A digital technology-based ancient building repair modeling reconstruction method

Abstract

The application provides an ancient building repair modeling reconstruction method based on digital technology, and belongs to the field of artificial intelligence and building repair technology. The method comprises the following steps: obtaining three-dimensional structure information and plane structure information of a target ancient building to be repaired; generating a three-dimensional structure diagram and a target plane structure diagram of the target ancient building to be repaired; performing matching search in a preset plane structure diagram database based on the target plane structure diagram to obtain at least one matching plane structure diagram; performing semantic analysis on the matching plane structure diagram to obtain a plurality of three-dimensional reconstruction prompt words; inputting the three-dimensional reconstruction prompt words and the three-dimensional structure diagram of the target ancient building to be repaired into a graph generation artificial intelligence engine to output a plurality of three-dimensional structure repair effect candidate diagrams; and evaluating to obtain a three-dimensional structure repair scheme diagram of the target ancient building to be repaired. The application realizes ancient building modeling reconstruction based on digital technology, and assists the artificial intelligence engine to obtain a plurality of candidate schemes and then evaluates to obtain an optimal repair scheme.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence and building restoration technology, and particularly relates to a method for modeling and reconstructing ancient building restoration based on digital technology, a computer-readable storage medium for implementing the method, a computer program product, and an electronic device. Background Technology

[0002] Ancient buildings, as important historical architectural relics, bear witness to the development of human civilization and history, reflecting the achievements of ancient working people in architectural engineering and cultural arts. These relics, having endured thousands of years of changes, have all suffered varying degrees of damage and destruction from both human activities and nature. Cultural inheritance and protection have become crucial aspects of the preservation of human cultural heritage. Ancient buildings represent an important stage in human development, and have been affected by time, environment, and human factors over a long period, with many gradually fading from historical view. Therefore, it is essential to adopt effective measures and artistic methods to maintain the original appearance of ancient buildings.

[0003] With the development of digital technology, significant achievements have been made in the preservation and restoration of ancient architecture. For example, in the process of restoring and protecting ancient buildings, corresponding models are established based on the original data of the buildings. Digital technology is used to predict the damage to the existing parts and simulate the restoration effect and make adjustments in advance. Chinese invention patent application CN202411203655.9 proposes a digital restoration model design system and restoration method for ancient buildings. By using deep learning algorithms to automatically detect damage to the three-dimensional digital model, it can more accurately identify signs of damage and quantify the degree of damage, avoiding the problem that manual inspection is difficult to discover damage in hidden parts.

[0004] As digital technology evolves to the stage of artificial intelligence, deeply evolved AI engines can now replace human designers in completing various repair simulation tasks. For example, by inputting relevant drawings of the building to be repaired into the AI ​​engine, the AI ​​engine can generate recommended repair simulation solutions based on big data models.

[0005] However, in practical applications, it has been found that if the existing 3D measurement data or drawings of the building to be repaired are simply input, the results (repair simulation effect) output by the AI ​​engine are not ideal. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention proposes a method for modeling and reconstructing ancient buildings based on digital technology, a computer-readable storage medium for implementing the method, a computer program product, and an electronic device.

[0007] In a first aspect of the present invention, a method for modeling and reconstructing ancient buildings based on digital technology is proposed. The method is implemented using electronic devices, including three-dimensional laser scanning devices, computer image processing devices, and database combination devices, etc.

[0008] The method includes the following steps:

[0009] S100: Obtain the three-dimensional and planar structural information of the ancient building to be restored;

[0010] S200: Based on the three-dimensional structural information, generate a three-dimensional structural diagram of the target ancient building to be restored; based on the planar structural information, obtain a target planar structural diagram of the target ancient building;

[0011] S300: Based on the target planar structure diagram, perform a matching search in a preset planar structure diagram database to obtain at least one matching planar structure diagram;

[0012] S400: Perform semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words;

[0013] S500: Input the three-dimensional reconstruction prompts and the three-dimensional structural diagram of the target ancient building to be repaired into the image-generated image artificial intelligence engine, and the image-generated image artificial intelligence engine outputs multiple candidate images of the three-dimensional structural repair effect of the target ancient building to be repaired;

[0014] S600: Evaluate the multiple candidate images of the three-dimensional structural repair effect to obtain the three-dimensional structural repair scheme of the target ancient building to be repaired.

[0015] Step S100 involves obtaining the three-dimensional structural information of the target ancient building to be restored, specifically including:

[0016] A laser scanner was used to scan the ancient building to be restored from multiple angles to obtain a discrete three-dimensional point cloud dataset.

[0017] The multiple angles include at least a first reference angle φ1, a second deviation angle φ2 that deviates from the first reference angle φ1 by a first preset arc θ1, and a third deviation angle φ3 that deviates from the first reference angle φ1 by the second preset arc θ2.

[0018] Gaussian mapping is performed on each subset of discrete 3D point cloud data collected at the same angle to obtain a 3D point cloud dataset after Gaussian mapping.

[0019] Each subset of discrete 3D point cloud data acquired at the same angle undergoes Gaussian mapping processing, specifically including:

[0020] Let the angle φ iScanning yields a discrete 3D point cloud data subset D i D i ={X i1 ,X i2 ,L,X iN}, i = 1, 2, 3;

[0021] For discrete 3D point cloud data subset D i Each element X in ij ,j=1,2,L,N; Perform Gaussian mapping.

[0022] The 3D structure information is obtained by stitching and fitting the 3D point cloud dataset processed by the Gaussian mapping.

[0023] Step S100 obtains the planar structural information of the target ancient building to be restored, specifically including:

[0024] Obtain the BIM simulation model of the target ancient building;

[0025] The BIM simulation model is improved based on the discrete 3D point cloud dataset.

[0026] The improved BIM simulation model is used to generate the planar structural information of the target ancient building to be restored.

[0027] Step S200, based on the aforementioned planar structure information, obtains the target planar structure diagram of the target ancient building, specifically including:

[0028] The improved BIM simulation model can be used to quickly generate standard CAD floor plans.

[0029] The matching planar structure diagram obtained in step S300 is a complete planar structure diagram containing the matching target building;

[0030] Step S400 performs semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words, specifically including:

[0031] Identify the differences between the complete planar structure diagram and the target planar structure diagram;

[0032] The differences are analyzed to determine the multiple 3D reconstruction prompts.

[0033] The process of analyzing the differences and determining the multiple 3D reconstruction prompts specifically includes:

[0034] Obtain the difference planar structure diagram corresponding to the difference portion;

[0035] Based on the aforementioned differential planar structure diagram, the morphological data of the differential parts are determined, including size, orientation, architectural style, axis, symmetry, or any combination thereof.

[0036] Based on each of the morphological data, the three-dimensional reconstruction prompt words are determined.

[0037] Furthermore, after obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes:

[0038] Obtain the planar structure repair diagram corresponding to the three-dimensional structure repair diagram;

[0039] The planar structure repair diagram is updated to the preset planar structure diagram database.

[0040] Furthermore, after obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes:

[0041] Digitalized three-dimensional structural repair plan diagram;

[0042] The three-dimensional structural repair plan diagram and the target plan structural diagram are associated and stored in the ancient building information database.

[0043] The aforementioned method for modeling and reconstructing ancient buildings based on digital technology can be automatically implemented through various forms of electronic devices and computer program instructions; the computer program instructions can be stored in different forms of storage media and loaded into computer electronic devices for execution.

[0044] Therefore, in a second aspect of the invention, a computer-readable storage medium is also provided for storing computer instructions that, when executed on an electronic device, cause the electronic device to perform the digital technology-based ancient building restoration modeling and reconstruction method of the first aspect.

[0045] In a third aspect of the invention, a computer device is also provided, the computer device including a processor and a memory, the memory being used to store instructions, and the processor being used to invoke the instructions in the memory, causing the computer device to execute the digital technology-based ancient building restoration modeling and reconstruction method of the first aspect.

[0046] In a fourth aspect of the invention, a computer program product is also provided, the product comprising a computer program, which, when executed, implements the digital technology-based ancient building restoration modeling and reconstruction method of the first aspect.

[0047] This invention utilizes digital technology to model and reconstruct ancient buildings, and then uses an artificial intelligence engine to generate multiple candidate solutions before evaluating and determining the optimal restoration plan. In this invention, the large model input to the AI ​​engine is no longer limited to existing 3D measurement data or drawings of the building to be restored, but includes 3D reconstruction prompts and a 3D structural diagram of the target ancient building. The 3D reconstruction prompts are derived by matching and retrieving from a relatively comprehensive pre-designed database of planar structural diagrams, which aligns with the reality that the database of planar drawings in the field of ancient architecture is relatively complete, while the database of 3D structural diagrams is relatively lacking. Therefore, the method of this invention is scalable, and the derived AI prompts can improve the targeting of the candidate solutions output by the AI ​​engine, ensuring the standardization of the ancient building restoration modeling and reconstruction process.

[0048] Further advantages of the present invention will be further detailed in the Specific Embodiments section in conjunction with the accompanying drawings. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a schematic diagram of the main process of a digital technology-based method for modeling and reconstructing ancient buildings for restoration, according to an embodiment of the present invention.

[0051] Figure 2 yes Figure 1 The flowchart illustrating the process of obtaining the three-dimensional structural information of the target ancient building in the embodiment is shown below.

[0052] Figure 3 yes Figure 1 The embodiment provides a schematic diagram illustrating the process principle of three-dimensional reconstruction prompts;

[0053] Figure 4 yes Figure 1 The embodiment provides a schematic diagram of the subsequent optimization steps for obtaining the three-dimensional structure repair scheme diagram;

[0054] Figure 5 Is to implement Figure 1 A schematic diagram of the connection of the functional hardware combination modules in the embodiment of the method. Detailed Implementation

[0055] First, it should be noted that the embodiments of the ancient building restoration modeling and reconstruction method based on digital technology mentioned in this section can be implemented by computer programs on electronic devices or systems equipped with memory and processors. The electronic devices or systems can be physical machines, virtual machines, servers, clusters, or any combination thereof.

[0056] Preferably, the electronic device can also be a human-computer interaction terminal, which can be a desktop terminal, a smart handheld terminal, a mobile terminal, etc., with a human-computer interaction interface.

[0057] First see Figure 1 , Figure 1 This is a schematic diagram of the main process of a digital technology-based method for modeling and reconstructing ancient buildings for restoration, according to an embodiment of the present invention.

[0058] Figure 1 The method is implemented based on three-dimensional laser scanning equipment, computer image processing equipment, and database combination equipment.

[0059] Figure 1 The method described includes the following steps:

[0060] S100: Obtain the three-dimensional and planar structural information of the ancient building to be restored;

[0061] S200: Based on the three-dimensional structural information, generate a three-dimensional structural diagram of the target ancient building to be restored; based on the planar structural information, obtain a target planar structural diagram of the target ancient building;

[0062] S300: Based on the target planar structure diagram, perform a matching search in a preset planar structure diagram database to obtain at least one matching planar structure diagram;

[0063] S400: Perform semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words;

[0064] S500: Input the three-dimensional reconstruction prompts and the three-dimensional structural diagram of the target ancient building to be repaired into the image-generated image artificial intelligence engine, and the image-generated image artificial intelligence engine outputs multiple candidate images of the three-dimensional structural repair effect of the target ancient building to be repaired;

[0065] S600: Evaluate the multiple candidate images of the three-dimensional structural repair effect to obtain the three-dimensional structural repair scheme of the target ancient building to be repaired.

[0066] Next, combined Figures 2-4 ,right Figure 1 The steps of each method embodiment are described in an adaptive manner.

[0067] Based on existing literature, 3D laser scanning technology offers significant advantages for surveying ground surfaces and walls. Specifically, it is highly adaptable, operating and applicable in various environments. Secondly, it greatly improves the accuracy of survey data and significantly shortens the surveying cycle, thus comprehensively enhancing the efficiency and quality of surveying work. After the survey data is entered into a specific computer system, the system can automatically create a 3D model of the ancient building, providing crucial information support for staff to develop conservation plans or schemes. Most importantly, staff can directly use 3D laser scanning technology to reconstruct ancient buildings that can no longer be seen by the public.

[0068] In practical applications, most buildings undergo laser scanning using only a single reference plane data set, upon which Gaussian mapping is applied. In this case, understanding the spatial layout of the point cloud and the surface form it occupies, combined with mathematical formulas to identify classes and feature structures, improves stitching accuracy and efficiency. Based on the Gaussian mapping data, the actual reference plane fitting work for the point cloud needs to be carried out in two steps: "shape type recognition" and "feature extraction."

[0069] However, when modeling and reconstructing ancient buildings for restoration, if only point cloud data from a single reference plane is considered, the subsequent fitting process needs to be divided into multiple steps and multiple modalities, and the resulting simulated 3D structure effect is relatively simple. Even if the shape type and features can be identified through the aforementioned Gaussian mapping processing, the subsequent stitching process still requires a thorough understanding of the data types contained in the point cloud data to determine which point cloud processing scheme to use to complete the research.

[0070] Therefore, as the first improvement of the present invention, see [reference needed]. Figure 2 Step S100 involves obtaining the three-dimensional structural information of the target ancient building to be restored, specifically including:

[0071] A laser scanner was used to scan the ancient building to be restored from multiple angles to obtain a discrete three-dimensional point cloud dataset.

[0072] The multiple angles include at least a first reference angle φ1, a second deviation angle φ2 that deviates from the first reference angle φ1 by a first preset arc θ1, and a third deviation angle φ3 that deviates from the first reference angle φ1 by the second preset arc θ2.

[0073] The discrete 3D point cloud data subsets collected at the same angle are processed by multi-angle Gaussian mapping to obtain the 3D point cloud dataset after multi-angle Gaussian mapping.

[0074] The 3D point cloud dataset processed by the multi-angle Gaussian mapping is stitched together and fitted to obtain the 3D structural information.

[0075] As an improvement of the present invention, in this embodiment, the discrete 3D point cloud data subsets collected at the same angle are subjected to multi-angle Gaussian mapping processing, specifically including:

[0076] Let the angle φ i Scanning yields a discrete 3D point cloud data subset D i D i ={X i1 ,X i2 ,L,X iN}, i = 1, 2, 3;

[0077] For discrete 3D point cloud data subset D i Each element X in ij j = 1, 2, ..., N; N > 1; perform the following Gaussian mapping:

[0078]

[0079] ω1=1,

[0080] in, X for element ij The result after Gaussian mapping.

[0081] As can be seen, the above process considers the offset of the current angle relative to the reference angle as a weight influence value when performing Gaussian mapping on discrete 3D point cloud data subsets at different angles, thereby obtaining the corresponding Gaussian mapping data for each angle. On this basis, the existing reference surface fitting method mentioned above is used, that is, the Gaussian mapping data of the three different angles are all used as new "reference surface" data. This ensures that the subsequent stitching process does not need to consider the data type contained in each point cloud data, and the data fitting and stitching process can be directly carried out, thus improving the efficiency of the entire 3D reconstruction and stitching.

[0082] Of course, in practical applications, if the existing ordinary Gaussian mapping process is actually adopted, after processing the discrete 3D point cloud data subsets at each angle using ordinary Gaussian mapping, the subsequent fitting and stitching process can continue, and the technical solution of this application can also be achieved, although the effect is not as good as the embodiment of this application. Therefore, this process (multi-angle Gaussian mapping) is the preferred technical solution.

[0083] Furthermore, step S100 obtains the planar structural information of the target ancient building to be restored, specifically including:

[0084] Obtain the BIM simulation model of the target ancient building;

[0085] The BIM simulation model is improved based on the discrete 3D point cloud dataset.

[0086] The improved BIM simulation model is used to generate the planar structural information of the target ancient building to be restored.

[0087] Floor plans can be viewed using CAD software, and BIM 3D models can be viewed using Revit;

[0088] Preferably, step S200, based on the planar structure information, obtains a target planar structure diagram of the target ancient building, specifically including:

[0089] The improved BIM simulation model can be used to quickly generate standard CAD floor plans.

[0090] The BIM simulation model is improved based on the discrete 3D point cloud dataset, specifically including:

[0091] Based on software such as Revit, various interfaces have been developed to diversify the exported file types. For example, in addition to .rvt, Revit can export files in CAD formats (.dwg / .dxf / .dgn / .acis), DWF formats, ODBC database link formats, and image file formats (.png / jpeg / jpg). It can also accept the above file formats and allows the import of point cloud data in .DPI / .RCP formats, greatly improving the scalability of BIM model data files.

[0092] Based on this, point cloud models have the advantage of allowing for a comprehensive view of the object being measured. Furthermore, in Revit, BIM simulation models improved based on point cloud models can be used to quickly generate standard CAD floor plans. The model can be freely cut as needed to more quickly and accurately restore missing drawings in all aspects.

[0093] The matching planar structure diagram obtained in step S300 is a complete planar structure diagram containing the matching target building.

[0094] In the field of ancient architecture, due to historical reasons and limitations in technological development, the number, completeness, and sophistication of the planar structural drawings corresponding to a particular ancient building are, in most cases, significantly better than the number, completeness, and sophistication of the corresponding three-dimensional structural drawings.

[0095] With this in mind, a pre-built database D containing all or part of the floor plan diagrams of various target buildings can be constructed. For a target ancient building A to be restored, its target floor plan diagram Aa can be matched and searched in the pre-built database D to obtain at least one matching floor plan diagram A'a.

[0096] The resulting matching planar structure diagram A'a is a complete planar structure diagram of another main building A' that contains all or part of the structure of the matching target building A.

[0097] See next. Figure 3 .

[0098] Step S400 performs semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words, specifically including:

[0099] Identify the differences between the complete planar structure diagram and the target planar structure diagram;

[0100] The differences are analyzed to determine the multiple 3D reconstruction prompts.

[0101] Preferably, the step of parsing the difference portion and determining the plurality of 3D reconstruction prompt words specifically includes:

[0102] Obtain the difference planar structure diagram corresponding to the difference portion;

[0103] Based on the aforementioned differential planar structure diagram, the morphological data of the differential parts are determined, including size, orientation, architectural style, axis, symmetry, or any combination thereof.

[0104] Based on each of the morphological data, the three-dimensional reconstruction prompt words are determined.

[0105] Continuing the example above, the target ancient building A is an ancient building to be restored, which has some missing or damaged parts. Therefore, its corresponding target plan structure diagram Aa (derived from the actual 3D scan data through the BIM model) must also have some missing or damaged parts. That is, the target plan structure diagram Aa is incomplete. The resulting matching plan structure diagram A'a is a complete plan structure diagram of another main building A' that contains all or part of the structure of the matching target building A. Therefore, there must be differences between the two (including differences caused by the missing or damaged parts to be restored, as well as other differences).

[0106] For these differences, the corresponding planar structural diagrams can be further determined, thereby obtaining the morphological data of the differences. The morphological data includes one or any combination of dimensions, orientation, architectural style, axis, and symmetry.

[0107] As an example, if the morphological data represents the architectural style as "European, pointed roof" and the orientation as "missing in the northeast direction", then the corresponding 3D reconstruction prompt could be "The part to be repaired is in the northeast direction, and the repair options include European style and pointed roof".

[0108] The output of 3D reconstruction prompts can be obtained by querying a pre-established morphological data-prompt mapping database, or by pre-training a prompt generation model. The training samples include morphological data-prompt association data pairs. By inputting morphological data, the generation model can output prompts, etc. This embodiment does not impose specific limitations on this. Reference can also be made to existing technologies regarding prompt generation.

[0109] Based on this, proceed with steps S500-S600:

[0110] S500: Input the three-dimensional reconstruction prompts and the three-dimensional structural diagram of the target ancient building to be repaired into the image-generated image artificial intelligence engine, and the image-generated image artificial intelligence engine outputs multiple candidate images of the three-dimensional structural repair effect of the target ancient building to be repaired;

[0111] S600: Evaluate the multiple candidate images of the three-dimensional structural repair effect to obtain the three-dimensional structural repair scheme of the target ancient building to be repaired.

[0112] As mentioned earlier, advanced artificial intelligence engines in existing technologies can already replace human designers in completing various repair simulation tasks. For example, by inputting relevant drawings of the building to be repaired into the artificial intelligence (AI) engine, the AI ​​engine can derive recommended repair simulation solutions based on big data models.

[0113] However, the above-mentioned input-output process is simple, and the output of the AI ​​engine is mostly vague and lacks effectiveness due to the lack of effective prompts. This is related to the lack of historical 3D data in the field of ancient architecture and the insufficient prediction of 3D modeling database.

[0114] Based on this, the improvement of this application is to input the three-dimensional reconstruction prompt words and the three-dimensional structural diagram of the ancient building to be restored into the image-based artificial intelligence engine, thereby giving the AI ​​engine more guidance and correction capabilities.

[0115] As an example, the image-to-image AI engine uses partial image description text and an image as input prompts to intelligently generate preliminary design images. Typical examples include Stability AI's Stable Diffusion tool released in August 2022 and Tongji CAUP's FUgenerator tool released in March 2023. It's understandable that the general image-to-image industry-standard large-scale model configured in a typical image-to-image AI engine can also simultaneously possess text-to-text, text-to-image, and image-to-image generation capabilities.

[0116] Therefore, as Example 1, when the three-dimensional reconstruction prompt and the three-dimensional structural diagram of the target ancient building to be repaired are input into the image-based AI engine, the image-based AI engine may first use the three-dimensional reconstruction prompt to generate further three-dimensional reconstruction repair components, and then stitch and fuse them with the three-dimensional structural diagram of the target ancient building to be repaired to obtain a candidate image of the three-dimensional structural repair effect.

[0117] As another example 2, when the three-dimensional reconstruction prompt words and the three-dimensional structural diagram of the ancient building to be repaired are input into the image-based AI engine, the image-based AI engine may first directly generate multiple preliminary three-dimensional structural repair effect candidate images, and then consider the three-dimensional reconstruction prompt words to perform correction and adjustment, thereby obtaining the three-dimensional structural repair effect candidate images.

[0118] Of course, the image-based AI engine can also use other methods. The above description is only an example. The working principle of each image-based AI engine may be different. However, based on the above introduction, they can all achieve the purpose of outputting multiple candidate images of 3D structure repair effect in this application.

[0119] For example, Example 2 may also include generating further three-dimensional reconstruction and repair components using the three-dimensional reconstruction prompts, and then splicing and fusing them with the three-dimensional structural diagram of the ancient building to be repaired to obtain a candidate image of the three-dimensional structural repair effect, i.e., the process of fusing Example 1 and Example 2, etc.

[0120] Step S600 evaluates the multiple candidate images of the three-dimensional structure repair effect. This evaluation can be done by computer-automated scoring, expert experience scoring, or a combination of both. All evaluation and scoring methods are existing technologies and will not be described in detail here.

[0121] Based on the above embodiments, see Figure 4 , Figure 4 yes Figure 1 The embodiment provides a schematic diagram of the subsequent optimization steps for obtaining the three-dimensional structure repair scheme diagram.

[0122] Specifically, after obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes:

[0123] Obtain the planar structure repair diagram corresponding to the three-dimensional structure repair diagram;

[0124] The planar structure repair diagram is updated to the preset planar structure diagram database.

[0125] The above preferred embodiments enable standardized data utilization of the method, thereby achieving closed-loop learning and updating of the database throughout the entire process.

[0126] Furthermore, after obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes:

[0127] Digitalized three-dimensional structural repair plan diagram;

[0128] The three-dimensional structural repair plan diagram and the target plan structural diagram are associated and stored in the ancient building information database.

[0129] In the above example, by creating an ancient building information database, the information on ancient buildings can be digitally recorded and stored, making it easier for staff from various departments to quickly and accurately retrieve and apply the information they need, thereby promoting the efficient and stable development of ancient building protection work.

[0130] Preferably, a database of information on ancient buildings is created, specifically including:

[0131] First, it is necessary to consider the purpose of database construction and the data content (such as the name, age, and style of ancient buildings).

[0132] First, the database structure and fields should be rationally determined based on factors such as data characteristics to ensure the accuracy and comprehensiveness of the collected ancient building information. Second, technical personnel should select a suitable ancient building information database management system based on their needs, such as MySQL or Oracle. Different database management systems have different functions and features, and technical personnel can choose according to specific circumstances. Third, designers need to create table indexes based on the ancient building information query requirements of relevant departments to further improve the efficiency of data retrieval. Finally, technical personnel need to regularly perform database backup operations to prevent data loss, ensure the safe storage of backup files, and test the backup file recovery process to ensure that data information can be successfully restored after deletion.

[0133] Preferably, when entering and managing data on ancient buildings, the following are also included:

[0134] First, the collected information on ancient buildings will be promptly entered into the ancient building information database through manual or digital scanning methods, providing convenience for people to understand and access information about ancient buildings. Second, the database will be managed and organized, with basic functions such as adding, modifying, and deleting data added. It will also support operations such as classifying and tagging data information, and enable accurate retrieval and quick access to ancient building information by creating indexes. Finally, access control will be strengthened to ensure data security. Staff will ensure the confidentiality and integrity of the ancient building information in the database by setting security measures and defining access permissions for different personnel.

[0135] The aforementioned cloud-based digital technology-based ancient building restoration modeling and reconstruction method can be automatically implemented through various forms of electronic devices and computer-readable program instructions; the computer-readable program instructions can be stored in different forms of storage media and loaded into computer electronic devices for execution.

[0136] According to an exemplary implementation of this disclosure, a computer-readable storage medium is provided that stores computer-executable instructions thereon, wherein the computer-executable instructions are executed by a processor to implement the methods described above. According to an exemplary implementation of this disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, which are executed by a processor to implement the methods described above.

[0137] In the introduction Figures 1-4 Based on the principles of the methods and embodiments involved, Figure 5 Implementation shown Figure 1 A schematic diagram of the connection of the functional hardware combination modules in the embodiment of the method.

[0138] exist Figure 5 The implementation is shown in the middle. Figure 1 The method includes the following related hardware functional components:

[0139] Three-dimensional laser scanning equipment is used to acquire the three-dimensional structural information of the ancient building to be restored;

[0140] BIM simulation model, based on the improved point cloud model, quickly generates standard CAD floor plans;

[0141] Computer image processing equipment is used to build three-dimensional models of ancient buildings and reproduce them; data resources are acquired to form a comprehensive database; comparison technology is used to compare the acquired images related to the shape, color, texture, etc. of ancient buildings, and to give an accurate and comprehensive evaluation based on the comparison results, providing an important reference for creating ancient building protection plans.

[0142] Specifically, based on the three-dimensional structural information, a three-dimensional structural diagram of the target ancient building to be restored is generated; based on the planar structural information, a computer image processing device uses a BIM simulation model to obtain a target planar structural diagram of the target ancient building.

[0143] Figure 5 It also shows Figure 1 The functions of the preset planar structure database, ancient building information database, and graph-based artificial intelligence engine used in the method embodiments have been described in the foregoing method embodiments and will not be repeated here.

[0144] Other technologies, principles, algorithms, or models not elaborated in detail in this application can be found in the prior art.

[0145] This invention utilizes digital technology to model and reconstruct ancient buildings, and then uses an artificial intelligence engine to generate multiple candidate solutions before evaluating and determining the optimal restoration plan. In this invention, the large model input to the AI ​​engine is no longer limited to existing 3D measurement data or drawings of the building to be restored, but includes 3D reconstruction prompts and a 3D structural diagram of the target ancient building. The 3D reconstruction prompts are derived by matching and retrieving from a relatively comprehensive pre-designed database of planar structural diagrams, which aligns with the reality that the database of planar drawings in the field of ancient architecture is relatively complete, while the database of 3D structural diagrams is relatively lacking. Therefore, the method of this invention is scalable, and the derived AI prompts can improve the targeting of the candidate solutions output by the AI ​​engine, ensuring the standardization of the ancient building restoration modeling and reconstruction process.

[0146] In the foregoing embodiments section, the present invention provides multiple embodiments, each of which can constitute an independent technical solution and may contribute to the prior art, and solve corresponding technical problems. However, it should be noted that different embodiments can be combined with each other without violating logic; at the same time, each embodiment can solve at least one technical problem, but it is not required that each individual embodiment solve multiple or all technical problems.

[0147] Furthermore, in specific embodiments of this application, if user-related data is involved, user permission or consent must be obtained when the embodiments of this application are applied to specific products or technologies, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0148] Various implementations of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed implementations. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described implementations. The terminology used herein is chosen to best explain the principles, practical applications, or improvements to technology in the market, or to enable others skilled in the art to understand the various implementations disclosed herein.

Claims

1. A method for modeling and reconstructing ancient buildings based on digital technology, the method being implemented using electronic devices, characterized in that... The method includes the following steps: S100: Obtain the three-dimensional and planar structural information of the ancient building to be restored; S200: Based on the three-dimensional structural information, generate a three-dimensional structural diagram of the ancient building to be restored; Based on the aforementioned planar structure information, a target planar structure diagram of the target ancient building is obtained; S300: Based on the target planar structure diagram, perform a matching search in a preset planar structure diagram database to obtain at least one matching planar structure diagram; S400: Perform semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words; S500: Input the three-dimensional reconstruction prompts and the three-dimensional structural diagram of the target ancient building to be repaired into the image-generated image artificial intelligence engine, and the image-generated image artificial intelligence engine outputs multiple candidate images of the three-dimensional structural repair effect of the target ancient building to be repaired; S600: Evaluate the multiple candidate images of the three-dimensional structural repair effect to obtain a three-dimensional structural repair scheme image of the target ancient building to be repaired; the matching planar structural image obtained in step S300 is a complete planar structural image containing the matching target building; Step S400 performs semantic parsing on the matching planar structure graph to obtain multiple three-dimensional reconstruction prompt words, specifically including: Identify the differences between the complete planar structure diagram and the target planar structure diagram; The differences are analyzed to determine the multiple 3D reconstruction prompts.

2. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 1, characterized in that, Step S100 involves obtaining the three-dimensional structural information of the target ancient building to be restored, specifically including: A laser scanner was used to scan the ancient building to be restored from multiple angles to obtain a discrete three-dimensional point cloud dataset. Gaussian mapping is performed on each subset of discrete 3D point cloud data collected at the same angle to obtain a 3D point cloud dataset after Gaussian mapping. The 3D structure information is obtained by stitching and fitting the 3D point cloud dataset processed by the Gaussian mapping.

3. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 2, characterized in that, The multiple angles include at least a first reference angle φ1, a second deviation angle φ2 that deviates from the first reference angle φ1 by a first preset arc θ1, and a third deviation angle φ3 that deviates from the first reference angle φ1 by the second preset arc θ2. Each subset of discrete 3D point cloud data acquired at the same angle undergoes Gaussian mapping processing, specifically including: Let the angle φ i Scanning yields a discrete 3D point cloud data subset D i D i ={X i1 ,X i2 ,……,X iN }, i = 1, 2, 3; For discrete 3D point cloud data subset D i Each element X in ij j = 1, 2, ..., N; N > 1; perform Gaussian mapping.

4. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 2, characterized in that, Step S100 obtains the planar structural information of the target ancient building to be restored, specifically including: Obtain the BIM simulation model of the target ancient building; The BIM simulation model is improved based on the discrete 3D point cloud dataset. The improved BIM simulation model is used to generate the planar structural information of the target ancient building to be restored.

5. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 4, characterized in that, Step S200, based on the aforementioned planar structure information, obtains the target planar structure diagram of the target ancient building, specifically including: The improved BIM simulation model can be used to quickly generate standard CAD floor plans.

6. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 1, characterized in that, The process of analyzing the differences and determining the multiple 3D reconstruction prompts specifically includes: Obtain the difference planar structure diagram corresponding to the difference portion; Based on the aforementioned differential planar structure diagram, the morphological data of the differential parts are determined, including size, orientation, architectural style, axis, symmetry, or any combination thereof. Based on each of the morphological data, the three-dimensional reconstruction prompt words are determined.

7. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 1, characterized in that, After obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes: Obtain the planar structure repair diagram corresponding to the three-dimensional structure repair diagram; The planar structure repair diagram is updated to the preset planar structure diagram database.

8. The method for modeling and reconstructing ancient buildings based on digital technology as described in claim 1, characterized in that, After obtaining the three-dimensional structural restoration plan of the target ancient building in step S600, the method further includes: Digitalized three-dimensional structural repair plan diagram; The three-dimensional structural repair plan diagram and the target plan structural diagram are associated and stored in the ancient building information database.

9. A computer program product, comprising a computer program or computer-executable instructions, wherein when the computer program or computer-executable instructions are executed by a processor, they implement the ancient building restoration modeling and reconstruction method based on digital technology as described in any one of claims 1 to 8.

Images