools and cultural relic fragment elements, providing a user interactive interface for splicing cultural relic fragments. The system provides real-time feedback on the splicing effect based on the user's selection and operation. After the splicing is completed, the user can export the virtually restored cultural relics in the form of a three-dimensional model and share it on social media.

Images