Wood structure cultural relic ancient building restoration modeling reconstruction method based on digital technology
Through multi-sensor fusion and intelligent algorithms, problems such as data acquisition accuracy and structural modeling difficulty in restoring wooden structures in ancient buildings are solved, and high-precision three-dimensional model construction and repair strategy optimization are realized, which improves the scientificity and efficiency of repair work. Ensures accuracy and aesthetic consistency of repair effects.
Patent Information
- Application Number
- CN202411203658.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-08
Smart Images

Figure CN118916971A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of ancient building restoration, and in particular to a digital technology-based restoration modeling and reconstruction method for wooden structure cultural relics ancient buildings. Background Art
[0002] China is an ancient country with a long history. In the era when metallurgical technology was lacking in ancient China, the relative lack of metallurgical technology in its ancient civilization gave rise to a unique style of architectural art, dominated by exquisite wooden structures. With extraordinary wisdom and ingenuity, the ancients skillfully used the natural properties of wood materials and connection wisdom to create exquisite structural connection methods such as brackets, mortise and tenon joints, tooth joints and pin joints. These techniques not only show extremely high aesthetic value, but also create architectural miracles with stable structures and lasting newness.
[0003] However, as time goes by and wind and rain erode, even these ingenious wooden structures cannot escape the test of time, and their surfaces and even internal structures gradually show varying degrees of damage and aging. Faced with this challenge, traditional restoration methods tend to directly replace damaged wood. Although this method can restore the original appearance of the building to a certain extent, it is regrettable that this method has limited retention of the original materials, high restoration costs, and a long construction period.
[0004] In view of this, it is particularly urgent and important to explore an innovative in-situ restoration technology for ancient building wooden structures.
[0005] Although modern technology has great advantages in protecting and restoring ancient buildings, there are still some challenges and problems in existing technologies. The following are several major technical issues:
[0006] 1. Accuracy and completeness of data collection, difficulty in scanning complex structures: Ancient wooden buildings usually have complex geometric shapes and details. Sometimes in narrow or obstructed areas, laser scanning and photogrammetry are difficult to obtain comprehensive and accurate data, resulting in the possibility of missing details in the model.
[0007] Material reflectivity and transparency issues: The optical properties of the wood surface (such as reflectivity, texture differences) will affect the effect of laser scanning, resulting in data accuracy, especially when dealing with aged, corroded or unevenly coated wood.
[0008] Traditional methods of restoring wooden cultural relics usually rely on a single or a small number of sensors for data collection, resulting in incomplete information being obtained. Especially in complex restoration scenarios, it is difficult to ensure the accuracy and completeness of the data.
[0009] Existing technologies, such as conventional laser scanning or photogrammetry, can collect three-dimensional data, but they still rely on manual or simple algorithms for data fusion and processing, making it difficult to achieve intelligent data fusion effects. Especially in complex restoration scenarios, the model may be incomplete or inaccurate due to inaccurate data.
[0010] 2. Complexity of digital modeling and difficulty in modeling structural connections: The key part of ancient wooden buildings is often their complex mortise and tenon structures. These connection methods are not only complex and diverse, but also require extremely high precision and meticulous work when modeling. Existing modeling software sometimes finds it difficult to fully simulate these details.
[0011] In traditional point cloud processing methods, point cloud denoising and surface reconstruction usually use classic algorithms such as RANSAC (Random Sampling Consistency) or Poisson surface reconstruction. However, these methods may cause loss of details or incomplete surface reconstruction when dealing with complex wooden structure surfaces. Current restoration guidance usually relies on drawings or two-dimensional images, without achieving intuitive integration with the real environment.
[0012] Insufficient historical information: Due to the age of some ancient buildings, there is a lack of complete historical records and design drawings. Therefore, speculation and assumptions need to be made during digital modeling, which may result in the model being inconsistent with the actual structure.
[0013] 3. Limitations of structural analysis, uncertainty in material properties: The physical properties of wood, such as density, strength, and elastic modulus, will change due to aging, humidity changes, and environmental influences. These changes are difficult to accurately consider in structural analysis, which may cause the simulation results to deviate from the actual situation.
[0014] Simulation of complex loading conditions: Ancient buildings are often subjected to complex loading conditions, including long-term weathering, earthquakes and other external forces. The simulation accuracy of these conditions in structural analysis still needs to be improved.
[0015] 4. The difference between virtual restoration and actual restoration, and the matching problem between restoration materials and raw materials: Although digital models can accurately simulate the restoration process, in actual restoration, it is still challenging to find materials that completely match the texture, grain, and color of the original wood, which may affect the overall effect and aesthetic consistency of the restoration.
[0016] Differences between simulation and actual operation: Repair simulation in digital technology may be effective under ideal conditions, but in actual operation, unforeseen problems may occur, such as process deviations or environmental changes, resulting in repair effects that are not as good as expected. In existing repair methods, the optimization of repair strategies mostly relies on manual experience, making it difficult to achieve real-time feedback and adjustments.
[0017] 5. Technical challenges of long-term monitoring and protection, accuracy and durability of sensors: Although sensors can be used to monitor the health status of buildings in real time, these sensors need to work in harsh environments for a long time, and their accuracy and durability are challenged, especially in high temperature, humidity changes or insect infestations.
[0018] Data processing and analysis: Real-time monitoring generates a large amount of data, which needs to be processed and analyzed in a timely manner to make accurate judgments. Existing data processing systems may not be able to cope with such a large amount of data, resulting in delayed or inaccurate analysis results.
[0019] 6. Imperfect technical standards and specifications, and lack of industry standards: At present, restoration methods combined with digital technology lack unified standards and specifications in the industry, which leads to large differences in technical applications between different projects and makes it difficult to form a universally recognized technical flow. Summary of the invention
[0020] In order to solve one of the above technical problems, a restoration modeling and reconstruction method for wooden structure cultural relics and ancient buildings based on digital technology is provided, and a restoration modeling and reconstruction method based on multi-sensor fusion and intelligent algorithm is provided. This method can accurately collect building data in a complex environment, and generate a high-precision three-dimensional model through advanced modeling and analysis technology, thereby providing a scientific basis for subsequent restoration work.
[0021] In order to achieve the above purpose, the technical solution adopted by the present invention is:
[0022] A first aspect: A digital technology-based restoration modeling and reconstruction method for wooden structure cultural relics and ancient buildings, comprising the following steps:
[0023] Acquire three-dimensional data of the cultural relics and ancient buildings to be restored, collect information of the three-dimensional data based on multiple data acquisition devices and deployments, pre-process the data in real time according to the embedded processing unit, input the data collected by the multiple data acquisition devices and deployments into a three-dimensional coordinate system in response to a multi-view stereo matching algorithm, and generate a three-dimensional point cloud model;
[0024] The three-dimensional point cloud model automatically identifies and corrects errors in the data collection of the three-dimensional point cloud model based on an adaptive correction algorithm of machine learning, performs denoising on the preliminary point cloud data according to a clustering algorithm, and automatically identifies the mortise and tenon structure in the point cloud according to a pattern recognition algorithm to construct a three-dimensional mortise and tenon model;
[0025] Based on historical data and material principles, the aging process of wood is simulated, finite element analysis is used for mechanical simulation, the mechanical behavior of wood structures is predicted, and three-dimensional point cloud models are used to test different repair schemes and simulate the repair process;
[0026] A sensor network is deployed in the restored building to monitor the moisture, temperature, stress and deformation of the wood. An artificial intelligence algorithm is used to analyze long-term monitoring data and automatically assess the health of the wooden structure.
[0027] Preferably: the multiple data acquisition devices and deployments specifically include:
[0028] 3D laser scanners: Multiple high-precision 3D laser scanners are deployed around and inside the cultural relics and ancient buildings to capture the overall geometric structure of the cultural relics and ancient buildings;
[0029] High-resolution cameras: Install high-resolution cameras on the exterior of cultural relics and ancient buildings to capture surface details and textures of cultural relics and ancient buildings;
[0030] Infrared scanner: Detect cavities, cracks and aging inside wooden structures using infrared scanners;
[0031] Environmental sensors: Place temperature and humidity sensors inside cultural relics and ancient buildings, and record environmental parameters during scanning.
[0032] Preferably: the 3D laser scanner, high-resolution camera, infrared scanner and environmental sensor communicate via wireless network or optical fiber to achieve synchronous data collection of multiple sensors;
[0033] Based on the multi-view stereo matching algorithm, laser scanning data, photogrammetry data and infrared scanning data are input into the three-dimensional coordinate system, and a three-dimensional point cloud model is generated through feature point matching and rigid body transformation.
[0034] Preferably: the real-time data preprocessing is performed according to the embedded processing unit, and the preprocessing includes noise filtering, error calibration and point cloud stitching.
[0035] Preferably, the denoising process is performed on the preliminary point cloud data according to the clustering algorithm to remove isolated points and noise points, and the mortise and tenon structure in the point cloud is automatically identified according to the pattern recognition algorithm to construct the mortise and tenon three-dimensional model, specifically including:
[0036] The preliminary point cloud data is denoised using a density-based clustering algorithm to remove isolated and noisy points, and the point cloud is streamlined using a voxel grid simplification algorithm;
[0037] According to the Poisson surface reconstruction algorithm, the simplified point cloud data is converted into a continuous three-dimensional surface model;
[0038] Construct a 3D model of mortise and tenon joints, and automatically identify mortise and tenon joint structures with typical features in the point cloud based on historical data and traditional craftsmanship through pattern recognition algorithms;
[0039] In response to the mortise and tenon structure with typical features in the point cloud, a high-resolution texture is mapped onto the surface of the mortise and tenon three-dimensional model according to the photogrammetry data;
[0040] The optical BRDF model is used to simulate the optical properties of wood and reproduce the real material of the wood structure.
[0041] Preferred: Based on historical data and material principles, simulate the aging process of wood, use finite element analysis for mechanical simulation, predict the mechanical behavior of wood structures, test different repair schemes with a 3D point cloud model, and simulate the repair process, including:
[0042] Digital twin technology, real-time simulation of the physical behavior and environmental response of wooden structure cultural relics during the restoration process;
[0043] The environmental response includes ambient temperature, ambient humidity, earthquake and wind load, and the expression formula is:
[0044] T virtual (t) = M (u real (t), p external );
[0045] Among them, T virtual (t): virtual repair state at time t;
[0046] u real (t): displacement and deformation in the actual restoration state;
[0047] p external : Environmental response.
[0048] Preferably, the method of analyzing long-term monitoring data by an artificial intelligence algorithm to automatically assess the health of the wooden structure includes:
[0049] The repair strategy is adjusted based on the automatic assessment of the health status of the wood structure, expressed as:
[0050]
[0051] in,
[0052] θ: Parameters of the machine learning model for wooden structure artifacts;
[0053] G AR (t i ): time t i Artificial intelligence algorithm guidance information;
[0054] T virtual (t i ): time t i The status of the virtual restoration of wooden structure heritage;
[0055] L(θ): loss function that measures the difference between the guidance of the AI algorithm and the virtual restoration.
[0056] Preferred: The method includes: during the actual repair construction process, using sensors and real-time monitoring systems, providing real-time feedback through intelligent systems, guiding construction personnel to make corrections, and using AR equipment at the construction site to project digital models and repair plans in real time.
[0057] Preferably, the intelligent system is expressed as:
[0058] G AR =R(S reconstruct , I user );
[0059] Among them, G AR : Guidance information generated by the AR system;
[0060] R: rendering function of the AR system;
[0061] I user : Information or operation entered by the user.
[0062] The second aspect: A computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor implements the digital technology-based method for restoring, modeling and reconstructing wooden structure cultural relics and ancient buildings.
[0063] Compared with the prior art, the present invention has the following beneficial effects:
[0064] The present invention collects data through a variety of sensors (such as laser scanning, infrared imaging, and optical cameras) and uses artificial intelligence algorithms for intelligent fusion, which can capture the detailed features of cultural relics more accurately.
[0065] Multimodal data fusion technology can obtain more comprehensive information about wooden structure cultural relics than a single sensor method, significantly improving the accuracy of the three-dimensional model and providing more accurate basic data for subsequent restoration and reconstruction.
[0066] A deep learning model is used to perform point cloud denoising and a neural network is used for surface reconstruction, which can better maintain the details of the wooden structure and automatically adapt to different types of noise and surface complexity.
[0067] Deep learning algorithms can learn the characteristics of surface details and noise of cultural relics through a large amount of training data. They have higher denoising capabilities and surface reconstruction effects than traditional algorithms. Especially when facing complex wooden structure surfaces, they can retain more details and generate smoother and more natural three-dimensional surfaces.
[0068] Applying digital twin technology to the process of cultural relics restoration can predict the effect of restoration operations in advance and reduce errors in actual operations by simulating the physical behaviors and environmental responses during the restoration process in real time.
[0069] This allows restorers to test different restoration plans in a virtual environment and select the best plan for actual operation, greatly reducing the uncertainty and risk in the restoration process. By simulating the response of cultural relics under different environmental conditions, the structural integrity and historical value of cultural relics can be better protected.
[0070] Through AR technology, restoration personnel can obtain virtual guidance information in real time and project restoration suggestions directly onto the cultural relics during the operation, improving the accuracy and intuitiveness of the operation.
[0071] Machine learning algorithms can automatically optimize repair strategies based on real-time data and repair effects.
[0072] By minimizing the difference between AR guidance and virtual repair status, each step of the repair process can be optimized to ensure the accuracy and stability of the final repair effect. This intelligent feedback and adaptive optimization mechanism makes the repair process more intelligent and refined.
[0073] Machine learning-driven adaptive optimization can automatically adjust the restoration strategy to ensure the best effect of each step of the operation, greatly improving the accuracy, efficiency and safety of the restoration of wooden structure cultural relics, and providing a more advanced and scientific restoration method. BRIEF DESCRIPTION OF THE DRAWINGS
[0074] Figure 1 The present invention is a flow chart of the digital technology-based method for restoring and modeling ancient wooden structures. DETAILED DESCRIPTION
[0075] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.
[0076] Embodiment 1:
[0077] Reference Figure 1 As shown, a digital technology-based restoration modeling and reconstruction method for wooden structure cultural relics and ancient buildings includes the following steps:
[0078] Acquire three-dimensional data of the cultural relics and ancient buildings to be restored, collect information of the three-dimensional data based on multiple data acquisition devices and deployments, pre-process the data in real time according to the embedded processing unit, input the data collected by the multiple data acquisition devices and deployments into a three-dimensional coordinate system in response to a multi-view stereo matching algorithm, and generate a three-dimensional point cloud model;
[0079] Specifically: the multiple data collection devices and deployments specifically include:
[0080] 3D laser scanners: Multiple high-precision 3D laser scanners are deployed around and inside the cultural relics and ancient buildings to capture the overall geometric structure of the cultural relics and ancient buildings;
[0081] High-resolution cameras: Install high-resolution cameras on the exterior of cultural relics and ancient buildings to capture surface details and textures of cultural relics and ancient buildings;
[0082] Infrared scanner: Detect cavities, cracks and aging inside wooden structures using infrared scanners;
[0083] Environmental sensors: Place temperature and humidity sensors inside cultural relics and ancient buildings, and record environmental parameters during scanning.
[0084] Specifically: the 3D laser scanner, high-resolution camera, infrared scanner and environmental sensor communicate via wireless network or optical fiber to achieve synchronous data collection of multiple sensors;
[0085] Based on the multi-view stereo matching algorithm, laser scanning data, photogrammetry data and infrared scanning data are input into the three-dimensional coordinate system, and a three-dimensional point cloud model is generated through feature point matching and rigid body transformation.
[0086] Specifically: the real-time data preprocessing is performed according to the embedded processing unit, and the preprocessing includes noise filtering, error calibration and point cloud stitching.
[0087] The three-dimensional point cloud model automatically identifies and corrects errors in the data collection of the three-dimensional point cloud model based on an adaptive correction algorithm of machine learning, performs denoising on the preliminary point cloud data according to a clustering algorithm, and automatically identifies the mortise and tenon structure in the point cloud according to a pattern recognition algorithm to construct a three-dimensional mortise and tenon model;
[0088] Specifically: the 3D laser scanner, high-resolution camera, infrared scanner and environmental sensor communicate via wireless network or optical fiber to achieve synchronous data collection of multiple sensors;
[0089] Based on the multi-view stereo matching algorithm, laser scanning data, photogrammetry data and infrared scanning data are input into the three-dimensional coordinate system, and a three-dimensional point cloud model is generated through feature point matching and rigid body transformation.
[0090] Specifically: the real-time data preprocessing is performed according to the embedded processing unit, and the preprocessing includes noise filtering, error calibration and point cloud stitching.
[0091] Specifically: the preliminary point cloud data is denoised according to the clustering algorithm to remove isolated points and noise points, the mortise and tenon structure in the point cloud is automatically identified according to the pattern recognition algorithm, and the mortise and tenon three-dimensional model is constructed, specifically including:
[0092] The preliminary point cloud data is denoised using a density-based clustering algorithm to remove isolated and noisy points, and the point cloud is streamlined using a voxel grid simplification algorithm;
[0093] According to the Poisson surface reconstruction algorithm, the simplified point cloud data is converted into a continuous three-dimensional surface model;
[0094] Construct a 3D model of mortise and tenon joints, and automatically identify mortise and tenon joint structures with typical features in the point cloud based on historical data and traditional craftsmanship through pattern recognition algorithms;
[0095] In response to the mortise and tenon structure with typical features in the point cloud, a high-resolution texture is mapped onto the surface of the mortise and tenon three-dimensional model according to the photogrammetry data;
[0096] The optical BRDF model is used to simulate the optical properties of wood and reproduce the real material of the wood structure.
[0097] Based on historical data and material principles, the aging process of wood is simulated, finite element analysis is used for mechanical simulation, the mechanical behavior of wood structures is predicted, and three-dimensional point cloud models are used to test different repair schemes and simulate the repair process;
[0098] Specific: Based on historical data and material principles, simulate the aging process of wood, use finite element analysis for mechanical simulation, predict the mechanical behavior of wood structures, test different repair schemes with a 3D point cloud model, and simulate the repair process, including:
[0099] Digital twin technology, real-time simulation of the physical behavior and environmental response of wooden structure cultural relics during the restoration process;
[0100] The environmental response includes ambient temperature, ambient humidity, earthquake and wind load, and the expression formula is:
[0101] T virtual (t) = M (u real (t), p external );
[0102] Among them, T virtual (t): virtual repair state at time t;
[0103] u real (t): displacement and deformation in the actual restoration state;
[0104] p external : Environmental response.
[0105] A sensor network is deployed in the restored building to monitor the moisture, temperature, stress and deformation of the wood. An artificial intelligence algorithm is used to analyze long-term monitoring data and automatically assess the health of the wooden structure.
[0106] Specifically: The long-term monitoring data is analyzed by artificial intelligence algorithms to automatically assess the health status of wooden structures, including:
[0107] The repair strategy is adjusted based on the automatic assessment of the health status of the wood structure, expressed as:
[0108]
[0109] in,
[0110] θ: Parameters of the machine learning model for wooden structure artifacts;
[0111] G AR (t i ): time t i Artificial intelligence algorithm guidance information;
[0112] T virtual (t i ): time t i The status of the virtual restoration of wooden structure heritage;
[0113] L(θ): loss function that measures the difference between the guidance of the AI algorithm and the virtual restoration.
[0114] 3D laser scanner: Multiple high-precision 3D laser scanners are deployed at key locations around the building to capture the overall geometric structure of the building. The scanners should be able to cover all angles of the building to reduce data occlusion and blind spots.
[0115] High-resolution cameras: Install high-resolution cameras at strategic locations to capture building surface details and textures. These cameras should support multi-angle shooting and be able to be effectively integrated with laser scanning data.
[0116] Infrared scanner: Infrared scanners are introduced to detect cavities, cracks and aging inside wooden structures, making up for the shortcomings of laser scanning and ordinary photography in detecting internal structures.
[0117] Environmental sensors: Temperature and humidity sensors are arranged to record environmental parameters during scanning and provide a reference for post-processing of data.
[0118] Through wireless networks or optical fiber communications, data from multiple sensors can be collected synchronously. During each collection, the data timestamps between sensors are aligned to ensure that data from different sensors can be fused in the same coordinate system.
[0119] During the data acquisition process, an embedded processing unit is used to perform real-time data preprocessing, including noise filtering, error calibration, and preliminary point cloud stitching. This step can effectively reduce the complexity of later data processing.
[0120] Using a multi-view stereo matching algorithm, the data from laser scanning, photogrammetry, and infrared scanning are fused in a unified 3D coordinate system. Through feature point matching and rigid body transformation, a preliminary, high-precision 3D point cloud model is generated.
[0121] Introducing an adaptive correction algorithm based on machine learning to automatically identify and correct errors in data acquisition, especially dealing with noise and missing data caused by optical properties (such as reflectivity).
[0122] The preliminary point cloud data is denoised using a density-based clustering algorithm to remove isolated points and noise points. The point cloud is then streamlined using a voxel grid simplification algorithm to reduce the amount of data and retain key structural features.
[0123] The Poisson surface reconstruction algorithm is used to transform the simplified point cloud data into a continuous three-dimensional surface model. In order to retain the details of the wooden structure, the density of the surface mesh should be appropriately increased during the reconstruction process.
[0124] Using pattern recognition algorithms, we can automatically identify typical mortise and tenon structures in point clouds. Combining historical data and traditional craftsmanship, we can build a detailed 3D model of mortise and tenon joints to ensure the authenticity and historical accuracy of the model.
[0125] Texture Mapping and Material Reproduction: Using photogrammetry data, high-resolution textures are mapped onto the surface of 3D models.
[0126] The optical BRDF model is used to simulate the optical properties of wood and reproduce the real material feel of the wood structure.
[0127] Based on historical data and material science principles, the aging process of wood and damage caused by environmental factors (such as humidity and temperature) are simulated. Finite element analysis (FEA) is used for mechanical simulation to predict the mechanical behavior of wooden structures under long-term loads.
[0128] Test different repair solutions in the 3D model to simulate structural deformation and stress concentration problems that may occur during the repair process. Select the best repair solution through multiple simulations.
[0129] Before repair, a multi-field coupling analysis of mechanics and environment was carried out on the wooden structure to simulate the stability and durability of the repaired structure under different environmental conditions.
[0130] Combined with external dynamic load conditions such as earthquakes and wind loads, the response of the repaired structure under extreme conditions is analyzed to ensure the reliability of the repair plan.
[0131] During the actual repair and construction process, sensors and real-time monitoring systems are used to monitor the deformation, stress and environmental conditions of the building. Through real-time feedback from the intelligent system, construction workers are guided to make corrections to ensure the accuracy and safety of the construction.
[0132] AR equipment is used at the construction site to project digital models and repair plans in real time, assisting workers in performing precise repair operations and ensuring that the repair effect meets expectations.
[0133] During the construction process, the repair strategy is automatically adjusted based on the real-time data collected by sensors. For example, when local stress concentration is detected, the repair method is optimized in real time to avoid secondary damage.
[0134] Develop and use intelligent tools, such as self-adjusting mechanical support devices, to ensure that every step in the repair process is carried out within safe limits.
[0135] A sensor network is deployed in the restored building to monitor key indicators such as wood moisture, temperature, stress and deformation. The sensors should have self-diagnosis and self-calibration functions to ensure the reliability of long-term monitoring.
[0136] Wireless data transmission and cloud storage: Sensor data is transmitted to cloud storage in real time via wireless networks, and combined with big data technology for long-term trend analysis and anomaly detection.
[0137] By analyzing long-term monitoring data through artificial intelligence algorithms, the health status of wooden structures can be automatically assessed and possible damage and deterioration trends can be predicted in advance.
[0138] Based on the health assessment results, an adaptive maintenance and repair plan is formulated, and combined with the aforementioned virtual simulation technology, pre-repair assessment and solution optimization are carried out.
[0139] Specifically: The method includes: during the actual repair construction process, using sensors and real-time monitoring systems, providing real-time feedback through intelligent systems, guiding construction workers to make corrections, and using AR equipment at the construction site to project digital models and repair plans in real time.
[0140] Specifically: the intelligent system is expressed as follows:
[0141] G AR =R(S reconstruct , I user );
[0142] Among them, G AR : Guidance information generated by the AR system;
[0143] R: rendering function of the AR system;
[0144] I user : Information or operation entered by the user.
[0145] The present invention realizes high-precision restoration modeling and reconstruction of wooden cultural relics and ancient buildings by combining multi-sensor fusion, intelligent algorithms, virtual simulation and real-time monitoring technology. This method not only overcomes the data acquisition accuracy and integrity problems in the prior art, but also improves the accuracy and reliability of the restoration work through intelligent means, providing a scientific and innovative solution for the protection and restoration of cultural relics and ancient buildings.
[0146] 1. Multi-view data fusion, expression formula:
[0147]
[0148] in:
[0149] Represents the coordinates of a certain point of the ancient wooden building at the jth viewing angle.
[0150] R j : corresponds to the rotation adjustment matrix of the image or point cloud data obtained from the jth view. This represents how to adjust the rotation of data from different viewpoints in the digital modeling process so that it is aligned to a unified three-dimensional space.
[0151] T j : It is the translation adjustment vector when collecting data from the jth perspective, which is used to correct the displacement error between different perspectives during the modeling process.
[0152] Summary: In the process of restoring cultural relics and ancient buildings, multi-perspective data fusion can accurately align data from different perspectives to ensure that the generated 3D model accurately reflects the true geometric structure of the building.
[0153] 2. Point cloud denoising and simplification, expression formula:
[0154] N ∈ (p i )={p j ∈P:||p i -p j ||≤∈};
[0155] in:
[0156] N ∈ (p i ): represents the neighborhood point set within a given radius ∈ of a certain point of a wooden structure cultural relic. In restoration modeling, this neighborhood point set is used to identify noise points to ensure that the retained point set reflects the true surface of the cultural relic.
[0157] ∈: defines the size of the neighborhood, which is used to determine which points are valid data points and which are noise that needs to be removed during the modeling process.
[0158] Summary: By setting a reasonable neighborhood radius ∈ and removing interference data, the retained point cloud data can more realistically represent the actual surface of the wooden structure cultural relics.
[0159] 3. Surface reconstruction and refinement, expression formula:
[0160]
[0161] in:
[0162] △u i In digital modeling of wooden artifacts, the Laplacian operator represents the target function (i.e. the generated 3D surface). This means how points on the model surface are distributed in 3D space relative to the surrounding points.
[0163] v: A field representing the surface normal vectors of wooden artifacts. In the modeling process, it represents the normal direction of each point, which is essential for reconstructing a smooth surface.
[0164] Summary: Poisson surface reconstruction generates a smooth and continuous three-dimensional surface by solving this equation, ensuring that the digital model can accurately reproduce the real appearance of wooden structure artifacts.
[0165] 4. Finite element analysis, expression formula:
[0166] Ku=F;
[0167] Combined with design ideas:
[0168] K represents the stiffness matrix of the wooden structure artifact, which reflects the rigidity of each part of the structure and its deformation characteristics when subjected to external forces.
[0169] u i It represents the displacement of each node of the structure and is used to predict the deformation of the wood after being subjected to stress during the repair process.
[0170] F: External forces acting on the structure, simulating external forces or environmental loads that may be applied during repair or maintenance.
[0171] Summary: Finite element analysis can be used to simulate and predict the mechanical response of restored cultural relics structures under external forces, thereby helping to select appropriate restoration materials and processes.
[0172] 5. Adaptive correction driven by machine learning, expressed as:
[0173]
[0174] in:
[0175] θ: Parameters of the machine learning model, which are optimized through training during the repair process to achieve the most accurate error correction effect.
[0176] x i : Input data, which represents the raw data collected by the sensor and may include noise or errors.
[0177] y i : Ideal correction result, reflecting the most accurate data expected during the repair process.
[0178] Summary: Machine learning models are trained to correct errors in sensor data, improving the accuracy and reliability of digital models during the repair process.
[0179] 6. Multi-field coupling analysis, expression formula:
[0180]
[0181] in:
[0182] T: Temperature field, reflecting the temperature distribution of wooden structure cultural relics under different environmental conditions.
[0183] The thermal conductivity term describes how heat is transferred through the wood.
[0184] Q: Heat source term, which may be the ambient temperature or the heat introduced during the repair process.
[0185] Summary: Multi-field coupling analysis helps predict the behavior of restored cultural relics under long-term environmental influences and ensures that the structure remains stable under various environmental conditions.
[0186] Embodiment 2
[0187] A digital technology-based restoration modeling and reconstruction method for wooden structure cultural relics and ancient buildings, comprising the following steps:
[0188] Step 1: Intelligent collection and fusion of multimodal data: AI-based multimodal sensor fusion uses artificial intelligence algorithms to intelligently fuse data from multiple sensors such as laser scanning, infrared thermal imaging, and optical cameras.
[0189] Expression formula:
[0190]
[0191] F i : Data characteristics of the i-th sensor.
[0192] w i: The weight of each sensor data learned based on the AI model.
[0193] In practical applications, F fusion Represents the fused multimodal data features, which are used to build a more accurate 3D model. The AI algorithm dynamically adjusts the weight w i , automatically optimizes the fusion between different sensor data to generate more accurate model data.
[0194] Step 2: Point cloud denoising and surface reconstruction based on deep learning: 2. Point cloud denoising based on deep learning, introduce a point cloud denoising method based on deep learning, and automatically identify and remove noise points through a neural network.
[0195] Expression formula:
[0196] p clean =f DL (p raw );
[0197] p raw : Original point cloud data.
[0198] f DL : Deep learning model for denoising.
[0199] p clean : Point cloud data after denoising.
[0200] p raw is the deep learning model f DL The processed point cloud data can remove noise in complex environments and ensure the accuracy and integrity of the surface of wooden structure cultural relics.
[0201] Neural network-based surface reconstruction: The neural network-based surface reconstruction method can generate smoother and more detailed three-dimensional surfaces on complex wooden structure artifacts.
[0202] Expression formula:
[0203]
[0204] Neural network models for surface reconstruction.
[0205] S reconstruct : The reconstructed surface model.
[0206] Represents the neural network in the point cloud data p clean Application on the generation of a smooth three-dimensional surface model S reconstruct , which is more adaptable to complex surface details of cultural relics than traditional Poisson reconstruction methods.
[0207] Step 3: Virtual restoration simulation based on digital twins: Introduce digital twin technology to simulate the physical behavior and environmental response of wooden structure cultural relics during the restoration process in real time. Expression formula:
[0208] T virtual (t) = M (u real (t), p external );
[0209] T virtual (t): virtual repair state at time t.
[0210] u real (t): displacement and deformation in the realistic restoration state.
[0211] p external : External conditions (such as ambient temperature, humidity, etc.).
[0212] During the restoration process, digital twin technology is used to simulate the restoration status of cultural relics in real time. virtual (t), and combined with the actual environmental conditions p external , ensuring the accuracy and stability of the repair process.
[0213] Step 4: AR-assisted restoration guidance: The AR-based restoration guidance system uses AR technology to provide real-time restoration guidance to restorers and optimize restoration operations by projecting virtual restoration suggestions on cultural relics.
[0214] Expression formula:
[0215] G AR =R(S reconstruct , I user );
[0216] G AR : Guidance information generated by the AR system.
[0217] R: Rendering function of the AR system.
[0218] I user : Information or operation entered by the user.
[0219] The AR system uses the rendering function R to render the repair suggestion G AR Projected on the cultural relics, combined with the operation information input by the user user , providing precise and real-time guidance for the repair process.
[0220] Step 5: Intelligent feedback and adaptive optimization: Machine learning-driven adaptive optimization of the repair process, which automatically optimizes the repair process through machine learning algorithms and adjusts the repair strategy based on real-time data and feedback on the repair effect.
[0221] Expression formula:
[0222]
[0223] θ: Parameters of the machine learning model.
[0224] G AR (t i ): time t i AR guidance information.
[0225] T virtual (t i ): time t i Virtual repair status.
[0226] L(θ): loss function that measures the difference between AR guidance and virtual restoration.
[0227] By minimizing the loss function L(θ), the machine learning model can continuously optimize the repair guidance G AR The actual repair effect T virtual The differences between them ensure the efficiency and accuracy of the repair process.
[0228] Combining cutting-edge technologies such as artificial intelligence, deep learning, digital twins, augmented reality and machine learning, the paper explains the key technical elements in each step and their practical application in the restoration modeling of wooden cultural relics through detailed mathematical formulas. These technical means have significantly improved the accuracy, efficiency and controllability of restoration, ensuring that the restoration process of cultural relics is scientific and reasonable and meets the application requirements of digital technology.
[0229] Embodiment 3
[0230] An embodiment of the present application also provides a computer-readable storage medium, which can be set in a server to store at least one instruction or at least one program related to a method for repairing, modeling and reconstructing wooden structures, cultural relics and ancient buildings based on digital technology in a method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement a method for repairing, modeling and reconstructing wooden structures, cultural relics and ancient buildings based on digital technology provided in the above method embodiment.
[0231] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.
[0232] Embodiment 4
[0233] The embodiment of the present invention also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes a digital technology-based wooden structure cultural relic ancient building restoration modeling and reconstruction method provided in the above various optional embodiments.
[0234] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0235] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment and storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0236] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.
[0237] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.
[0238] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A digital technology-based restoration modeling and reconstruction method for wooden structure cultural relics and ancient buildings, characterized in that: The following steps are involved: Acquire three-dimensional data of the cultural relics and ancient buildings to be restored, collect information of the three-dimensional data based on multiple data acquisition devices and deployments, pre-process the data in real time according to the embedded processing unit, input the data collected by the multiple data acquisition devices and deployments into a three-dimensional coordinate system in response to a multi-view stereo matching algorithm, and generate a three-dimensional point cloud model; The three-dimensional point cloud model automatically identifies and corrects errors in the data collection of the three-dimensional point cloud model based on an adaptive correction algorithm of machine learning, performs denoising on the preliminary point cloud data according to a clustering algorithm, and automatically identifies the mortise and tenon structure in the point cloud according to a pattern recognition algorithm to construct a three-dimensional mortise and tenon model; Based on historical data and material principles, the aging process of wood is simulated, finite element analysis is used for mechanical simulation, the mechanical behavior of wood structures is predicted, and three-dimensional point cloud models are used to test different repair schemes and simulate the repair process; A network of sensors is deployed in the restored building to monitor the moisture, temperature, stress and deformation of the wood; Long-term monitoring data is analyzed through artificial intelligence algorithms to automatically assess the health status of wooden structures.
2. The method according to claim 1, characterized in that: The multiple data collection devices and deployments specifically include: 3D laser scanners: Multiple high-precision 3D laser scanners are deployed around and inside the cultural relics and ancient buildings to capture the overall geometric structure of the cultural relics and ancient buildings; High-resolution cameras: Install high-resolution cameras on the exterior of cultural relics and ancient buildings to capture surface details and textures of cultural relics and ancient buildings; Infrared scanner: Detect cavities, cracks and aging inside wooden structures using infrared scanners; Environmental sensors: Place temperature and humidity sensors inside cultural relics and ancient buildings, and record environmental parameters during scanning.
3. The method according to claim 2, characterized in that: The 3D laser scanner, high-resolution camera, infrared scanner and environmental sensor communicate with each other via wireless network or optical fiber to achieve synchronous data collection of multiple sensors; Based on the multi-view stereo matching algorithm, laser scanning data, photogrammetry data and infrared scanning data are input into the three-dimensional coordinate system, and a three-dimensional point cloud model is generated through feature point matching and rigid body transformation.
4. The method according to claim 2, characterized in that: The real-time data preprocessing is performed according to the embedded processing unit, and the preprocessing includes noise filtering, error calibration and point cloud stitching.
5. The method according to claim 1, characterized in that: The method of performing denoising on the preliminary point cloud data according to the clustering algorithm, removing isolated points and noise points, automatically identifying the mortise and tenon structure in the point cloud according to the pattern recognition algorithm, and constructing the mortise and tenon three-dimensional model specifically includes: The preliminary point cloud data is denoised using a density-based clustering algorithm to remove isolated and noisy points, and the point cloud is streamlined using a voxel grid simplification algorithm; According to the Poisson surface reconstruction algorithm, the simplified point cloud data is converted into a continuous three-dimensional surface model; Construct a 3D model of mortise and tenon joints, and automatically identify mortise and tenon joint structures with typical features in the point cloud based on historical data and traditional craftsmanship through pattern recognition algorithms; In response to the mortise and tenon structure with typical features in the point cloud, a high-resolution texture is mapped onto the surface of the mortise and tenon three-dimensional model according to the photogrammetry data; The optical BRDF model is used to simulate the optical properties of wood and reproduce the real material of the wood structure.
6. The method according to claim 1, characterized in that: Based on historical data and material principles, the aging process of wood is simulated, finite element analysis is used for mechanical simulation, the mechanical behavior of wood structures is predicted, and three-dimensional point cloud models are used to test different repair schemes and simulate the repair process, including: Digital twin technology, real-time simulation of the physical behavior and environmental response of wooden structure cultural relics during the restoration process; The environmental response includes ambient temperature, ambient humidity, earthquake and wind load, and the expression formula is: T virtual (t)=M(u real (t),p external ); Among them, T virtual (t): virtual repair state at time t; u real (t): displacement and deformation in the actual restoration state; p external : Environmental response.
7. The method according to claim 1, characterized in that: The method uses artificial intelligence algorithms to analyze long-term monitoring data and automatically assess the health status of wooden structures, including: The repair strategy is adjusted based on the automatic assessment of the health status of the wood structure, expressed as: in, θ: Parameters of the machine learning model for wooden structure artifacts; G AR (t i ): time t i Artificial intelligence algorithm guidance information; T virtual (t i ): time t i The status of the virtual restoration of wooden structure heritage; L(θ): loss function that measures the difference between the guidance of the AI algorithm and the virtual restoration.
8. The method according to claim 1, characterized in that: The method includes: during the actual repair construction process, using sensors and real-time monitoring systems, providing real-time feedback through intelligent systems to guide construction workers to make corrections, and using AR equipment at the construction site to project digital models and repair plans in real time.
9. The method according to claim 8, characterized in that: The intelligent system is expressed as follows: G AR =R(S reconstruct ,I user ); Among them, G AR : Guidance information generated by the AR system; R: rendering function of AR system; I user : Information or operation entered by the user.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the processor implements the digital technology-based restoration, modeling and reconstruction method for wooden structure cultural relics and ancient buildings as described in any one of claims 1 to 9.
Citation Information
Cited By
Indoor decoration engineering intelligent management system applying BIM technology
CN119251018A
An intelligent management system for interior decoration projects applying BIM technology
CN119251018B
Crisp moon cake production safety quality monitoring method and system based on artificial intelligence
CN119539584A
Original village building restoration method and system based on 3D printing
CN119559085A
Reconstruction method of historic building structure analysis model
CN119760813A
