A method for detailed modeling of ancient building components using BIM technology

Abstract

This invention discloses a method for detailed modeling of ancient building components using BIM technology, belonging to the field of ancient building conservation technology. The invention includes the following steps: S1: Preliminary survey and data collection of ancient building components; S2: Point cloud data preprocessing: the collected raw point cloud data is sequentially denoised, registered, segmented, and simplified; S3: Constructing a parameterized BIM family library for ancient building components; S4: Component-level detailed modeling and information fusion; S5: Overall model assembly and verification; S6: Lightweight model processing and multi-terminal adaptation; S7: Dynamic model update and iteration. This invention solves the problem of low accuracy in existing ancient building BIM modeling, and simultaneously constructs an integrated BIM model of "geometric information - attribute information - damage information," enabling visualized query and management of information, providing comprehensive and accurate data support for ancient building damage assessment and precise restoration design.

Description

Technical Field

[0001] This invention relates to the field of ancient building conservation technology, specifically to a method for detailed modeling of ancient building components using BIM technology. Background Technology

[0002] In the process of new urbanization and urban renewal in my country, the protection and restoration of ancient buildings face problems such as structural safety disturbance, damage to historical features, and prominent contradictions between protection and development. Digital modeling is the core technical support for achieving precise restoration of ancient buildings.

[0003] BIM technology, with its advantages of 3D visualization, information integration, and full-process control, has been gradually applied to the protection of ancient buildings. However, existing technologies still have the following core shortcomings: 1. Existing BIM modeling mostly focuses on the overall appearance of ancient buildings, failing to reproduce the details of intricate components such as mortise and tenon joints, brackets, and brick carvings, making it difficult to support core tasks such as damage assessment and precise restoration design; 2. Existing BIM models only contain basic geometric information and lack key attribute information such as component materials, historical craftsmanship, and degree of damage. Summary of the Invention

[0004] The purpose of this invention is to provide a method for fine modeling of ancient building components that integrates BIM technology, which solves the problem of low accuracy in existing ancient building BIM modeling and realizes integrated fine modeling of geometric information, attribute information and damage information of ancient building components.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for detailed modeling of ancient building components using BIM technology includes the following steps:

[0007] S1: Preliminary Survey and Data Collection of Ancient Architectural Components

[0008] First, a comprehensive on-site survey of the ancient building was conducted. The ancient building was sorted out according to the hierarchy of "whole-zone-component" to determine the modeling zones and core modeling components. The core modeling components include characteristic components of ancient buildings such as wooden mortise and tenon joints, brackets, wooden beams, wooden columns, brick carvings, walls, and stone foundations. Each component is an independent modeling unit.

[0009] A combination of 3D laser scanning equipment and close-range photogrammetry is used to collect data from all dimensions of each modeling unit: the nominal ranging accuracy of the 3D laser scanning equipment is no higher than 0.1mm (the smaller the value, the higher the accuracy); the scanning point spacing is adaptively adjusted according to the complexity of the components, using a point spacing of 0.05-0.1mm for delicate components such as brackets and mortise and tenon joints, and 0.1-0.5mm for large components such as wooden beams and walls, to collect high-precision point cloud data of the components; close-range photogrammetry is used to collect high-definition texture data of the components, providing a foundation for subsequent model texture mapping;

[0010] Meanwhile, through on-site surveys, interviews with craftsmen, and review of historical data, non-geometric information of each component was collected, including material type, historical craftsmanship, damage information, and protection level. The non-geometric information was standardized and coded in three levels. The coding rule is component material code - zone code - component serial number. For example, ML-01 represents wooden component - East zone - No. 01. The code is globally uniquely bound to the corresponding component modeling unit.

[0011] S2: Point cloud data preprocessing

[0012] The collected raw point cloud data is sequentially processed by denoising, registration, segmentation and simplification;

[0013] S3: Constructing a BIM parametric family library for ancient building components

[0014] Based on the standards and specifications of the "Guidelines for the Construction Techniques of Ancient Chinese Buildings" and component data collected on-site, a component-level parametric family library adapted to BIM technology was developed on the Revit platform, specifically targeting characteristic components of ancient Chinese architecture. The family library is divided into three sub-libraries according to component material: timber component families, brick component families, and stone component families. Each sub-library contains basic form families and damaged adaptation families.

[0015] Basic Form Families: Standard form development for various types of components, including core geometric parameters, material parameters, and process parameters of the components, supporting parametric adjustment, and adapting to modeling of the same type of components in different ancient buildings; the basic form families of the wooden component family include the mortise and tenon family, the bracket family, the wooden beam family, and the wooden column family; the basic form families of the brick component family include the wall family, the brick carving family, and the blue brick family; the basic form families of the stone component family include the stone foundation family, the stone carving family, and the stone wall family.

[0016] Damage Adaptation Families: Based on the basic structural family, pre-set parametric modules for typical damage characteristics such as cracks, corrosion, loosening, and missing parts are provided. These modules can flexibly adjust the location, size, and degree of damage based on on-site collected damage data to adapt to the modeling of ancient architectural components in different damage states, achieving accurate modeling of damaged components. The damage adaptation families for wooden components include crack modules, corrosion modules, loosening modules, and missing parts modules; for brick components, they include cracking modules, detachment modules, and weathering modules; and for stone components, they include breakage modules, weathering modules, and displacement modules.

[0017] All parameters of the parametric family support integration with subsequent repair design and construction management systems, enabling parameterized model-driven operation.

[0018] S4: Component-level fine modeling and information fusion

[0019] Import the segmented component point cloud data into a BIM modeling platform based on Revit secondary development. This platform is equipped with a special plugin for modeling ancient building components, which supports real-time comparison of point cloud data, accurate model drawing, and one-click embedding of non-geometric information.

[0020] Using the component point cloud model as a precise benchmark, the parametric family in step S3 is called to perform fine modeling of the component. The point cloud real-time comparison function of the plugin is used to accurately match the parametric family with the component point cloud. The geometric features of the component are drawn and corrected one by one to achieve accurate mapping of point cloud data to BIM three-dimensional solid model and restore the core features of the component such as mortise and tenon interlocking relationship, bracket hierarchical structure, and brick carving texture details.

[0021] The plugin's one-click information embedding function embeds non-geometric information such as material, process, and damage collected and standardized in step S1 into the attribute library of the corresponding BIM component model, establishing an integrated component-level BIM model of "geometric information-attribute information-damage information" to realize the visual query and management of component information; at the same time, texture mapping is performed on the component model, and high-definition texture data is attached to the BIM three-dimensional solid model to make the model fit the actual appearance of the ancient building components.

[0022] S5: Overall Model Assembly and Verification

[0023] According to the original structural logic, spatial relationship and construction techniques of the ancient building, the BIM models of each component are precisely assembled to form the overall BIM model of the ancient building, ensuring that the connection between the components is completely consistent with the ancient building itself.

[0024] The assembled model is then subjected to accuracy verification and correction. The model accuracy is quantified using three indicators: average deviation, maximum deviation, and standard deviation of deviation. The formula for calculating the average deviation is as follows: The formula for calculating the maximum deviation is: The formula for calculating the standard deviation is: ,in The average deviation of the model. For the maximum deviation of the model, The standard deviation is the deviation. The total number of checkpoints. For the verification point number, For the first BIM model The three-dimensional spatial coordinates of each verification point For the ancient building entity / measured point cloud and The corresponding number The three-dimensional spatial coordinates of each verification point for and Euclidean distance between them;

[0025] Verification points should prioritize key feature points of components such as mortise and tenon joints, bracket set levels, and wall corners. The verification criteria for models of national-level cultural relics protection units are as follows: , , Provincial level and below , , For parts that exceed the threshold, manual refinement is carried out to ensure the consistency between the model and the ancient building itself.

[0026] S6: Lightweight Model Processing and Multi-Platform Adaptation

[0027] The overall BIM model undergoes LOD (Level of Detail) hierarchical management and lightweight optimization in accordance with the "Unified Standard for Building Information Modeling Applications" GB / T51212-2016, and is divided into three application levels:

[0028] LOD400 (Component Fabrication Level): Preserves complete component geometry parameters, attribute information, and damaged details for precise repair design and component fabrication; when exporting a visual mesh model, the number of triangles per component is controlled to within 50,000.

[0029] LOD300 (Construction Level): Retains the core geometric features and key attribute information of components for construction briefings and on-site management; when exporting a visual mesh model, the number of triangular faces per component is controlled between 10,000 and 50,000.

[0030] LOD200 (Scheme Demonstration Level): Preserves the overall architectural form and spatial relationships of components for multi-terminal lightweight display; when exporting the visual mesh model, the number of triangular faces of a single component is controlled between 5,000 and 10,000.

[0031] The lightweighting effect is quantified by the patch simplification rate and Hausdorff distance. The formula for calculating the patch simplification rate is: ,in To simplify the dough, To reduce the number of triangles in the original model, The number of triangles in the lightweight model; Hausdorff distance formula. Quantify geometric bias, among which The Hausdorff distance between the models before and after lightweighting. The original BIM model before lightweighting. For the lightweight BIM model, for Any three-dimensional point on, for Any three-dimensional point on, for and The Euclidean distance between them; the lightweight qualification conditions are ≥70% and This ensures that fine-grained component-level information is retained after lightweighting, while reducing the amount of model data.

[0032] Simultaneously, we developed multi-terminal data interfaces supporting mainstream formats such as IFC, DWG, and FBX. Among them, the IFC format is used for full-information model integration, realizing seamless transfer of integrated "geometry-attribute-damage" data with restoration design, construction management, and operation and maintenance monitoring systems; the DWG format is used for two-dimensional drawing interaction, connecting with CAD restoration design software; and the FBX format is used for visual model export, connecting with display and disclosure scenarios, realizing seamless integration of component-level BIM models with CAD restoration design software, construction management systems, and digital operation and maintenance monitoring systems for the entire life cycle of ancient buildings, completing the synchronization and storage of model data, so that the model can directly support restoration scheme design, construction process control, and real-time damage monitoring.

[0033] S7: Dynamic Model Updates and Iterations

[0034] Establish a dynamic updating and maintenance mechanism for BIM models at the component level of ancient buildings, and assign dedicated personnel to be responsible for the full lifecycle management of the models: During the restoration and construction of ancient buildings, the BIM models are modified and updated in a timely manner according to on-site construction adjustments and the actual state of the components after restoration; the models are iteratively optimized by combining the new damage data and monitoring data generated by the regular inspection of ancient buildings and urban renewal projects; the updated models are synchronized to the digital management platform for the full lifecycle of ancient buildings in real time to ensure that the models are consistent with the actual state of the ancient buildings, providing accurate digital support for subsequent operation and maintenance monitoring and secondary restoration; at the same time, the original data, processed data, model files, and modification records of the entire modeling process are uniformly stored and managed to achieve traceability of the entire modeling process.

[0035] Furthermore, the specific steps of S2 are as follows:

[0036] S21: Point Cloud Denoising: Noise points are identified using a combination of neighborhood density and normal curvature methods. The neighborhood density calculation formula is as follows: ,in For the first The neighborhood density of a point cloud, For the first The number of nearest neighbors in a point cloud. for The area of ​​the neighborhood formed by nearest neighbors; formula for calculating normal curvature. ,in For the first The normal curvature of a point cloud, These are the eigenvalues ​​of the neighborhood covariance matrix of the point cloud; the formula for calculating the neighborhood covariance matrix is: , for Mean coordinates of nearest neighbors For the first The three-dimensional coordinates of the nearest neighbor points; simultaneously satisfying and The point cloud points were identified as noise points and removed. The minimum neighborhood density threshold is the threshold corresponding to the component type. The minimum neighborhood density threshold for fine components is... Large components are ;

[0037] S22: Point Cloud Registration: An improved ICP algorithm is used for coarse and fine registration. First, coarse registration of multi-view point clouds is achieved using FPFH (Fast Point Feature Histogram), and then fine registration is completed using the ICP algorithm. The registration accuracy is calculated using the root mean square error formula. Quantification, among which The root mean square error for point cloud registration, To effectively match the total number of point pairs, To match point pair sequence numbers, The first point in the source cloud The three-dimensional spatial coordinates of the points The 4×4 registration transformation matrix (including the 3×3 rotation matrix) and 3×1 translation vector ,satisfy ), In the target point cloud The corresponding number The three-dimensional spatial coordinates of the matching points for With scriptures Mapped The Euclidean distance between them; the registration convergence condition is And the two adjacent iterations Difference ;

[0038] S23: Point Cloud Segmentation: Point clouds of the same component are segmented using the feature similarity method. The feature similarity calculation formula is as follows: ,in For the first , Feature similarity of point clouds We set the weighting coefficient to 0.6. For the first , The angle between the normal vectors of the point cloud. for cosine value, For the first , Euclidean distance between points in the cloud Distance threshold; Points with a value ≥0.7 are considered to be the same component point cloud, with fine components being... Set to 1mm, for large components Set to 5mm; after traversing all point clouds, obtain the point cloud model of each independent component, complete the component-level segmentation, and at the same time perform preliminary matching between the high-definition texture data and the corresponding component point cloud;

[0039] S24: Point Cloud Simplification: While preserving the key details of the components, voxel downsampling is performed on the dense point cloud to simplify it, thereby reducing the amount of data processing required for subsequent modeling.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] 1. This invention targets distinctive components of ancient Chinese architecture, such as mortise and tenon joints and bracket sets. It achieves precise point cloud processing through dual-index denoising, improved ICP registration, and feature similarity segmentation. Combined with a dedicated BIM parametric family library, it enables detailed component-level modeling, accurately reproducing the geometric features, craftsmanship details, and damage status of the components, thus solving the problem of insufficient accuracy in existing modeling. Simultaneously, it constructs an integrated BIM model encompassing "geometric information, attribute information, and damage information," deeply integrating key information such as the component's material, craftsmanship, and damage with the 3D model. This allows for visualized querying and management of information, providing comprehensive and accurate data support for the assessment of damage to ancient buildings and precise restoration design.

[0042] 2. This invention constructs a parametric family library adapted to ancient wooden, brick, and stone buildings in my country, forming a standardized component-level fine modeling process. It can be widely applied to the digital restoration and repair of various ancient buildings in urban renewal areas, reducing rework rates and lowering repair costs. Attached Figure Description

[0043] Figure 1This is a flowchart of the modeling method of the present invention;

[0044] Figure 2 This is a schematic diagram of the structural design of the ancient building component-level BIM parametric family library of the present invention;

[0045] Figure 3 This is a schematic diagram of the integrated component-level BIM model information fusion architecture of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Please see Figure 1-3 This embodiment focuses on the component-level detailed modeling of a wooden and brick ancient building in a certain region. This building is a provincial-level cultural relic protection unit. Its core components include mortise and tenon joints, seven-tiered bracket sets, blue brick walls, wooden beams and columns, and stone foundations. Simultaneously, urban renewal and underground space development projects are underway in the surrounding area. The constructed model is needed to support the precise restoration and active protection of the ancient building. The implementation steps are as follows (e.g....). Figure 1 (as shown)

[0048] S1: Preliminary Survey and Data Collection of Ancient Architectural Components

[0049] The ancient building was divided into four sections: east, west, south, and north. Within each section, core modeling units were identified based on the structure of "beams-columns-bracket sets-walls-stone foundations," resulting in 86 independent component modeling units. A 3D laser scanner with a scanning precision of 0.1mm was used to scan each component. Fine components such as bracket sets and mortise and tenon joints were scanned using a 0.05mm point spacing, while large components such as wooden beams and walls were scanned using a 0.3mm point spacing, collecting complete point cloud data. A high-resolution SLR camera combined with close-range photogrammetry was used to acquire high-definition texture data for each component.

[0050] Meanwhile, on-site surveys recorded non-geometric information of each component: for example, the No. 1 wooden beam in the East Zone is made of Phoebe zhennan wood and has corrosion, cracks and damage, with a crack length of 1.2m; the bracket set in the East Zone is a seven-step bracket set from the Qing Dynasty, with two components missing; the blue brick wall is built with glutinous rice mortar and has cracks and brick falling off. All information was standardized and coded (e.g., ML-01 represents wooden beam-East Zone-01) and bound to the corresponding component modeling unit;

[0051] S2: Point cloud data preprocessing

[0052] S21: Point cloud denoising: Neighborhood density method combined with normal curvature method is used. Take 15, including fine components such as brackets, mortise and tenon joints, etc. Large components such as wooden beams and walls Eliminate those that simultaneously satisfy and After noise removal, the effective point retention rate was 98.7%, and the noise point removal rate was 96.2%.

[0053] S22: Point cloud registration:

[0054] Using an improved ICP algorithm, the number of matching point pairs n=1260, after registration Difference between adjacent iterations This satisfies the convergence condition and enables precise stitching of point clouds from multiple perspectives.

[0055] S23: Point cloud segmentation: using the feature similarity method. Take 0.6, for fine components Large components , Points with a value ≥0.7 were identified as the same component point cloud. The final segmentation yielded 86 independent point cloud models of components with a segmentation accuracy of 99.2%. At the same time, the initial matching of high-definition texture data and corresponding component point clouds was completed.

[0056] S24: Point cloud simplification: Voxel downsampling is used to simplify the dense point cloud, reducing the point cloud data volume by 58% while retaining all key details of the components.

[0057] S3: Constructing a BIM parametric family library for ancient building components

[0058] Based on ancient building construction standards and on-site data, three parametric family libraries (such as timber, brick, and stone structures) were developed on the Revit platform. Figure 2 (As shown in the image) The wooden structure family library includes the mortise and tenon family, the seven-step bracket family, and the wooden beam and column family, with preset modules for damage such as corrosion, cracks, and missing parts; the brick structure family library includes the blue brick wall family and the brick carving family; the stone structure family library includes the stone foundation family, and all family parameters can be flexibly adjusted.

[0059] S4: Component-level fine modeling and information fusion

[0060] The segmented component point clouds were imported into a Revit-based BIM modeling platform for ancient buildings. Parametric families were then used to create component BIM models based on the point clouds, accurately reproducing the mortise and tenon joints and the hierarchical structure of the bracket sets. Standardized coded material, damage, and process information were embedded into the corresponding model attribute library using a dedicated plugin. For example, information such as the nanmu wood material of ML-01, a crack length of 1.2m, and a corrosion rate of 30% were embedded into the wooden beam model. Texture mapping was completed to ensure the model closely matched the actual appearance of the components. Finally, 86 integrated "geometry-attribute-damage" component-level BIM models were established (e.g., ...). Figure 3 (as shown)

[0061] S5: Overall Model Assembly and Verification

[0062] Following the structural logic of ancient timber-framed buildings, 86 component-level BIM models were assembled into a unified BIM model of the ancient building, ensuring consistency between beam-column connections, bracket set overlaps, and mortise-and-tenon joints with the building itself. Key feature points of each component were selected as check points, with ≥50 check points per component and a total of 4580 check points. The average deviation of the model was calculated. , the maximum deviation Standard deviation It meets the accuracy requirements for provincial-level cultural relics protection units;

[0063] S6: Lightweight Model Processing and Multi-Platform Adaptation

[0064] The overall model was optimized using LOD hierarchical lightweighting. The LOD400 model retained 62,000 triangles, the LOD300 model retained 28,000 triangles, and the LOD200 model retained 7,500 triangles. Lightweighting reduced the triangle simplification rate. =76%≥70%, Hausdorf distance It meets the lightweighting requirements;

[0065] Develop multi-terminal data interfaces to support exporting in IFC and DWG formats, and synchronize the model to the ancient building restoration design software, construction management system and digital operation and maintenance monitoring platform. The model can be smoothly queried and operated on construction tablets, computers and web pages.

[0066] S7: Dynamic Model Updates and Iterations

[0067] A designated person is responsible for model management. During the restoration of ancient buildings, the model is updated in a timely manner for the repaired wooden beams and the replacement bracket components. By combining vibration monitoring data from the development of surrounding underground spaces, vibration deformation data of ancient building components are embedded into the model to achieve dynamic iteration of the model and provide data support for the active protection of ancient buildings. At the same time, the data and files of the entire modeling process are stored and managed in a unified manner to achieve full traceability.

[0068] This embodiment successfully supported the precise restoration design and construction control of the ancient building by constructing a component-level BIM model. Compared with the traditional manual modeling restoration mode, the rework rate of this project was reduced by 62%, and the collaborative efficiency of restoration scheme design and on-site construction was improved by 56%. The model is also connected to the urban renewal planning system, providing comprehensive and accurate digital data support for the coordinated promotion of ancient buildings and urban renewal.

[0069] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for detailed modeling of ancient building components integrating BIM technology, characterized in that: Includes the following steps: S1: Preliminary Survey and Data Collection of Ancient Architectural Components First, a comprehensive on-site survey of the ancient building was conducted. The ancient building was sorted out according to the hierarchy of "whole-zone-component" to determine the modeling zones and core modeling components. The core modeling components include characteristic components of ancient buildings such as wooden mortise and tenon joints, brackets, wooden beams, wooden columns, brick carvings, walls, and stone foundations. Each component is an independent modeling unit. A 3D laser scanning device combined with close-range photogrammetry was used to collect data from all dimensions of each modeling unit: the nominal ranging accuracy of the 3D laser scanning device is no higher than 0.1mm; the scanning point spacing is adaptively adjusted according to the complexity of the components, with a point spacing of 0.05-0.1mm for delicate components such as brackets and mortise and tenon joints, and 0.1-0.5mm for large components such as wooden beams and walls, to collect high-precision point cloud data of the components; close-range photogrammetry was used to collect high-definition texture data of the components, providing a foundation for subsequent model texture mapping; Meanwhile, through on-site surveys, interviews with craftsmen, and review of historical data, non-geometric information of each component was collected, including material type, historical craftsmanship, damage information, and protection level. The non-geometric information was then standardized into a three-level coding system, with the coding rule being component material code - partition code - component serial number. The code is globally and uniquely bound to the corresponding component modeling unit. S2: Point cloud data preprocessing The collected raw point cloud data is sequentially processed by denoising, registration, segmentation and simplification; S3: Constructing a component-level BIM parametric family library for ancient buildings Based on the standards and specifications of the "Guidelines for the Construction Techniques of Ancient Chinese Buildings" and component data collected on-site, a component-level parametric family library adapted to BIM technology was developed on the Revit platform, specifically targeting characteristic components of ancient Chinese architecture. The family library is divided into three sub-libraries according to component material: timber component families, brick component families, and stone component families. Each sub-library contains basic form families and damaged adaptation families. Basic Form Families: Standard form development for various types of components, including core geometric parameters, material parameters, and process parameters of the components, supporting parametric adjustment, and adapting to modeling of the same type of components in different ancient buildings; the basic form families of the wooden component family include the mortise and tenon family, the bracket family, the wooden beam family, and the wooden column family; the basic form families of the brick component family include the wall family, the brick carving family, and the blue brick family; the basic form families of the stone component family include the stone foundation family, the stone carving family, and the stone wall family. Damage Adaptation Families: Based on the basic structural family, pre-set parametric modules for typical damage characteristics such as cracks, corrosion, loosening, and missing parts are provided. These modules can flexibly adjust the location, size, and degree of damage based on on-site collected damage data to adapt to the modeling of ancient architectural components in different damage states, achieving accurate modeling of damaged components. The damage adaptation families for wooden components include crack modules, corrosion modules, loosening modules, and missing parts modules; for brick components, they include cracking modules, detachment modules, and weathering modules; and for stone components, they include breakage modules, weathering modules, and displacement modules. All parameters of the parametric family support integration with subsequent repair design and construction management systems, enabling parameterized model-driven operation. S4: Component-level fine modeling and information fusion Import the segmented component point cloud data into a BIM modeling platform based on Revit secondary development. This platform is equipped with a special plugin for modeling ancient building components, which supports real-time comparison of point cloud data, accurate model drawing, and one-click embedding of non-geometric information. Using the component point cloud model as a precise benchmark, the parametric family in step S3 is called to perform fine modeling of the component. The point cloud real-time comparison function of the plugin is used to accurately match the parametric family with the component point cloud. The geometric features of the component are drawn and corrected one by one to achieve accurate mapping of point cloud data to BIM three-dimensional solid model and restore the core features of the component such as mortise and tenon interlocking relationship, bracket hierarchical structure, and brick carving texture details. The plugin's one-click information embedding function embeds non-geometric information such as material, process, and damage collected and standardized in step S1 into the attribute library of the corresponding BIM component model, establishing an integrated component-level BIM model of "geometric information-attribute information-damage information" to realize the visual query and management of component information; at the same time, texture mapping is performed on the component model, and high-definition texture data is attached to the BIM three-dimensional solid model to make the model fit the actual appearance of the ancient building components. S5: Overall Model Assembly and Verification According to the original structural logic, spatial relationship and construction techniques of the ancient building, the BIM models of each component are precisely assembled to form the overall BIM model of the ancient building, ensuring that the connection between the components is completely consistent with the ancient building itself. The assembled model is then subjected to accuracy verification and correction. The model accuracy is quantified using three indicators: average deviation, maximum deviation, and standard deviation of deviation. The formula for calculating the average deviation is as follows: The formula for calculating the maximum deviation is: The formula for calculating the standard deviation is: ,in The average deviation of the model. For the maximum deviation of the model, The standard deviation is the deviation. The total number of verification points. For the verification point number, For the first BIM model The three-dimensional spatial coordinates of each verification point For the ancient building entity / measured points on the cloud and The corresponding number The three-dimensional spatial coordinates of each verification point for and Euclidean distance between them; Verification points should prioritize key feature points of components such as mortise and tenon joints, bracket set levels, and wall corners. The verification criteria for models of national-level cultural relics protection units are as follows: , , Provincial level and below , , For parts that exceed the threshold, manual refinement is carried out to ensure the consistency between the model and the ancient building itself. S6: Lightweight Model Processing and Multi-Platform Adaptation The overall BIM model undergoes LOD (Level of Detail) hierarchical management and lightweight optimization in accordance with the "Unified Standard for Building Information Modeling Applications" GB / T51212-2016, and is divided into three application levels: LOD400 Component Fabrication Level: Preserves complete component geometry parameters, attribute information, and damaged details for precise repair design and component fabrication; when exporting a visual mesh model, the number of triangular faces per component is controlled to within 50,000. LOD300 Construction Level: Retains core geometric features and key attribute information of components for construction briefings and on-site management; when exporting a visual mesh model, the number of triangular faces per component is controlled between 10,000 and 50,000. LOD200 Scheme Demonstration Level: Preserves the overall architectural form and spatial relationships of components for multi-terminal lightweight display; when exporting the visual mesh model, the number of triangular facets of a single component is controlled between 5,000 and 10,000. The lightweighting effect is quantified by the patch simplification rate and Hausdorff distance. The formula for calculating the patch simplification rate is: ,in To simplify the dough, To reduce the number of triangles in the original model, The number of triangles in the lightweight model; Hausdorff distance formula. Quantify geometric bias, among which The Hausdorff distance between the models before and after lightweighting. The original BIM model before lightweighting. For the lightweight BIM model, for Any three-dimensional point on, for Any three-dimensional point on, for and The Euclidean distance between them; the lightweight qualification conditions are ≥70% and This ensures that fine-grained component-level information is retained after lightweighting, while reducing the amount of model data. Simultaneously, we developed multi-terminal data interfaces supporting mainstream formats such as IFC, DWG, and FBX. Among them, the IFC format is used for full-information model integration, realizing seamless transfer of integrated "geometry-attribute-damage" data with restoration design, construction management, and operation and maintenance monitoring systems; the DWG format is used for two-dimensional drawing interaction, connecting with CAD restoration design software; and the FBX format is used for visual model export, connecting with display and handover scenarios, realizing seamless integration of component-level BIM models with CAD restoration design software, construction management systems, and digital operation and maintenance monitoring systems for the entire life cycle of ancient buildings, completing the synchronization and storage of model data, so that the model can directly support restoration scheme design, construction process control, and real-time damage monitoring. S7: Dynamic Model Updates and Iterations Establish a dynamic updating and maintenance mechanism for BIM models at the component level of ancient buildings, and designate a specific person to be responsible for the full life cycle management of the model: during the construction of ancient building restoration, modify and update the BIM model in a timely manner according to the on-site construction adjustments and the actual state of the components after restoration; By combining regular inspections of ancient buildings and new damage and monitoring data generated by urban renewal projects, the model is iteratively optimized. The updated model is synchronized in real time to the digital management platform for the entire life cycle of ancient buildings to ensure that the model is consistent with the actual state of the ancient buildings, providing accurate digital support for subsequent operation and maintenance monitoring and secondary repair. At the same time, the original data, processed data, model files and modification records of the entire modeling process are uniformly stored and managed to achieve traceability of the entire modeling process.

2. The method for detailed modeling of ancient building components integrating BIM technology according to claim 1, characterized in that: The specific steps of S2 are as follows: S21: Point Cloud Denoising: Noise points are identified using a combination of neighborhood density and normal curvature methods. The neighborhood density calculation formula is as follows: ,in For the first The neighborhood density of a point cloud, For the first The number of nearest neighbors in a point cloud. for The area of ​​the neighborhood formed by the nearest neighbor points; Normal curvature calculation formula ,in For the first The normal curvature of a point cloud, These are the eigenvalues ​​of the point cloud neighborhood covariance matrix; The formula for calculating the neighborhood covariance matrix is ​​as follows: , for Mean coordinates of nearest neighbors For the first The three-dimensional coordinates of the nearest neighbor points; simultaneously satisfying and The point cloud points were identified as noise points and removed. The minimum neighborhood density threshold is the threshold corresponding to the component type. The minimum neighborhood density threshold for fine components is... Large components are ; S22: Point Cloud Registration: An improved ICP algorithm is used for coarse and fine registration. First, coarse registration of multi-view point clouds is achieved using FPFH (Fast Point Feature Histogram), and then fine registration is completed using the ICP algorithm. The registration accuracy is calculated using the root mean square error formula. Quantification, among which The root mean square error for point cloud registration, To effectively match the total number of point pairs, To match point pair numbers, The first point in the source cloud The three-dimensional spatial coordinates of the points The 4×4 registration transformation matrix (including the 3×3 rotation matrix) and 3×1 translation vector ,satisfy ), In the target point cloud The corresponding number The three-dimensional spatial coordinates of the matching points for With scriptures Mapped The Euclidean distance between them; the registration convergence condition is And the two adjacent iterations Difference ; S23: Point Cloud Segmentation: Point clouds of the same component are segmented using the feature similarity method. The feature similarity calculation formula is as follows: ,in For the first , Feature similarity of point clouds We set the weighting coefficient to 0.

6. For the first , The angle between the normal vectors of the point cloud. for cosine value, For the first , Euclidean distance between points in the cloud Distance threshold; Points with a value ≥0.7 are considered to be the same component point cloud, with fine components being... Set to 1mm, for large components Set to 5mm; after traversing all point clouds, obtain the point cloud model of each independent component, complete the component-level segmentation, and at the same time perform preliminary matching between the high-definition texture data and the corresponding component point cloud; S24: Point Cloud Simplification: While preserving the key details of the components, voxel downsampling is performed on the dense point cloud to simplify it, thereby reducing the amount of data processing required for subsequent modeling.

Images