37 results about "Digital restoration" patented technology

The invention belongs to the technical field of digital restoration of computer vision and cultural heritage, and particularly relates to a traditional picture restoration method fusing low-resolution prior and efficient visual selection, which comprises the following steps: constructing a multi-source image data set, taking images in the multi-source image data set as high-resolution images, preprocessing the high-resolution images to obtain low-resolution images, and carrying out high-resolution priori and high-efficiency visual selection on the low-resolution images. The high-resolution image and the low-resolution image are respectively masked to generate simulated damage mask images, and the simulated damage mask images comprise a regular damage mask image and an irregular damage mask image; taking the multi-source image data set and the preprocessed multi-source image data set as training data, and training a multi-source image model; the dual-stage repair network comprises a coarse repair network and a fine repair network; according to the method, the problems of structural semantic loss, high priori information dependency and insufficient global and local coordination when an existing image restoration method is used for processing a complex scene and a large-range missing region are solved.

The invention belongs to the technical field of wall painting restoration, and discloses a damaged image automatic digital restoration method comprising the following steps: a data collection module collects geological data and image data; the three-dimensional structure scanning unit obtains a three-dimensional structure of the surface of the mural and space coordinates of a damaged area; the infrared imaging unit extracts a bottom layer pattern; the intelligent damage detection module locates a damage area; the multi-modal generative repair module complements a damaged physical form through a structural layer repair unit, and the repair scheme planning module plans an entity repair route and a material scheme; the damaged area color partitioning and boundary extraction module is used for performing partitioning and boundary extraction on the color for repairing the damaged area; and the entity repair execution module completes physical repair. According to the invention, through linkage of entity repair and digital achievements, the entity repair execution module performs physical repair based on a digital repair model, so that excessive dependence of traditional repair on artificial experience is broken through, and historical authenticity and artistic integrity of the mural are guaranteed by a scientific method.

The invention discloses a historical document restoration method and system based on implicit interpolation network enhancement. The method comprises the following steps: acquiring a damaged historical document image, a position mask image for identifying a to-be-restored area and a content guide image for guiding the semantic meaning of restored content; constructing a historical document repair model based on a de-noising diffusion probability model; training the historical document repair model by using a training data set containing damaged image and intact image pairs; and inputting the damaged historical document image to be repaired, the position mask image and the content guide image into the historical document repair model, and outputting the repaired historical document image. According to the method, the accuracy of the character content in the restoration area and the naturalness of the visual boundary are remarkably improved, and high-fidelity digital restoration of complex degraded literatures such as ancient books and inscription rubbings is realized.

The invention discloses an AI-based historical defective fresco digital restoration system, and relates to the technical field of AI restoration, a two-channel filter and a U-Net network are adopted to extract a fresco damage boundary, and edge detection is enhanced in combination with manual labeling; calculating a multi-scale Hurst index, and constructing a matrix quantization fractal feature change rule; dynamically generating a probabilistic grammar rule set based on the fractal incidence matrix, and defining a state transition probability function of texture growth; recursively executing texture generation driven by grammar rules from boundary points, and introducing random disturbance attenuating along with depth to simulate a stroke natural form; textures are generated through Laplacian pyramid layered correction, and the phase of a reference image is aligned in a Fourier domain to reserve structural consistency; and solving an affine transformation matrix with unchanged illumination in the transition zone, and globally optimizing color connection between the restoration area and the original wall painting. According to the method, the dynamic grammar rule base is established by analyzing the fractal features of the residual strokes, so that the generated texture follows creation logic, and the visual coherence of a recovery area is improved.

This invention provides a method and system for batch digital restoration of ancient books. The method includes: scanning the ancient book to be restored and generating multiple original images of the book based on all its pages; performing image preprocessing on the original images to generate multiple digital images of the book, including noise detection, normalization, and format conversion; generating spatial feature maps and content feature maps based on the digital images; determining a joint restoration decision based on the spatial and content feature maps; and batch restoring the multiple digital images of the book based on the joint restoration decision. By connecting edges, the discrete areas to be restored are organized into an organic whole, ensuring visual consistency and semantic coherence in the restoration results while improving the efficiency of batch restoration of ancient books.

A steel structure digital restoration device and method for digital twin construction, the device comprises a structure recognition module, the structure recognition module is configured to: the point cloud information of the acquired steel structure is executed in a virtual three-dimensional space according to its embedded space coordinate information based on the corresponding restoration operation of position, the total model field about the distribution of point cloud in space is acquired, and the restored point cloud information is in the form of a point model in the total model field;Based on the distribution of point cloud, the position with the best point model density is automatically selected, slicing is carried out to obtain the basic cross-section data of the structure;Based on the image in the basic cross-section data, the identification and judgment of the type of steel structure are executed, after correctly identifying the type of steel structure, the three-dimensional model fitting restoration of slicing is executed based on the identified type of steel structure.The present application can directly restore the three-dimensional pipeline model with high precision, and the engineering personnel can obtain the information of all steel structures in the whole factory by directly observing the three-dimensional steel structure model with sufficient rich details.

The invention proposes a real-time Thangka image semantic segmentation method and system based on improved PIDNet, and the method comprises the steps: carrying out the optimization of three branches of PIDNet for solving the problems of the fine texture of Thangka, and the wrong division, missing division and poor real-time performance caused by the interweaving of a main body and a background: sequentially embedding an ECA-Pag channel attention module and an LGFM local-global fusion module at the tail end of a P branch, details of the main esterm and the summator are strengthened; pConv3 mask partial convolution is introduced into a D branch, background noise is suppressed, and boundary definition is improved; the I branch keeps multi-scale pooling and provides global semantics. During reasoning, three-branch optimization features are fused after scale alignment and channel unification, and a pixel-level segmentation result is output through Softmax. According to the method, while the light weight of the network is kept, the confusion and edge fuzzy defects of the Thangka similarity summator are remarkably improved, and the method can be operated in real time, is suitable for Thangka digital repair, interactive labeling and culture research, and has the capability of migrating to other semantic segmentation tasks.

The application relates to the field of non-heritage culture 3D generation, and particularly relates to a Lingnan non-heritage culture 3D experience system based on generative artificial intelligence, which comprises a 3D digital restoration module, a virtual reality module, an interactive experience module, a culture activity preview module and an intelligent question and answer guide module; a 3D generation circulation reconstruction model is used for non-heritage culture 3D, the model cyclically uses a generation module based on 2D diffusion and a feedforward 3D reconstruction module in a multi-step diffusion process, and the diversity and texture consistency of 3D generation of non-heritage cultural articles are enhanced; a neural accumulator method is used to continuously converge low-resolution visual information between one frame and another frame, a neural super-resolution module is used, in the gaze point super-resolution reconstruction stage, a feature block grouping and processing strategy based on a quality mark is used, the calculation efficiency is improved, and the output image is significantly improved in terms of detail expressiveness and visual definition.

The invention discloses a digital intelligent oral cavity restoration process simulation system and method based on an oral cavity video, and belongs to the technical field of oral cavity digital restoration. The system comprises an oral cavity video acquisition module which is responsible for multi-angle video acquisition, preprocessing and oral cavity key point extraction; the modeling module is used for completing multi-view matching, three-dimensional basic model construction and defect identification and occlusion analysis; the three-dimensional repair design module is used for designing a three-dimensional tooth repair model in a digital environment; the virtual assembly detection module carries out assembly and biting fit detection to obtain fit indexes; the process simulation module adjusts process parameters based on the adaptation indexes and simulates and repairs the tooth state; and the repair scheme determination module iteratively optimizes the design parameters according to the state data and the ideal target to obtain a final optimized repair scheme. According to the scheme, the oral cavity three-dimensional model is reconstructed based on the multi-angle video, repair design is completed in combination with defect and occlusion analysis, a personalized scheme is generated through simulation evaluation and iterative optimization, and the repair precision is improved.

The application discloses a 3D printing resin jaw fork fixing method and system for complex dentition patients, relates to the technical field of oral digital restoration, and comprises the following steps: obtaining intraoral scanning data of the dentition patient, extracting mandibular dentition support points, undercut locking areas and dynamic occlusion interference boundaries based on the occlusion point trajectory, establishing a three-dimensional digital model of the mandibular full dentition and occlusion relationship with bearing stability weight and motion avoidance constraints, and performing coordinate unification and assembly gap matching according to the interface specifications of the scanning face bow, the occlusion recorder and the mandibular metal jaw fork to obtain a preliminary docking feature set. The 3D printing resin jaw fork fixing method and system for complex dentition patients improve the accuracy, repeatability and clinical operation efficiency of occlusion recording and mandibular motion analysis of complex dentition patients.

The application discloses an intelligent quality control system and method for manufacturing of full-welding pressure plate heat exchangers, and relates to the field of intelligent manufacturing, and comprises: a detection module for detecting heat exchanger surface structure parameters and constructing a heat exchanger three-dimensional model based on the heat exchanger surface structure parameters; and a derivation module for receiving the heat exchanger three-dimensional model, picking up incomplete models on the surface of the heat exchanger three-dimensional model, and respectively deriving bounding boxes of the incomplete models based on the picked-up incomplete models. The application realizes high-precision digital restoration of the heat exchanger structure by means of infrared distance measuring sensors densely arranged on a ring guide rail, combination of position and time stamp information, conversion of a large number of distance measuring results into three-dimensional space line segments, and splicing of the three-dimensional space line segments into a closed three-dimensional model, so as to provide quality control services for the heat exchanger.

This invention proposes a method for digital restoration of damaged historical buildings based on multi-source historical data and generative models. The method includes: acquiring historical images and mapping point clouds; generating a structurally perceptual representation that fuses image boundaries and spatial traces through multi-layer convolution and spatial gridding statistics; performing reverse inference using an inference model containing one-dimensional convolution and bidirectional gating units, and generating a complete structural expression by combining existing trace constraints; introducing this structural expression into a neural radiation field (NeRF) to construct an initial voxel field and scene guidance kernel controlled by the building's structural logic, achieving 3D reconstruction with spatial thickness and vertical differences; calculating the display reliability coefficient based on multi-view appearance consistency and geometric stability, and performing differentiated rendering and multi-scene applications on the model. This invention achieves integrated restoration from trace alignment and structural inference to reliable display, effectively solving the problem of missing geometric and structural logic in the digital restoration of damaged historical buildings.

The application discloses an ancient armor digital restoration system and method based on three-dimensional modeling, relates to the technical field of digital restoration, and comprises the following steps: extracting local thickness and defining a neutral point, and forming a residual part neutral surface by all the neutral points; screening the nearest neighbor set of reliable areas for each damaged area, and defining a target curvature field; defining a reliable connection boundary according to a displacement field, performing rigid body registration on the reliable connection boundaries of different residual parts, screening connection relationship candidates, introducing a binary selection variable to establish a global constraint, defining a global assembly topology set for the connection relationship candidates, and performing boundary closure and difference set correction to form a restoration model. The application combines the use of the geodetic distance nearest neighbor, local coordinate system principal curvature solving and target curvature propagation, can keep the curvature transmission along the residual part body expansion path when there are fractures, curling and edge missing, and avoids the cross-layer misfit caused by the Euclidean proximity.

This invention discloses a method for the collaborative restoration of structure and texture in Dunhuang murals based on structural uncertainty, belonging to the field of image restoration technology. The method includes: a structure prediction step, which generates a structure reconstruction result through an encoder-decoder architecture based on codebook quantization and simultaneously outputs pixel-level structural uncertainty parameters; a texture reconstruction step, which models the texture completion process as texture flow propagation on a graph structure and adaptively adjusts the propagation range and constraint strength according to the structural uncertainty parameters; and a fusion step, which fuses structural features and texture features at multiple scales and dynamically adjusts the fusion strength according to the structural uncertainty parameters to generate the final restoration result. This invention achieves effective linkage between structure prediction and texture reconstruction through structural uncertainty parameters, suppressing the transmission of structural errors to the texture generation process when there is large-area damage or severe information loss, while maintaining overall consistency and natural detail, making it suitable for the digital restoration of Dunhuang murals.

The invention relates to the technical field of three-dimensional data scanning, and discloses a high-fidelity cultural heritage three-dimensional data scanning system. The system comprises the following modules: a multi-source data acquisition module, which is used for obtaining original data of cultural heritage in an omnibearing and multi-dimensional manner; the data processing and geometric modeling module is used for converting the original data into an accurate and complete initial three-dimensional model and intelligently diagnosing the geometric integrity of the initial three-dimensional model; the intelligent geometric repair and coordination analysis module is used for intelligently reasoning and reconstructing the original style and appearance of the damaged area to realize digital recovery; the texture and material high-fidelity restoration module is used for giving vivid color and material feeling to the geometric model; and the restoration guidance data output module is used for converting the high-fidelity digital model into a specific scheme capable of guiding an actual restoration action. According to the method, high-precision three-dimensional data acquisition, damage geometry intelligent reasoning and material high-fidelity restoration are integrated, and accurate guidance is provided from data acquisition to physical restoration.

This invention relates to the field of digital preservation and restoration of cultural relics, specifically a method for extracting and digitally restoring ancient artifact patterns based on point cloud models. It includes modules for manifold frequency domain mapping, spectral domain signal separation, topological skeleton repair, and Poisson fusion generation. The system constructs a Laplace-Beltramm operator to map the point cloud to the frequency domain, uses a bandpass filter to remove weathering noise and enhance the pattern layer, extracts the pattern skeleton to identify breakpoints, and combines a knowledge graph for non-rigid registration to generate a deformation guidance field. Its core is to use Poisson fusion to fill in missing information and output a restoration confidence heatmap, effectively solving the problem of aliasing between weathering noise and weak feature signals, achieving micron-level high-fidelity restoration.

The invention discloses a digital jaw position recording and occlusal relationship collecting device and method, and relates to the technical field of oral medicine. Comprising the following steps: acquiring upper and lower jaw tissue surface data of a to-be-detected person, generating an oral cavity model of the to-be-detected person, performing three-dimensional data design based on the oral cavity model to obtain a tray model, importing the tray model into 3D printing software to obtain a personalized occlusion tray, wearing the tray by the to-be-detected person, and obtaining a static jaw position relationship by utilizing a three-dimensional scanner. The method comprises the following steps: acquiring opening and closing, lateral and front-back movement track data by using a lower jaw movement analysis system to form acquisition data, preprocessing the acquisition data, and inputting the preprocessed data into digital restoration / orthodontic design software to carry out subsequent digital design and analysis of a restoration / orthodontic appliance. According to the invention, high-precision jaw position recording and occlusal relationship acquisition are realized, data can be seamlessly jointed with a digital design process, the prosthesis adaptation rate and the treatment efficiency are improved, and the method has remarkable clinical application value and popularization prospect.

This invention provides a method for restoring the structure of a diagnosis and treatment flowchart based on a multimodal large model. The method includes: preprocessing the diagnosis and treatment flowchart; establishing a standardized structure format and prompt templates for the flowchart; designing a graph structuring loss function for the multimodal large model; fine-tuning the multimodal large model to restore the flowchart structure; inputting the flowchart and prompt templates into the fine-tuned model, outputting structured content conforming to the flowchart format specifications; restoring the structured content to a standardized, clear flowchart; and storing the structured content output by the multimodal large model, the edited structured content, and the corresponding flowchart in a database. This invention can automatically extract the node content and its connections from the input diagnosis and treatment flowchart image and convert it into a standardized, clear flowchart structure, achieving digital restoration of the diagnosis and treatment process.

The present disclosure relates to a display method, device and storage medium of a virtual jaw frame. The display method of the virtual jaw frame comprises: acquiring first facial three-dimensional data, second facial three-dimensional data, tooth three-dimensional data and radiographic three-dimensional data of a patient; aligning the first facial three-dimensional data and the radiographic three-dimensional data based on the tooth three-dimensional data; extracting facial feature points from the second facial three-dimensional data and tooth arrangement feature points from the tooth three-dimensional data or the radiographic three-dimensional data; constructing a reference axis and a plurality of reference planes of the patient according to the facial feature points and the tooth arrangement feature points; and displaying a facial three-dimensional model and a virtual jaw frame of the patient on a display interface, wherein the positional relationship between the virtual jaw frame and the facial three-dimensional model is determined based on the reference axis and the plurality of reference planes. The method provided by the present disclosure can accurately transfer the occlusion relationship of the patient to the virtual jaw frame, thereby providing a reliable basis for digital restoration.

The invention discloses a Buddha figure digital restoration and color restoration system based on deep learning in the technical field of image processing. The system comprises an image acquisition module used for acquiring a digital image of a Buddha figure; the semantic segmentation module is used for carrying out semantic segmentation on Buddha figure digitization; the knowledge graph construction module is used for constructing Buddha statue colors and forms; the deep learning network module is used for analyzing a special area of a Buddhist image corresponding part in each semantic area; the structure repairing module is used for repairing different Buddhist statue structures; the texture and color repairing module is used for repairing the texture and color areas of the Buddha statue; the fusion module is used for outputting a completely repaired Buddhist image; and the environment detection module is used for establishing an aging prediction model, assisting in correcting aging related information in the knowledge graph and providing an environment reference basis for color restoration and texture restoration. According to the method, the auxiliary environment prediction factors are introduced in the Buddha statue repairing process, the repairing parameters of the Buddha statue are corrected, and the repairing accuracy is further improved.

This invention discloses a novel digital restoration method for jar-shaped cultural relics, comprising: digitizing cultural relic fragments using a scanner; simulating the assembly of the fragments using scanning or design software and checking for seamless data connection; measuring curvature and assembling the fragments based on the cultural relic data; simulating the original cultural relic based on its restored position and designing an outer shell to enclose the fragments and a padding located on the upper inner side of the jar-shaped cultural relic, wherein the outer shell serves as the main load-bearing component and positioning grooves for positioning the cultural relic fragments are provided on both the outer shell and the padding; printing the outer shell, the padding, and the simulated lost fragments using 3D printing; and manually assembling and restoring the jar-shaped cultural relic fragments using the outer shell and the padding. This invention eliminates the need for any adhesives and allows for the replacement and aesthetic restoration of cultural relics with the support of transparent supports, enabling the display of the cultural relics.

The invention belongs to the field of cultural relic digital protection, relates to a wall painting digital full-automatic repairing method fusing a feature extractor, and aims to complete high-quality repairing of wall paintings. The method comprises the following steps: S1, constructing and training a mural style feature extractor which is used for extracting style feature vectors representing mural artistic characteristics from a mural image; s2, defect area detection is carried out on the mural image based on the detection model, and bounding box information of a defect area is output; s3, on the basis of the segmentation model, generating a binary segmentation mask of the defect area by taking bounding box information as a prompt; and S4, based on a diffusion restoration model, restoring the defect area in combination with the style feature vector and the binary segmentation mask, and outputting a restored mural image.

The invention discloses an intelligent museum digital vision multi-dimensional virtual interaction system, and belongs to the field of virtual interaction. Comprising a multi-source data acquisition module, a three-dimensional space modeling module, a behavior emotion recognition module, an environment illumination perception module, a cultural relic dynamic reconstruction module, a digital restoration enhancement module, a personalized narrative generation module, a multi-dimensional immersion rendering module, a user collaborative interaction module, a user mode analysis module and an operation decision support module. The method can generate a virtual museum space with strong reality sense and accurate structure, comprehensively captures moods and interests of audiences, realizes accurate behavior and psychological analysis, enhances immersion and interactivity, can generate high-fidelity rendering under any illumination condition, meets the personalized lighting requirements of the audiences, can realize thousand-person and thousand-face explanation experience, and improves the interestingness and interestingness of the audiences. And the passive visiting mode of the traditional museum is broken.

The invention discloses a wall painting digital restoration method based on structure guidance and a state space model, which comprises the following steps: constructing an edge restoration model with a mask image, a damaged edge image and a damaged grey-scale image as input and an edge restoration image as output, the method is characterized in that the method comprises the following steps: 1, sequentially constructing a sample convolution block, three down-sampling convolution blocks, eight double-path Mamba modules, three up-sampling convolution blocks, one output convolution block and one discriminator; constructing a main body repair model of which the input is the damaged mural picture, the mask picture and the edge repair picture and the output is the main body repair picture, and constructing a detail repair model of which the input is the mask picture, the main body repair picture and the edge repair picture and the output is the final repair picture; applying the trained model to a to-be-repaired actual mural; according to the method, the prior structure information of the mural is utilized in a multi-scale mode to guide the repairing process, the line structure and texture detail information of the mural are better restored, the mural repairing effect is remarkably improved, and the generated image is closer to a real mural.

The invention relates to the technical field of data processing, and discloses a digital restoration scene recognition method and device, and the method comprises the steps: carrying out the color analysis of received original point cloud data, and recognizing candidate mark points with a preset color range; performing local geometric analysis on the candidate mark points to obtain host geometry surface information; on the basis of the host geometry surface information, whether an attachment relationship exists between the candidate mark point groups and the host geometry surface is judged, and the mark point groups with the attachment relationship are identified; executing point cloud data filtering operation based on the identification information of the identified mark point group, and reserving the identified mark point group; and associating the semantic information carried by the reserved mark point group with the host geometry, and generating or updating a digital scene model containing the geometric information and the semantic information. According to the invention, the tiny temporary mark bearing the key semantic information in the point cloud data is effectively identified and reserved; and the noise is prevented from being misjudged as noise to be lost in the data processing process.

The application discloses a kind of three-dimensional matching and digital repair method of cultural relic fragments, it is related to cultural relic repair technical field, the method comprises the following steps: obtaining the point cloud data of each fragment belonging to the same cultural relic;Using Poisson surface reconstruction method, the triangular mesh model of each fragment is constructed;The corrosion edge in the triangular mesh model of each fragment is repaired;The triangular mesh model of each fragment is registered after edge repair is completed using multi-level matching algorithm, and the optimal splicing position and attitude parameter of each fragment are obtained;According to the optimal splicing position and attitude parameter of each fragment, the triangular mesh model of each fragment after edge repair is completed is spliced, and the initial cultural relic digital model is obtained;The missing area of initial cultural relic digital model is repaired, and the repaired cultural relic digital model is obtained.The application not only improves the accuracy and efficiency of cultural relic repair, but also reduces the risk of manual operation, while ensuring the traceability and reversibility of repair steps.

The application discloses a historical document restoration method and system based on implicit interpolation network enhancement, and the method comprises the following steps: acquiring a damaged historical document image, a position mask image for identifying a region to be restored, and a content guide image for guiding the semantic content of the restoration; constructing a historical document restoration model based on a denoising diffusion probability model; training the historical document restoration model by using a training data set containing pairs of damaged images and intact images; inputting the damaged historical document image to be restored, the position mask image and the content guide image into the historical document restoration model, and outputting a restored historical document image. The application significantly improves the accuracy of the character content of the restoration region and the naturalness of the visual boundary, and realizes high-fidelity digital restoration of complex degraded documents such as ancient books, inscriptions and rubbing sheets.

This application relates to the field of artificial intelligence and discloses a method for extracting cultural elements based on intelligent recognition. It aims to address the problems of low segmentation accuracy, broken or adhered boundaries in cultural heritage images caused by low contrast, complex structure, and scarce annotations. The method includes: acquiring a cultural image and generating an initial coarse mask using pre-trained U-Net++; inputting the image and mask into a conditional generator to generate a high-fidelity reconstructed image; calculating a local adversarial loss using a PatchGAN discriminator; using the gradient of this loss to back-update the U-Net++ segmentation head parameters; iteratively optimizing until the mask converges, and outputting a refined binary mask. This application, through a generative adversarial self-supervised mechanism, significantly reduces the dependence on newly added pixel-level annotations, guiding the model to learn continuous natural boundaries that conform to human cognition, while requiring only a small number of pixel-level annotations for network initialization. The resulting mask can be directly used for vectorization, 3D reconstruction, and digital restoration, supporting high-fidelity digital applications of cultural heritage.