An intelligent fusion method and system for historical building surveying and repairing investigation

Abstract

The application discloses a kind of historical building surveying and mapping and repair survey intelligent fusion method, comprising the following steps: S1, the three-dimensional point cloud data and disease data of historical building are collected by non-contact surveying and mapping technology;S2, point cloud data and disease data are registered using adaptive fusion algorithm, to generate high-precision three-dimensional model;S3, three-dimensional model is automatically identified and quantified based on deep learning model;S4, disease data and repair suggestions are integrated into BIM model, and interaction is realized through mobile terminal. Through multi-source data fusion and adaptive algorithm, the surveying and mapping accuracy and disease detection reliability are improved;Using deep learning and historical data correlation, intelligent diagnosis and cause analysis of disease are realized;Through BIM and AR integration, the surveying and mapping efficiency and on-site operation convenience are improved;Whole-process platform management ensures data consistency and construction traceability.

Description

Technical Field

[0001] This invention relates to the field of architectural surveying and restoration technology, and in particular to an intelligent surveying and restoration method and system for historical buildings that combines non-contact surveying, deep learning and building information modeling (BIM). Background Technology

[0002] The protection and restoration of historical buildings requires high-precision surveying and diagnosing of damage, but traditional methods have the following problems: manual surveying is prone to causing secondary damage to the building and the data accuracy is low; damage identification relies on experience and lacks automated and quantitative means; the survey results are disconnected from design and construction, and information sharing is not timely.

[0003] While existing technologies employ drones or lidar for surveying, they are largely limited to single technologies and lack multi-source data fusion and intelligent analysis. Disease identification is primarily based on image processing but not deeply integrated with 3D models and historical data. BIM applications are mostly used for new buildings, lacking dynamic interaction and real-time updating capabilities in the renovation of historical buildings. Therefore, an integrated and intelligent solution is urgently needed. Summary of the Invention

[0004] This invention aims to solve the problems of insufficient accuracy, low efficiency, and information silos in traditional historical building surveying and restoration, and achieves full-process intelligence and visualization through the integration of multiple technologies.

[0005] To solve one of the aforementioned technical problems, the following technical solution is adopted: This application provides an intelligent fusion method for historical building surveying and restoration investigation, including the following steps: S1. Collect three-dimensional point cloud data and damage data of historical buildings through non-contact surveying technology; S2. Use an adaptive fusion algorithm to register point cloud data with disease data to generate a high-precision 3D model; S3. Automatic identification and quantification of defects in 3D models based on deep learning models; S4. Integrate the disease data and repair suggestions into the BIM model, and realize interactive communication through mobile terminals.

[0006] To better achieve the purpose of the invention, the present invention also has the following superior technical solutions: In some embodiments, the non-contact mapping technology includes unmanned aerial vehicle oblique photography, lidar, and infrared thermal imaging.

[0007] The adaptive fusion algorithm employs multi-scale feature matching and includes the following sub-steps: Geometric features of point cloud data and texture features of infrared data are extracted; matching is performed using feature descriptors, and misregistration is eliminated using the RANSAC algorithm.

[0008] In some embodiments, the deep learning model is a multimodal network that combines attention mechanisms and transfer learning, and the training sample library contains common types of damage to historical buildings.

[0009] In some embodiments, the BIM model supports lightweight mobile interaction and uses AR technology to achieve augmented reality display of disease information.

[0010] In some embodiments, this application also includes building a renovation project management platform to connect data with the construction system and support construction quality traceability.

[0011] The present invention also provides a system based on the above-described intelligent fusion method, comprising: The system includes a data acquisition module for non-contact surveying; a data processing module for data fusion and model generation; a disease diagnosis module for deep learning-based identification; a BIM interaction module for visual briefings; and a management platform module for end-to-end monitoring.

[0012] In some embodiments, the data acquisition module includes a drone, a lidar, and an infrared thermal imager.

[0013] In some embodiments, the disease diagnosis module is deployed on a cloud server and supports online model updates.

[0014] In some embodiments, the BIM interaction module supports multi-terminal synchronization and offline access.

[0015] By employing the aforementioned technical solutions, high-precision 3D point cloud data and damage data of historical buildings are collected using UAV oblique photography, LiDAR, and infrared thermal imaging technologies. An adaptive fusion algorithm is used to register the point cloud data with the infrared data, generating a 3D model with millimeter-level resolution, achieving non-contact mapping and data acquisition. Furthermore, based on a multimodal deep learning model (such as an enhanced version of the AegisYOLO algorithm), the 3D model is automatically inspected to identify the type, location, and extent of damage. Historical repair records and material parameters are correlated to generate a damage cause analysis report, achieving intelligent diagnosis and quantification of damage. Damage data, survey conclusions, and repair suggestions are integrated into the BIM model, and lightweight mobile interactive functions are developed. Real-time retrieval of model data, material test results, and process requirements is supported, and augmented reality display is achieved through AR technology, enabling visualization and interactive communication within the BIM framework. The repair project management platform constructed in this application achieves seamless data integration with the construction system, supports construction quality traceability and dynamic monitoring, and realizes intelligent management throughout the entire process.

[0016] Compared with existing technologies, multi-source data fusion and adaptive algorithms improve surveying accuracy and the reliability of disease detection; deep learning and historical data association enable intelligent disease diagnosis and cause analysis; BIM and AR integration enhances the efficiency of survey briefings and the convenience of on-site operations; and platform-based management throughout the entire process ensures data consistency and construction traceability. Attached Figure Description

[0017] Figure 1 This is the overall flowchart of the intelligent integrated survey method for historical buildings of this invention; Figure 2 Flowchart of the adaptive fusion process of point cloud and infrared data in this invention; Figure 3 The structural diagram of the multimodal deep learning model for disease identification in this invention; Figure 4 A schematic diagram illustrating the interaction between the BIM model and the mobile terminal in this invention. Figure 5 System architecture diagram of the present invention. Detailed Implementation

[0018] The present invention will be further described in detail below with reference to the embodiments. The embodiments are only intended to provide a clearer understanding of the technical features, objectives and effects of the present invention.

[0019] refer to Figure 1-5 This invention provides an intelligent fusion method for historical building surveying and restoration investigation, comprising the following steps: S1. Collect three-dimensional point cloud data and damage data of historical buildings through non-contact surveying technology; S2. Use an adaptive fusion algorithm to register point cloud data with disease data to generate a high-precision 3D model; S3. Automatic identification and quantification of defects in 3D models based on deep learning models; S4. Integrate the disease data and repair suggestions into the BIM model, and realize interactive communication through mobile terminals.

[0020] To better achieve the purpose of the invention, the present invention also has the following superior technical solutions: In some embodiments, the non-contact mapping technology includes unmanned aerial vehicle oblique photography, lidar, and infrared thermal imaging.

[0021] The adaptive fusion algorithm employs multi-scale feature matching and includes the following sub-steps: Geometric features of point cloud data and texture features of infrared data are extracted; matching is performed using feature descriptors, and misregistration is eliminated using the RANSAC algorithm.

[0022] In some embodiments, the deep learning model is a multimodal network that combines attention mechanisms and transfer learning, and the training sample library contains common types of damage to historical buildings.

[0023] In some embodiments, the BIM model supports lightweight mobile interaction and uses AR technology to achieve augmented reality display of disease information.

[0024] In some embodiments, this application also includes building a renovation project management platform to connect data with the construction system and support construction quality traceability.

[0025] The present invention also provides a system based on the above-described intelligent fusion method, comprising: The system includes a data acquisition module for non-contact surveying; a data processing module for data fusion and model generation; a disease diagnosis module for deep learning-based identification; a BIM interaction module for visual briefings; and a management platform module for end-to-end monitoring.

[0026] In some embodiments, the data acquisition module includes a drone, a lidar, and an infrared thermal imager.

[0027] In some embodiments, the disease diagnosis module is deployed on a cloud server and supports online model updates.

[0028] In some embodiments, the BIM interaction module supports multi-terminal synchronization and offline access.

[0029] By employing the aforementioned technical solutions, high-precision 3D point cloud data and damage data of historical buildings are collected using UAV oblique photography, LiDAR, and infrared thermal imaging technologies. An adaptive fusion algorithm is used to register the point cloud data with the infrared data, generating a 3D model with millimeter-level resolution, achieving non-contact mapping and data acquisition. Furthermore, based on a multimodal deep learning model (such as an enhanced version of the AegisYOLO algorithm), the 3D model is automatically inspected to identify the type, location, and extent of damage. Historical repair records and material parameters are correlated to generate a damage cause analysis report, achieving intelligent diagnosis and quantification of damage. Damage data, survey conclusions, and repair suggestions are integrated into the BIM model, and lightweight mobile interactive functions are developed. Real-time retrieval of model data, material test results, and process requirements is supported, and augmented reality display is achieved through AR technology, enabling visualization and interactive communication within the BIM framework. The repair project management platform constructed in this application achieves seamless data integration with the construction system, supports construction quality traceability and dynamic monitoring, and realizes intelligent management throughout the entire process.

[0030] Example 1, Data Acquisition Stage: A DJI Matrice 300 RTK drone equipped with a Zenmuse L1 LiDAR and an H20T infrared thermal imager was used to scan the historical building from multiple angles, generating point cloud data (resolution up to 2mm) and infrared images. The two were then registered using an adaptive fusion algorithm (based on feature point matching and scale-invariant transformation) to mark the locations of defects such as cracks and hollow areas.

[0031] Example 2, Disease Diagnosis Stage: A multimodal deep learning model was constructed within the PyTorch framework, based on AegisYOLO, incorporating attention mechanisms and transfer learning. The training sample library included 12 types of diseases, such as timber decay and brick and stone weathering. The model outputs quantitative indicators of diseases (such as crack width and weathering area) and automatically generates analysis reports.

[0032] Example 3, BIM handover phase: A BIM model is created using Autodesk Revit, integrating defect data and repair suggestions. A mobile app is developed using Unity, supporting Android and iOS systems. On-site personnel can access the model by scanning a QR code and view a virtual model overlaid with defect information using AR glasses.

[0033] Example 4, System Integration: A management backend is built based on a cloud platform (such as Alibaba Cloud) to realize data storage, model updates, and construction progress tracking. The system interface supports integration with common engineering management software (such as Primavera) to achieve quality traceability.

[0034] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A smart fusion method for historical building surveying and restoration investigation, characterized in that, Includes the following steps: S1. Collect three-dimensional point cloud data and damage data of historical buildings through non-contact surveying technology; S2. Use an adaptive fusion algorithm to register point cloud data with disease data to generate a high-precision 3D model; S3. Automatic identification and quantification of defects in 3D models based on deep learning models; S4. Integrate the disease data and repair suggestions into the BIM model, and realize interactive communication through mobile terminals.

2. The intelligent fusion method according to claim 1, characterized in that, The non-contact mapping technologies include UAV oblique photography, lidar, and infrared thermal imaging.

3. The intelligent fusion method according to claim 1, characterized in that, The adaptive fusion algorithm employs multi-scale feature matching and includes the following sub-steps: S1. Extract the geometric features of point cloud data and the texture features of infrared data; S2. Matching is performed using feature descriptors, and misregistration is eliminated using the RANSAC algorithm.

4. The intelligent fusion method according to claim 1, characterized in that, The deep learning model described is a multimodal network that combines attention mechanisms and transfer learning. The training sample library contains common types of damage to historical buildings.

5. The intelligent fusion method according to claim 1, characterized in that, The BIM model supports lightweight mobile interaction and uses AR technology to display disease information in an augmented reality manner.

6. The intelligent fusion method according to claim 1, characterized in that, It also includes building a renovation project management platform to connect data with the construction system and support traceability of construction quality.

7. A system for implementing the intelligent fusion method according to any one of claims 1-6, characterized in that, include: Data acquisition module for non-contact surveying; The data processing module is used for data fusion and model generation; Disease diagnosis module, used for deep learning recognition; BIM interactive module, used for visual handover; The management platform module is used for end-to-end monitoring.

8. The system according to claim 7, characterized in that, The data acquisition module includes a drone, a lidar, and an infrared thermal imager.

9. The system according to claim 7, characterized in that, The disease diagnosis module is deployed on a cloud server and supports online model updates.

10. The system according to claim 7, characterized in that, The BIM interaction module supports multi-terminal synchronization and offline access.

Images