A method and system for modeling digital twins of smart buildings

Abstract

The present invention relates to the field of image processing technology, and in particular to a method and system for modeling a digital twin of a smart building, the method comprising: obtaining real-time construction data and aerial image data corresponding to a target building; constructing an initial digital twin model of a smart building corresponding to the target building according to the real-time construction data; matching the architectural structural features of the target building according to the aerial image data and the image of the initial digital twin model of the smart building in the same direction, and obtaining the degree of matching between the aerial image data and the initial digital twin model of the smart building according to the matched architectural structural features; and correcting the initial digital twin model of the smart building according to the degree of matching to obtain the digital twin model of the target smart building. In the present application, in the process of managing smart buildings, the digital twin model of the smart building is corrected by the aerial image data, thereby improving the authenticity of the digital twin model.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method and system for modeling a digital twin of a smart building. Background Art

[0002] In the context of the continuous high-level operation of the digital economy, smart buildings, as an important part of the digital transformation of cities, their digital twin modeling is a key technology to achieve the deep integration of physical space and digital space. This technology can promote the digitalization, networking and intelligent transformation of the construction industry. Digital twin technology is a product of the integration of new generation information technologies such as big data, cloud computing, 5G (fifth generation mobile communication technology), and artificial intelligence. It establishes accurate digital models for physical entities, realizes comprehensive mapping and real-time interaction of physical entities, and provides strong data support and decision-making basis for smart buildings. The models constructed by traditional digital twin modeling methods are usually static and cannot reflect the state changes of buildings in real time. Therefore, some existing technologies choose to use BIM (Building Information Modeling) to build building models in real time. For customers and construction parties, they need to have a certain understanding of the current status of the project during the construction process, but it may not be convenient for customers and construction parties to observe the construction progress and construction situation on site. The use of BIM models not only facilitates customers and construction parties to understand the progress of the project, but also facilitates the collaborative work between various construction departments.

[0003] However, since the BIM model is constructed through parameters such as material consumption, and there is inevitable waste of some materials during the construction process, there will be differences between the obtained digital twin model and the real building, which will lead to the lower authenticity of the digital twin model.

[0004] Therefore, how to improve the authenticity of the digital twin model is an urgent problem that needs to be solved. Summary of the invention

[0005] In order to solve the technical problem of how to improve the authenticity of the digital twin model, the purpose of the present invention is to provide a method and system for modeling a digital twin of a smart building. The technical solutions adopted are as follows:

[0006] The present application provides a method for modeling a digital twin of a smart building, the method comprising:

[0007] Obtain real-time construction data and aerial image data corresponding to the target building;

[0008] Constructing an initial smart building digital twin model corresponding to the target building according to the real-time construction data;

[0009] According to the aerial image data and the image of the initial smart building digital twin model in the same direction, the architectural structure features of the target building are matched, and according to the matched architectural structure features, the matching degree between the aerial image data and the initial smart building digital twin model is obtained;

[0010] According to the matching degree, the initial smart building digital twin model is modified to obtain the target smart building digital twin model.

[0011] In some embodiments, the real-time construction data includes: building material consumption data, building geometry data, construction cost data, construction progress data, construction environment data, construction safety data, and construction equipment data of the target building during real-time construction.

[0012] In some embodiments, obtaining aerial image data corresponding to a target building includes:

[0013] The aerial image data is obtained by photographing the entire target building through a plurality of pre-configured drones, wherein each of the drones flies around the target building.

[0014] In some embodiments, after obtaining the aerial image data corresponding to the target building, the method further includes:

[0015] The aerial image data is processed by a preset histogram equalization function to obtain aerial image data without illumination influence;

[0016] The aerial image data without illumination influence is processed by a preset grayscale function to obtain preprocessed aerial image data, and the preprocessed aerial image data is used to match the architectural structure features of the target building.

[0017] In some embodiments, matching the architectural structure features of the target building according to the aerial image data and the image of the initial smart building digital twin model in the same direction, and obtaining the matching degree between the aerial image data and the initial smart building digital twin model according to the matched architectural structure features, includes:

[0018] Perform regularity performance analysis on similar building structure features in the aerial image data and the image of the initial smart building digital twin model in the same direction, and divide the aerial image data into multiple regions according to the regularity performance analysis results, and divide the image of the initial smart building digital twin model into multiple regions according to the regularity performance analysis results;

[0019] Matching the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model;

[0020] According to the matching accuracy, the matching degree between the aerial image data and the initial smart building digital twin model is obtained.

[0021] In some embodiments, the performing regularity performance analysis on similar building structure features in the images of the aerial image data and the initial smart building digital twin model in the same direction, and dividing the aerial image data into a plurality of regions according to the regularity performance analysis results, and dividing the image of the initial smart building digital twin model into a plurality of regions according to the regularity performance analysis results, includes:

[0022] Acquire a first edge detection image and a first threshold segmentation image corresponding to the aerial image data, and a second edge detection image and a second threshold segmentation image corresponding to the image of the initial smart building digital twin model;

[0023] Dividing the data sequence corresponding to similar building structure features in the first edge detection image into a plurality of first feature groups, and performing regularity performance analysis on each of the first feature groups to obtain a first regularity performance degree corresponding to the first edge detection image;

[0024] Dividing the data sequence corresponding to similar building structure features in the first threshold segmentation image into a plurality of second feature groups, and performing regularity performance analysis on each of the second feature groups to obtain a second regularity performance degree corresponding to the first threshold segmentation image;

[0025] The image corresponding to the regularity expression degree closest to zero among the first regularity expression degree and the second regularity expression degree is used as a target aerial image, and similar building structure features in the target aerial image are divided into a plurality of first main areas, and each of the first main areas is divided into a plurality of first sub-areas;

[0026] Dividing the data sequence corresponding to the similar building structure features in the second edge detection image into a plurality of third feature groups, and performing regularity performance analysis on each of the third feature groups to obtain a third regularity performance degree corresponding to the second edge detection image;

[0027] Dividing the data sequence corresponding to similar building structure features in the second threshold segmentation image into a plurality of fourth feature groups, and performing regularity performance analysis on each of the fourth feature groups to obtain a fourth regularity performance degree corresponding to the second threshold segmentation image;

[0028] The image corresponding to the regularity expression degree closest to zero among the third regularity expression degree and the fourth regularity expression degree is taken as the target model image, and similar architectural structure features in the target model image are divided into multiple second main areas, and each of the second main areas is divided into multiple second sub-areas.

[0029] In some embodiments, the matching of the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model, includes:

[0030] Matching the building structure features in the target aerial image and the target model image to obtain a plurality of matching feature groups;

[0031] Determine the position difference of each of the matching feature groups according to the first main area and the first sub-area corresponding to the building structure feature in each of the matching feature groups, and the second main area and the second sub-area corresponding to the building structure feature;

[0032] The matching accuracy corresponding to each matching feature group is determined by combining the position difference of each matching feature group and the corresponding regularity expression degree of the target aerial image and the target model image.

[0033] In some embodiments, obtaining the degree of matching between the aerial image data and the initial smart building digital twin model according to the matching accuracy includes:

[0034] The matching feature groups whose matching accuracy is greater than or equal to a preset matching accuracy threshold are taken as target matching feature groups, and the number of the target matching feature groups is obtained;

[0035] Comparing the quantity of the target matching feature group with the quantity of all the matching feature groups to obtain a quantity comparison result;

[0036] The matching degree is determined by combining the quantitative comparison result with the average of the matching accuracy corresponding to all the matching feature groups.

[0037] In some embodiments, the initial smart building digital twin model is modified according to the matching degree to obtain a target smart building digital twin model, including:

[0038] comparing the matching degree with a preset matching degree threshold;

[0039] When the matching degree is less than or equal to the preset matching degree threshold, correcting the real-time construction data according to the initial smart building digital twin model;

[0040] Based on the corrected real-time construction data, a digital twin model of the target smart building is constructed.

[0041] The embodiment of the present application also provides a smart building digital twin modeling system, the system comprising:

[0042] A data acquisition module is used to acquire real-time construction data and aerial image data corresponding to the target building;

[0043] An initial model building module, used to build an initial smart building digital twin model corresponding to the target building according to the real-time construction data;

[0044] A matching module is used to match the architectural structure features of the target building according to the aerial image data and the image of the initial smart building digital twin model in the same direction, and obtain the matching degree between the aerial image data and the initial smart building digital twin model according to the matched architectural structure features;

[0045] The correction module is used to correct the initial smart building digital twin model according to the matching degree to obtain the target smart building digital twin model.

[0046] The present invention has the following beneficial effects:

[0047] First, obtain the real-time construction data and aerial image data corresponding to the target building; then, according to the real-time construction data, construct the initial smart building digital twin model corresponding to the target building; then, according to the image of the aerial image data and the initial smart building digital twin model in the same direction, match the architectural structure features of the target building, and according to the matched architectural structure features, obtain the matching degree between the aerial image data and the initial smart building digital twin model; finally, according to the matching degree, correct the initial smart building digital twin model to obtain the target smart building digital twin model. In this application, the real-time construction data provides the real-time status and detailed information of the target building to ensure that the initial smart building digital twin model can reflect the current actual situation of the target building; the aerial image data provides an external view of the target building, which can capture the details and appearance features of the building structure; by matching the architectural structure features of the target building and obtaining the matching degree, it is possible to verify whether the geometric structure and appearance features of the initial smart building digital twin model are consistent, and find the difference between the initial smart building digital twin model and the real building; correcting the initial smart building digital twin model according to the matching degree can eliminate the above differences and improve the authenticity of the digital twin model. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A schematic diagram of an implementation environment of a smart building digital twin modeling method provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of a flow chart of a method for modeling a digital twin of a smart building provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of a drone provided by an embodiment of the present invention photographing a target building;

[0052] Figure 4 A schematic diagram of a first edge detection image provided by an embodiment of the present invention;

[0053] Figure 5 A schematic diagram of a first threshold value segmented image provided by an embodiment of the present invention;

[0054] Figure 6 A schematic diagram of area division provided by an embodiment of the present invention;

[0055] Figure 7 A schematic structural diagram of a smart building digital twin modeling system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a smart building digital twin modeling method and system proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0057] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0058] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0059] The specific scheme of a smart building digital twin modeling method and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0060] See also Figure 1 , Figure 1 A schematic diagram of an implementation environment of a smart building digital twin modeling method provided by an embodiment of the present invention. Figure 1As shown, the implementation environment includes a modeling terminal 101, a real-time construction data acquisition terminal 102, and an aerial image data acquisition terminal 103. The modeling terminal 101 can be a terminal device equipped with a modeling system, including but not limited to a laptop computer, a tablet computer, a handheld computer, a PAD (tablet computer), a desktop computer, etc. with local computing capabilities; the modeling system can be implemented in the form of a target client, and the target client can be a video client, an instant messaging client, a browser client, etc. that supports the digital twin modeling of smart buildings; the modeling terminal 101 can, but is not limited to, communicate with the real-time construction data acquisition terminal 102 and the aerial image data acquisition terminal 103 through a network, and the above network can include but is not limited to: a wired network, a wireless network, wherein the wired network includes: a local area network, a metropolitan area network, and a wide area network, and the wireless network includes: Bluetooth, WIFI (Wireless Fidelity, a technology that allows electronic devices to connect to a wireless local area network) and other networks that implement wireless communication. The above modeling terminal 101 can, but is not limited to, include a human-computer interaction screen, a processor, and a memory. The above human-computer interaction screen can, but is not limited to, be used to display images in the process of digital twin modeling of smart buildings. The above processor may be used, but is not limited to, to respond to human-computer interaction operations, execute corresponding operations, or generate corresponding instructions.

[0061] As an optional method, the real-time construction data corresponding to the target building can be collected through the real-time construction data collection terminal 102, and the aerial image data corresponding to the target building can be collected through the aerial image data collection terminal 103. The real-time construction data collection terminal 102 can include, for example, sensors, cameras, radio frequency identification tags and readers, global positioning devices, mobile devices for collecting on-site data and recording construction progress, and Internet of Things devices. The aerial image data collection terminal 103 can include, for example, cameras mounted on drones, fixed-wing aircraft, helicopters, balloons, airships, towers, and cranes.

[0062] As an optional manner, the modeling terminal 101 may also be a server, which may be a single server, a server cluster composed of multiple servers, or a cloud server. The above is only an example, and this embodiment does not make any limitation to this.

[0063] As an optional method, the following steps of the smart building digital twin modeling method can be performed on the modeling terminal 101:

[0064] Obtain real-time construction data and aerial image data corresponding to the target building;

[0065] Constructing an initial smart building digital twin model corresponding to the target building according to the real-time construction data;

[0066] According to the aerial image data and the image of the initial smart building digital twin model in the same direction, the architectural structure features of the target building are matched, and according to the matched architectural structure features, the matching degree between the aerial image data and the initial smart building digital twin model is obtained;

[0067] According to the matching degree, the initial smart building digital twin model is modified to obtain the target smart building digital twin model.

[0068] In the above method, real-time construction data provides the real-time status and detailed information of the target building, ensuring that the initial smart building digital twin model can reflect the current actual situation of the target building; aerial image data provides an external view of the target building, which can capture the details and appearance characteristics of the building structure; by matching the architectural structure characteristics of the target building and obtaining the degree of matching, it is possible to verify whether the geometric structure and appearance characteristics of the initial smart building digital twin model are consistent, and find the differences between the initial smart building digital twin model and the real building; by correcting the initial smart building digital twin model according to the degree of matching, the above differences can be eliminated and the authenticity of the digital twin model can be improved.

[0069] As an optional example, this embodiment does not limit the execution subject of the above-mentioned smart building digital twin modeling method. The above-mentioned smart building digital twin modeling method can be executed on the modeling terminal 101. For example, when the modeling terminal 101 is a desktop computer, some or all steps of the above-mentioned smart building digital twin modeling method can be executed on the desktop computer.

[0070] The above section introduces the contents of an exemplary implementation environment for applying the technical solution of the present application. Next, we will continue to introduce the smart building digital twin modeling method of the present application.

[0071] In order to solve the problem of how to improve the authenticity of the digital twin model in the prior art, the embodiments of the present application respectively propose a smart building digital twin modeling method and a smart building digital twin modeling system, and these embodiments will be described in detail below.

[0072] See also Figure 2 , Figure 2 A flow chart of a method for modeling a digital twin of a smart building provided by an embodiment of the present invention, which can be applied to Figure 1 It should be understood that the method can also be applied to other exemplary implementation environments and specifically executed by devices in other implementation environments, and this embodiment does not limit the implementation environment to which the method is applicable.

[0073] like Figure 2As shown, in an exemplary embodiment, the smart building digital twin modeling method includes at least steps S210 to S240, which are described in detail as follows:

[0074] In step S210, real-time construction data and aerial image data corresponding to the target building are obtained.

[0075] Among them, real-time construction data refers to various types of data during the construction of the target building; aerial image data refers to the image data of the target building taken from the air. For example, sensors can be used to collect environmental data and physical state data corresponding to the target building; cameras can be used to collect video image data corresponding to the target building for security monitoring and quality inspection; radio frequency identification tags and readers can be used to track and manage material information, equipment information and personnel information during the construction of the target building; global positioning equipment can be used to provide location information of materials, equipment and personnel; mobile devices can be used to collect and record data such as construction progress; Internet of Things devices can be used to achieve interconnection between devices and collect and transmit various data during construction.

[0076] In step S220, an initial smart building digital twin model corresponding to the target building is constructed based on the real-time construction data.

[0077] Among them, the initial smart building digital twin model can be, for example, a BIM model.

[0078] Exemplarily, first, a basic geometric model is created through the pre-selected BIM software. Then, based on the geometric information in the real-time construction data, a three-dimensional model of the building is created, for example: drawing the floor plan of each floor, defining the spatial layout of rooms, corridors, etc.; drawing the elevation of the building, defining the elements such as the exterior walls, windows, doors, etc.; drawing the section of the building, defining the internal structure and hierarchy. Then, the attribute data (such as materials, costs, schedules, etc.) in the real-time construction data are associated with the corresponding building structure features. Finally, by developing a data interface, the rest of the real-time construction data is automatically imported into the model.

[0079] In step S230, the architectural structure features of the target building are matched according to the image of the aerial image data and the initial smart building digital twin model in the same direction, and the matching degree between the aerial image data and the initial smart building digital twin model is obtained according to the matched architectural structure features.

[0080] Among them, building structural characteristics refer to a series of physical and geometric properties exhibited by a building during the structural design and construction process. Building structural characteristics can be, for example: geometric shape, including the overall shape of the building, floor layout, roof design, window and door layout, etc.; size and proportion, involving the size of the building, such as height, width, depth, and the proportional relationship between different parts; structural system, refers to the structural system used in the building, such as frame structure, shear wall structure, steel structure, wooden structure, etc.

[0081] Among them, the image of the aerial image data and the initial smart building digital twin model in the same direction can be the image of the aerial image data and the initial smart building digital twin model in front of the target building, the image of the aerial image data and the initial smart building digital twin model behind the target building, the image of the aerial image data and the initial smart building digital twin model on the left side of the target building, and the image of the aerial image data and the initial smart building digital twin model on the right side of the target building.

[0082] Exemplarily, the building structure features are extracted from the aerial image and the image of the initial digital twin model in the same direction. These features may include the edges, corners, textures, shapes, etc. of the building. The extracted features are matched to find the correspondence between the aerial image and the digital twin model image. When matching features, the SIFT (Scale Invariant Feature Transform) algorithm, SURF (Speeded Up Robust Features) algorithm, etc. can be used. The matching degree can be obtained by calculating the number, quality, and matching error of the matched features through the matched building structure features.

[0083] In step S240, the initial smart building digital twin model is modified according to the matching degree to obtain a target smart building digital twin model.

[0084] It can be seen from the above steps S210 to S240 that in the solution proposed in this embodiment, the real-time construction data provides the real-time status and detailed information of the target building, ensuring that the initial smart building digital twin model can reflect the current actual situation of the target building; the aerial image data provides an external view of the target building, which can capture the details and appearance features of the building structure; by matching the architectural structure features of the target building and obtaining the degree of matching, it is possible to verify whether the geometric structure and appearance features of the initial smart building digital twin model are consistent, and find the differences between the initial smart building digital twin model and the real building; by correcting the initial smart building digital twin model according to the degree of matching, the above differences can be eliminated and the authenticity of the digital twin model can be improved.

[0085] In one embodiment of the present application, the real-time construction data includes: building material consumption data, building geometry data, construction cost data, construction progress data, construction environment data, construction safety data, and construction equipment data of the target building during real-time construction.

[0086] Among them, building material consumption data is used to record the types, quantities and usage of building materials, and can be collected using sensors, tablets and other equipment; building geometry data is used to record the size, shape and spatial layout of the building, and can be collected using laser scanners, scanning equipment, etc.; construction cost data is used to record the project's budget, actual expenditure and cost changes, and can be collected using financial management systems, tablets and other equipment; construction progress data is used to record the various stages and time nodes of the project, and can be collected using project management software, tablets, sensors and other equipment; construction environment data is used to record the temperature, humidity, light and other environmental conditions at the construction site, and can be collected using environmental sensors, meteorological stations and other equipment; construction safety data is used to record safety accidents and inspection results at the construction site, and can be collected using security monitoring cameras, sensors, tablets and other equipment; construction equipment data is used to record the working status and maintenance records of mechanical equipment, and can be collected using equipment management systems, sensors, radio frequency tags and other equipment.

[0087] In one embodiment of the present application, obtaining aerial image data corresponding to a target building includes:

[0088] The aerial image data is obtained by photographing the entire target building through a plurality of pre-configured drones, wherein each of the drones flies around the target building.

[0089] For example, see Figure 3 , Figure 3 A schematic diagram of a drone provided by an embodiment of the present invention photographing a target building. The drone is used to photograph the building from four directions. Figure 3 As shown, multiple drones fly around a target building, and during the flight, they take photos of the entire target building to obtain aerial image data.

[0090] In this embodiment, each drone can shoot the target building from different heights and angles, providing a variety of perspectives. Multi-angle shooting can make up for the blind spots that may exist in a single drone shooting, ensuring that every detail of the entire building is recorded. The drone can be equipped with a high-resolution camera to capture clear and detailed images; the drone can also shoot at predetermined time intervals to record the changes of the building at different construction stages; in addition, the drone can move quickly at high altitudes, without being restricted by ground obstacles, improving collection efficiency.

[0091] In one embodiment of the present application, after obtaining the aerial image data corresponding to the target building, the method further includes:

[0092] The aerial image data is processed by a preset histogram equalization function to obtain aerial image data without illumination influence;

[0093] The aerial image data without illumination influence is processed by a preset grayscale function to obtain preprocessed aerial image data, and the preprocessed aerial image data is used to match the architectural structure features of the target building.

[0094] Among them, the histogram equalization function adjusts the histogram of the image and nonlinearly stretches the distribution interval with relatively concentrated brightness values, so that the pixel values ​​of the entire brightness range are evenly distributed. This processing can significantly enhance the contrast of the image and make the details that were originally difficult to distinguish become clearer. For images with too low or too high brightness, the histogram equalization function can make its pixel brightness histogram more evenly distributed, thereby enhancing the local contrast of the image without affecting the overall contrast.

[0095] Among them, the grayscale function is the process of converting a color image into a grayscale image. It removes the color information in the image and only retains the grayscale value. This processing can simplify the image information and reduce the complexity of subsequent processing. The grayscale processed image can more accurately match the architectural structure characteristics of the target building, improving the accuracy and reliability of the match.

[0096] For example, since the drone is affected by sunlight when taking photos, the light intensity in the aerial image data is inconsistent. Therefore, it is necessary to perform equalization processing on the image to obtain aerial image data without the influence of light, and then perform grayscale processing on the image to obtain pre-processed aerial image data. The pre-processed aerial image data is the grayscale image set J N , J N The representation can be:

[0097] J N ={J 1 ,J 2 …,J n ,…J N}

[0098] Among them, J N represents a grayscale image set; J 1 Represents the grayscale image corresponding to the first aerial image collected by the drone; J 2 represents the grayscale image corresponding to the second aerial image collected by the UAV; J n represents the grayscale image corresponding to the nth aerial image collected by the UAV; J NRepresents the grayscale image corresponding to the Nth aerial image collected by the drone; N represents the total number of aerial images collected by the drone.

[0099] In one embodiment of the present application, the matching of the building structure features of the target building according to the image of the aerial image data and the initial smart building digital twin model in the same direction, and obtaining the matching degree between the aerial image data and the initial smart building digital twin model according to the matched building structure features, includes:

[0100] Perform regularity performance analysis on similar building structure features in the aerial image data and the image of the initial smart building digital twin model in the same direction, and divide the aerial image data into multiple regions according to the regularity performance analysis results, and divide the image of the initial smart building digital twin model into multiple regions according to the regularity performance analysis results;

[0101] Matching the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model;

[0102] According to the matching accuracy, the matching degree between the aerial image data and the initial smart building digital twin model is obtained.

[0103] For example, when performing regularity analysis on similar building structural features in the same direction of the aerial image data and the initial smart building digital twin model, a multi-scale feature extraction method can be used to extract feature points in the image, and then the distribution pattern of the feature points can be analyzed to identify the geometric shape, structural type and other features of the building. Through regularity analysis, key structural features of the building, such as columns, beams, windows, doors, etc., can be identified; regularity analysis helps reduce mismatches and improve matching accuracy.

[0104] For example, based on the results of regularity performance analysis, the aerial image data and the image of the initial smart building digital twin model are divided into multiple regions, each region can be labeled to mark different structural features, and the matching granularity can be refined to improve the matching accuracy.

[0105] For example, when matching the building structure features in the aerial image data and the initial smart building digital twin model in the same direction, the feature points in each area can be matched according to the divided areas. Through regional matching, feature points can be matched more accurately and the error of global matching can be reduced. The matching of local areas can better handle the local structural features of the building and improve the robustness of the matching.

[0106] For example, when calculating the degree of matching between the aerial image data and the initial smart building digital twin model based on the matching accuracy, the number of matching points can be counted to determine the density of the matching points; the proportion of matching points to the total feature points can be calculated to evaluate the overall quality of the match; the geometric error of the matching points can be calculated to evaluate the positional deviation of the matching points; and the distribution of errors can be analyzed to identify areas with larger errors.

[0107] In one embodiment of the present application, the regularity performance analysis is performed on the similar building structure features in the images of the aerial image data and the initial smart building digital twin model in the same direction, and the aerial image data is divided into multiple regions according to the regularity performance analysis results, and the image of the initial smart building digital twin model is divided into multiple regions according to the regularity performance analysis results, including:

[0108] Acquire a first edge detection image and a first threshold segmentation image corresponding to the aerial image data, and a second edge detection image and a second threshold segmentation image corresponding to the image of the initial smart building digital twin model;

[0109] Dividing the data sequence corresponding to similar building structure features in the first edge detection image into a plurality of first feature groups, and performing regularity performance analysis on each of the first feature groups to obtain a first regularity performance degree corresponding to the first edge detection image;

[0110] Dividing the data sequence corresponding to similar building structure features in the first threshold segmentation image into a plurality of second feature groups, and performing regularity performance analysis on each of the second feature groups to obtain a second regularity performance degree corresponding to the first threshold segmentation image;

[0111] The image corresponding to the regularity expression degree closest to zero among the first regularity expression degree and the second regularity expression degree is used as a target aerial image, and similar building structure features in the target aerial image are divided into a plurality of first main areas, and each of the first main areas is divided into a plurality of first sub-areas;

[0112] Dividing the data sequence corresponding to the similar building structure features in the second edge detection image into a plurality of third feature groups, and performing regularity performance analysis on each of the third feature groups to obtain a third regularity performance degree corresponding to the second edge detection image;

[0113] Dividing the data sequence corresponding to similar building structure features in the second threshold segmentation image into a plurality of fourth feature groups, and performing regularity performance analysis on each of the fourth feature groups to obtain a fourth regularity performance degree corresponding to the second threshold segmentation image;

[0114] The image corresponding to the regularity expression degree closest to zero among the third regularity expression degree and the fourth regularity expression degree is taken as the target model image, and similar architectural structure features in the target model image are divided into multiple second main areas, and each of the second main areas is divided into multiple second sub-areas.

[0115] Among them, since the buildings in the aerial image data taken by drones have great structural similarities, when the aerial image data is directly used to match the building structure features with the images collected at the same angle by the initial smart building digital twin model, the matching results may have large errors. Therefore, it is necessary to analyze the regularity of similar building structure features to make the matching results more accurate.

[0116] Among them, through the aerial image data, it can be seen that the buildings have extremely high structural similarities, and there are obvious dividing lines at the edges of similar building structural features. If the image matching analysis is performed directly using corner point detection, there may be errors in the matching because many corner points are in similar structural areas. Therefore, in this embodiment, threshold segmentation and edge detection are used to process the aerial image data and the images of the initial smart building digital twin model in the same direction.

[0117] Among them, edge detection is to use a specific edge detection algorithm to identify places in the image where the pixel intensity changes significantly. These places usually correspond to the boundaries of objects. Edge detection algorithms can include: Sobel operator, which is used to detect edges by calculating the size and direction of the image gradient; Canny edge detection, a multi-level edge detection algorithm, which first uses a Gaussian filter to smooth the image, then calculates the gradient amplitude and direction, then performs non-maximum suppression, and finally determines the final edge through the double threshold method; Laplacian operator, which detects edges by performing secondary derivative operations on the image, is suitable for detecting small edges. When using different edge detection algorithms, different initial parameters need to be set, such as high and low thresholds, or the size of the filter. After edge detection is completed, post-processing is sometimes required. For example, edge detection will disconnect continuous edges, and these edges can be connected through morphological operations (such as dilation); length thresholds or connected region analysis methods can also be used to remove small fragments or isolated points that are unlikely to be actual edges. Through the above edge detection algorithms and post-processing methods, the images of the aerial image data and the initial smart building digital twin model in the same direction are processed to obtain edge detection images.

[0118] Among them, threshold segmentation is to divide the image into foreground and background according to the distribution of threshold and pixel value. The foreground part includes pixels with pixel values ​​greater than or equal to the threshold, and the background part includes pixels with pixel values ​​less than the threshold. When performing threshold segmentation, it is necessary to first determine the threshold, which can be a global threshold or a local threshold. Among them, when determining the global threshold, a fixed threshold can be selected to classify all pixels with pixel values ​​greater than the threshold as foreground, and pixels less than the threshold as background. Among them, when determining the local threshold, different thresholds can be selected according to different areas of the image, which is suitable for situations where the brightness of different areas in the image is greatly different. After determining the threshold, each pixel in the image is divided into foreground or background according to the threshold, that is, the threshold segmentation of the image is completed. Through the above threshold segmentation method, the aerial image data and the image of the initial smart building digital twin model in the same direction are processed to obtain a threshold segmented image.

[0119] For example, after obtaining the first edge detection image and the first threshold segmentation image corresponding to the aerial image data, since the main area of ​​the building has a large structural similarity from top to bottom, a vertical straight line is constructed from top to bottom at the same position using the structural characteristics, and the structural performance of the two images in the direction of the vertical straight line can be obtained. Figure 4 , Figure 4 A schematic diagram of a first edge detection image provided by an embodiment of the present invention, comprising Figure 4 It can be seen that for edge detection images, white pixels represent the boundaries of buildings; see Figure 5 , Figure 5 A schematic diagram of a first threshold segmentation image provided by an embodiment of the present invention, comprising Figure 5 It can be seen that for the threshold segmentation image, its black pixel points represent the boundaries of the building. Analyze from top to bottom along the constructed vertical line, count the continuous black pixel blocks distributed on the edge detection image each time, and construct the first data sequence

[0120] {n a,1 ,n a,2 …n a,i}

[0121] Among them, n a,1 is the first black pixel block on the edge detection image; n a,2 is the second black pixel block on the edge detection image; n a,i is the i-th black pixel block on the edge detection image; i is the number of black pixel blocks.

[0122] Analyze from top to bottom along the constructed vertical line, count the continuous white pixel blocks distributed on the threshold segmentation image each time, and construct the second data sequence

[0123] {n b,1 ,n b,2 …n b,i}

[0124] Among them, n b,1 is the first white pixel block on the threshold segmentation image; n b,2 is the second white pixel block on the threshold segmentation image; n b,i is the i-th white pixel block on the threshold segmentation image; i is the number of white pixel blocks.

[0125] Since buildings show great structural repetitiveness, the first data sequence and the second data sequence should have good regularity. The regularity is that in the data sequence, repeated data will appear every x data, so the x data are divided into a group to obtain multiple first feature groups and multiple second feature groups.

[0126] For example, the regularity between the feature groups is analyzed, that is, the normal regularity should be expressed as the same data at the corresponding positions in each group. The degree of regularity can be expressed as:

[0127]

[0128] Among them, τ is the degree of regularity; d k,i Represents the data corresponding to the i-th position in the k-th feature group in the data sequence; d k-1,iIt represents the data corresponding to the i-th position in the k-1-th feature group in the data sequence; K represents the number of feature groups in the data sequence; x is the number of data in the feature group, which is also the total number of positions corresponding to the data in the feature group.

[0129] Among them, |d k,i -d k-1,i | is the difference between two data at corresponding positions in adjacent groups. The difference can be the difference corresponding to the grayscale value or the difference corresponding to the geometric feature (such as shape). The larger the difference is, the more different the two data at corresponding positions in the adjacent groups are. For normal regularity, the difference between two data at corresponding positions in adjacent groups should be 0.

[0130] Among them, the closer τ is to 0, the better the regularity of the image is expressed.

[0131] For example, in the data sequence of the target aerial image and the target model image, every x data is divided into a group, and the data show good regularity. Figure 6 , Figure 6 A schematic diagram of area division provided by an embodiment of the present invention, comprising Figure 6 As shown, similar architectural structure features in the target aerial image are divided into multiple first main areas on the image every x data values ​​and the corresponding number of pixel blocks; and the data in each first main area is the same, and then each first main area is divided into multiple first sub-areas. For example, the first main area can be divided into multiple first sub-areas according to the number, position and shape of multiple architectural structure features contained in the first main area.

[0132] In this embodiment, by acquiring the first edge detection image, the first threshold segmentation image, the second edge detection image, and the second threshold segmentation image, the architectural structure features in the image can be highlighted, redundant information can be removed, and a basis can be provided for subsequent regularity performance analysis and regional division. Dividing the data sequence corresponding to similar architectural structure features in the edge detection image and the threshold segmentation image into multiple feature groups helps to classify similar features, which is convenient for subsequent analysis and processing. By performing regularity performance analysis on each feature group, the regularity performance degree of the architectural structure features in the image is obtained, and the distribution law of the architectural structure features in the image can be revealed. Selecting the image corresponding to the regularity performance degree closest to zero as the target image can screen out the image with the most regular architectural structure features and easy to analyze. Dividing similar architectural structure features in the target aerial image and the target model image into multiple main areas and sub-areas can divide the image into more detailed and manageable parts.

[0133] In one embodiment of the present application, the matching of the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model, includes:

[0134] Matching the building structure features in the target aerial image and the target model image to obtain a plurality of matching feature groups;

[0135] Determine the position difference of each of the matching feature groups according to the first main area and the first sub-area corresponding to the building structure feature in each of the matching feature groups, and the second main area and the second sub-area corresponding to the building structure feature;

[0136] The matching accuracy corresponding to each matching feature group is determined by combining the position difference of each matching feature group and the corresponding regularity expression degree of the target aerial image and the target model image.

[0137] Among them, since buildings have great structural similarities, the results of building structure feature matching may have large errors. Therefore, it is necessary to analyze the matching feature group to determine the accuracy of the matching results.

[0138] Among them, matching the building structure features in the target aerial image and the target model image usually includes the steps of feature extraction, feature descriptor generation, and feature matching. Feature extraction is the process of extracting building structure features from an image. These building structure features can be called key points. Commonly used feature extraction methods include SIFT (Scale Invariant Feature Transform) and SURF (Speeded Up Robust Features). Feature descriptors are detailed descriptions of the area around the key points, which are used for subsequent feature matching. Feature matching is to match the key points of the aerial image with the key points of the model image based on the feature descriptors to find the most similar key point pairs. BFMatcher (brute force matching) and FLANN (Fast Nearest Neighbor Search Library) can be used for feature matching. After feature matching, multiple matching feature groups can be obtained, each of which includes a building structure feature in a target aerial image and a building structure feature in a target model image. The geometric features of these two building structure features are the most similar. Exemplarily, the position difference of the matching feature groups can be expressed as:

[0139]

[0140] Among them, γ is the position difference of the matching feature group; M vIndicates the serial number of the first main area where the building structure feature in the vth matching feature group is located; Indicates the serial number of the second main area where the building structure feature in the vth matching feature group is located; J v Indicates the serial number of the first sub-region where the building structure feature in the vth matching feature group is located; Indicates the sequence number of the second sub-region where the building structure feature in the vth matching feature group is located.

[0141] in, It is the position difference of the building structure features in the same matching feature group in the main area where the target aerial image and the target model image are located. The larger the position difference is, the greater the position difference of the matching feature group is, and the lower the matching accuracy is.

[0142] in, It is the position difference of the building structure features in the same matching feature group in the sub-area where the target aerial image and the target model image are located. The larger the position difference is, the greater the position difference of the matching feature group is, and the lower the matching accuracy is.

[0143] Among them, if the building structure features in the target aerial image and the target model image are in the main area and sub-area at similar positions, the smaller γ is, the higher the accuracy of the building structure feature matching is.

[0144] Exemplarily, the matching accuracy corresponding to the matching feature group may be expressed as follows:

[0145]

[0146] Among them, FM v is the matching accuracy corresponding to the vth matching feature group; τ 1 Indicates the regularity of the target aerial image; τ 2 Indicates the regularity of the target model image; G v Represents the average gray value of the building structure feature in the target aerial image and its surrounding 8 neighborhood pixels in the vth matching feature group; It represents the average gray value of the building structure feature in the target model image in the vth matching feature group and the pixels in its 8 neighborhoods; e represents the base of the natural logarithm; norm represents the norm in mathematical calculation, which is used to measure the size or length of a vector or matrix.

[0147] Among them, e -γ The larger the value, the lower the accuracy of building structure feature matching; -γ The smaller it is, the higher the accuracy of building structure feature matching is.

[0148] Among them, |τ 1 -τ 2| represents the difference in the regularity between the target aerial image and the target model image. The larger the difference, the lower the accuracy of the building structure feature matching; the smaller the difference, the higher the accuracy of the building structure feature matching.

[0149] in, It represents the difference between the target model image and the average grayscale value of the pixels in the target model image and its 8-neighborhood neighborhood. The larger the difference, the lower the accuracy of the building structure feature matching; the smaller the difference, the higher the accuracy of the building structure feature matching.

[0150] Among them, when the building structure features are in the same position in the target aerial image and the target model image, the grayscale difference of the surrounding pixels is small, and the difference in the degree of regularity is small, the greater the matching accuracy, the greater the accuracy of the correct matching of the building structure features.

[0151] In one embodiment of the present application, obtaining the degree of matching between the aerial image data and the initial smart building digital twin model according to the matching accuracy includes:

[0152] The matching feature groups whose matching accuracy is greater than or equal to a preset matching accuracy threshold are taken as target matching feature groups, and the number of the target matching feature groups is obtained;

[0153] Comparing the quantity of the target matching feature group with the quantity of all the matching feature groups to obtain a quantity comparison result;

[0154] The matching degree is determined by combining the quantitative comparison result with the average of the matching accuracy corresponding to all the matching feature groups.

[0155] Exemplarily, the preset matching accuracy threshold T 1 =0.9, when the matching accuracy is greater than or equal to T 1 , the matching result of the matching feature group is considered normal.

[0156] Exemplarily, the matching degree may be expressed as:

[0157]

[0158] Among them, σ is the matching degree; n is the number of target matching feature groups; N is the number of all matching feature groups.

[0159] in, is the proportion of the target matching feature group. The larger the proportion, the higher the matching degree.

[0160] in, It is the mean of the matching accuracy corresponding to all matching feature groups. The larger the mean, the higher the matching degree.

[0161] Among them, when σ is larger, the building structure features in the target aerial image and the target model image are more matched, and the recognition degree of the target aerial image to the target model image is higher.

[0162] In one embodiment of the present application, the initial smart building digital twin model is modified according to the matching degree to obtain the target smart building digital twin model, including:

[0163] comparing the matching degree with a preset matching degree threshold;

[0164] When the matching degree is less than or equal to the preset matching degree threshold, correcting the real-time construction data according to the initial smart building digital twin model;

[0165] Based on the corrected real-time construction data, a digital twin model of the target smart building is constructed.

[0166] Exemplarily, the preset matching degree threshold is T 2 , T 2 =0.8, when the matching degree is greater than T 2 When the matching degree is less than or equal to T 2 It means that the initial smart building digital twin model is very different from the actual building. It is necessary to analyze the source of data differences based on the initial smart building digital twin model, correct the difference data in time, and build the target smart building digital twin model based on the corrected real-time construction data, so that the target smart building digital twin model is closer to the real building performance.

[0167] Exemplarily, the latest real-time construction data of the target building can be collected, and then the latest real-time construction data can be compared and analyzed with the initial smart building digital twin model to identify the source of the difference between the latest real-time construction data and the initial smart building digital twin model. The sources of these differences can be, for example: construction errors, which are errors that may exist during the construction process, resulting in deviations between the model and the actual building; material property differences, the materials actually used may be different from those in the model; environmental factors, changes in environmental conditions can also affect the real-time state of the building, such as foundation settlement. According to the results of the difference analysis, the initial smart building digital twin model is updated to ensure that it can accurately reflect the actual situation. The model update work can include: geometric model correction, adjusting the geometric shape and size of the building to make it closer to reality; material property update, if it is found that there are differences in material properties, it is necessary to update the properties of these materials in the model; functional simulation adjustment, according to actual use, adjust the simulation settings of the internal systems of the building, such as air conditioning systems, lighting systems, etc.; environmental condition correction, considering the latest environmental data, adjust the relevant environmental parameters in the model.

[0168] Figure 7 A schematic diagram of a smart building digital twin modeling system provided by an embodiment of the present invention. The system can be applied to Figure 1 The system can also be applied to other exemplary implementation environments and specifically configured in other devices. This embodiment does not limit the implementation environment to which the system is applied.

[0169] like Figure 7 As shown, the exemplary smart building digital twin modeling system includes:

[0170] The data acquisition module 701 is used to acquire the real-time construction data and aerial image data corresponding to the target building;

[0171] An initial model building module 702 is used to build an initial smart building digital twin model corresponding to the target building according to the real-time construction data;

[0172] A matching module 703 is used to match the architectural structure features of the target building according to the aerial image data and the image of the initial smart building digital twin model in the same direction, and obtain the matching degree between the aerial image data and the initial smart building digital twin model according to the matched architectural structure features;

[0173] The correction module 704 is used to correct the initial smart building digital twin model according to the matching degree to obtain the target smart building digital twin model.

[0174] In this exemplary smart building digital twin modeling system, real-time construction data provides the real-time status and detailed information of the target building, ensuring that the initial smart building digital twin model can reflect the current actual situation of the target building; aerial image data provides an external view of the target building, which can capture the details and appearance features of the building structure; by matching the architectural structure features of the target building and obtaining the degree of matching, it is possible to verify whether the geometric structure and appearance features of the initial smart building digital twin model are consistent, and to find the differences between the initial smart building digital twin model and the real building; by correcting the initial smart building digital twin model according to the degree of matching, the above differences can be eliminated, thereby improving the authenticity of the digital twin model.

[0175] It should be noted that the smart building digital twin modeling system provided in the above embodiment and the smart building digital twin modeling method provided in the above embodiment belong to the same concept, and the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment and will not be repeated here. In actual applications, the smart building digital twin modeling system provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the system into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0176] It should be noted that the sequence of the above embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0177] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for modeling a digital twin of a smart building, characterized in that: The method comprises: Obtain real-time construction data and aerial image data corresponding to the target building; Constructing an initial smart building digital twin model corresponding to the target building according to the real-time construction data; Acquire a first edge detection image and a first threshold segmentation image corresponding to the aerial image data, and a second edge detection image and a second threshold segmentation image corresponding to the image of the initial smart building digital twin model; Dividing the data sequence corresponding to similar building structure features in the first edge detection image into a plurality of first feature groups, and performing regularity performance analysis on each of the first feature groups to obtain a first regularity performance degree corresponding to the first edge detection image; Dividing the data sequence corresponding to similar building structure features in the first threshold segmentation image into a plurality of second feature groups, and performing regularity performance analysis on each of the second feature groups to obtain a second regularity performance degree corresponding to the first threshold segmentation image; The image corresponding to the regularity expression degree closest to zero among the first regularity expression degree and the second regularity expression degree is used as a target aerial image, and similar building structure features in the target aerial image are divided into a plurality of first main areas, and each of the first main areas is divided into a plurality of first sub-areas; Dividing the data sequence corresponding to the similar building structure features in the second edge detection image into a plurality of third feature groups, and performing regularity performance analysis on each of the third feature groups to obtain a third regularity performance degree corresponding to the second edge detection image; Dividing the data sequence corresponding to similar building structure features in the second threshold segmentation image into a plurality of fourth feature groups, and performing regularity performance analysis on each of the fourth feature groups to obtain a fourth regularity performance degree corresponding to the second threshold segmentation image; The image corresponding to the regularity expression degree closest to zero among the third regularity expression degree and the fourth regularity expression degree is used as the target model image, and similar building structure features in the target model image are divided into a plurality of second main areas, and each of the second main areas is divided into a plurality of second sub-areas; Matching the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model; According to the matching accuracy, the matching degree between the aerial image data and the initial smart building digital twin model is obtained; According to the matching degree, the initial smart building digital twin model is modified to obtain the target smart building digital twin model.

2. The smart building digital twin modeling method according to claim 1, characterized in that: The real-time construction data includes: building material consumption data, building geometry data, construction cost data, construction progress data, construction environment data, construction safety data, and construction equipment data of the target building during the real-time construction process.

3. The smart building digital twin modeling method according to claim 1, characterized in that: Obtain aerial image data corresponding to the target building, including: The aerial image data is obtained by photographing the entire target building through a plurality of pre-configured drones, wherein each of the drones flies around the target building.

4. The smart building digital twin modeling method according to claim 1, characterized in that: After obtaining the aerial image data corresponding to the target building, it also includes: The aerial image data is processed by a preset histogram equalization function to obtain aerial image data without illumination influence; The aerial image data without illumination influence is processed by a preset grayscale function to obtain preprocessed aerial image data, and the preprocessed aerial image data is used to match the architectural structure features of the target building.

5. The smart building digital twin modeling method according to claim 1, characterized in that: The matching of the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determining the matching accuracy of the building structure features after matching according to multiple areas corresponding to the aerial image data and multiple areas corresponding to the image of the initial smart building digital twin model, includes: Matching the building structure features in the target aerial image and the target model image to obtain a plurality of matching feature groups; Determine the position difference of each of the matching feature groups according to the first main area and the first sub-area corresponding to the building structure feature in each of the matching feature groups, and the second main area and the second sub-area corresponding to the building structure feature; The matching accuracy corresponding to each matching feature group is determined by combining the position difference of each matching feature group and the corresponding regularity expression degree of the target aerial image and the target model image.

6. The smart building digital twin modeling method according to claim 5, characterized in that: The obtaining, according to the matching accuracy, a matching degree between the aerial image data and the initial smart building digital twin model includes: The matching feature groups whose matching accuracy is greater than or equal to a preset matching accuracy threshold are taken as target matching feature groups, and the number of the target matching feature groups is obtained; Comparing the quantity of the target matching feature group with the quantity of all the matching feature groups to obtain a quantity comparison result; The matching degree is determined by combining the quantitative comparison result with the average of the matching accuracy corresponding to all the matching feature groups.

7. The smart building digital twin modeling method according to claim 1, characterized in that: The initial smart building digital twin model is modified according to the matching degree to obtain a target smart building digital twin model, including: comparing the matching degree with a preset matching degree threshold; When the matching degree is less than or equal to the preset matching degree threshold, correcting the real-time construction data according to the initial smart building digital twin model; Based on the corrected real-time construction data, a digital twin model of the target smart building is constructed.

8. A smart building digital twin modeling system, characterized in that: The system comprises: A data acquisition module is used to acquire real-time construction data and aerial image data corresponding to the target building; An initial model building module, used to build an initial smart building digital twin model corresponding to the target building according to the real-time construction data; A matching module is used to obtain a first edge detection image and a first threshold segmentation image corresponding to the aerial image data, and a second edge detection image and a second threshold segmentation image corresponding to the image of the initial smart building digital twin model; divide the data sequence corresponding to similar architectural structure features in the first edge detection image into multiple first feature groups, and perform regularity performance analysis on each of the first feature groups to obtain a first regularity performance degree corresponding to the first edge detection image; divide the data sequence corresponding to similar architectural structure features in the first threshold segmentation image into multiple second feature groups, and perform regularity performance analysis on each of the second feature groups to obtain a second regularity performance degree corresponding to the first threshold segmentation image; take the image corresponding to the regularity performance degree closest to zero among the first regularity performance degree and the second regularity performance degree as the target aerial image, and divide the similar architectural structure features in the target aerial image into multiple first main areas, and divide each of the first main areas into multiple first sub-areas; divide the data sequence corresponding to similar architectural structure features in the second edge detection image into multiple third feature groups. group, and perform regularity expression analysis on each of the third feature groups to obtain the third regularity expression degree corresponding to the second edge detection image; divide the data sequence corresponding to the similar building structure features in the second threshold segmentation image into multiple fourth feature groups, and perform regularity expression analysis on each of the fourth feature groups to obtain the fourth regularity expression degree corresponding to the second threshold segmentation image; take the image corresponding to the regularity expression degree closest to zero among the third regularity expression degree and the fourth regularity expression degree as the target model image, and divide the similar building structure features in the target model image into multiple second main areas, and divide each of the second main areas into multiple second sub-areas; match the building structure features in the image of the aerial image data and the initial smart building digital twin model in the same direction, and determine the matching accuracy of the building structure features after matching according to the multiple areas corresponding to the aerial image data and the multiple areas corresponding to the image of the initial smart building digital twin model; according to the matching accuracy, obtain the matching degree of the aerial image data with the initial smart building digital twin model; The correction module is used to correct the initial smart building digital twin model according to the matching degree to obtain the target smart building digital twin model.

Images