A method and system for constructing an urban building model based on climate environment simulation

Abstract

The present application provides a method and system for constructing an urban building model based on climate environment simulation, which relates to the field of information processing technology. The building image distribution domain of the urban building model in climate environment simulation is determined according to building point-of-interest data; the modified characteristic attributes of urban buildings under different climate environment categories are determined, and then the building form indicators for constructing the urban building model are determined through the modified characteristic attributes; the building attribute label image of the target urban building is added to the building category model to obtain the gradient influence coefficient of the target urban building image in different building areas, and then the building form interest points of the urban building model are determined; the urban building model is classified and constructed according to the building form indicators and the building form interest points, and batch automatic conversion is performed. The present application can construct a building model under the influence of diverse urban climate environments to improve the accuracy of constructing an urban building model based on climate simulation.

Description

Technical Field

[0001] This application relates to the field of information processing technologies. More specifically, this application relates to a method and system for constructing an urban building model based on climate environment simulation. Background Art

[0002] During the urban development process, the physical spatial form of a city has an important impact on the transfer, storage, and release of matter and energy in the environment. In particular, urban buildings not only directly act on the wind field and radiation field but also significantly change the distribution and variation of climate environment parameters. For example, the height, density, form, and material combination of the building facades all affect climate factors such as wind speed, heat radiation, and temperature distribution in the city, thus having a profound impact on the urban climate. Specifically, high-rise buildings may block the flow of wind and generate the heat island effect, while low-density residential areas may form a more uniform temperature distribution. However, with the acceleration of the urbanization process, the spatial form, functional zoning, and material combination of urban buildings have become increasingly complex, which poses a huge challenge to urban climate environment modeling. Traditional urban climate simulation methods usually have difficulty accurately reflecting the impact of buildings on the local climate, especially the combined effects of complex factors such as building form, functional zoning, and material selection. Due to the wide variety of building types, complex scales, and interlaced structures, how to effectively incorporate these building characteristics into climate environment simulation has become a key issue for accurately simulating and predicting urban climate change.

[0003] The existing methods for constructing urban building models based on climate environment simulation are mainly divided into mesoscale simulation and microscale simulation. Mesoscale simulation usually relies on land use and land cover data and uses the canopy model (LCZ) to analyze regional environmental parameters, with a resolution generally not higher than 1 km. The LCZ method simplifies the description of the building area by classifying the ground environment. Although it refines building characteristics to a certain extent, due to the use of fixed-type classification, it is difficult to fully reflect the complexity of the urban built environment and has problems of abstraction and insufficient generalization. In contrast, microscale simulation uses a higher spatial resolution and can reveal in detail the impact of complex urban underlying surfaces such as buildings, roads, and vegetation on the local climate environment. Microscale simulation relies on high-precision urban building geometric models that can accurately describe the geometric form, function, and location of buildings and directly apply these building characteristics as input parameters to numerical calculations, thus providing more accurate climate simulation results. However, the existing methods for constructing urban building models based on climate environment simulation have problems such as inaccurate acquisition of building attribute data and excessive simplification of building forms, making it difficult for the model to comprehensively capture the complex spatial forms and functional differences of buildings. As a result, in diverse urban climate environments, the impact of buildings on climate change cannot be accurately reflected, reducing the accuracy of building model construction in climate simulation. Therefore, how to construct building models under the influence of diverse urban climate environments to improve the accuracy of urban building model construction based on climate simulation is a problem faced by the industry. Summary of the Invention

[0004] This application provides a method and system for constructing an urban building model based on climate environment simulation, which can construct a building model under the influence of diverse urban climate environments to improve the accuracy of urban building model construction based on climate simulation.

[0005] In a first aspect, this application provides a method for constructing an urban building model based on climate environment simulation. The model construction method includes the following steps:

[0006] Obtain a set of input images of the target urban building from a climate environment simulation platform;

[0007] Perform hierarchical analysis on the set of input images to obtain building interest point data of the target urban building in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data;

[0008] Obtain the building geometric feature images of the target urban building complex in climate environment simulation, fill in the missing parts of the building geometric feature images to obtain the corrected characteristic attributes of urban buildings under different climate environment categories, and further determine the building form indicators when constructing the urban building model through the corrected characteristic attributes;

[0009] Determine the building category model corresponding to the target urban building complex in different climate environments, add the building attribute label images of the target urban buildings to the building category model, obtain the gradient influence coefficients of the target urban building images in different building areas, and then determine the building form interest points of the urban building model from the gradient influence coefficients and the building image distribution domain;

[0010] Classify and construct the urban building model according to the building form index and the building form interest points, and perform batch automatic conversion.

[0011] In this embodiment, performing hierarchical analysis on the input image set to obtain the building interest point data of the target urban buildings in different building areas specifically includes:

[0012] Extract features from multiple images in the input image set to obtain corresponding multiple feature images;

[0013] Perform class division according to all the feature images to obtain a class mapping table corresponding to each class of feature images;

[0014] Project the class mapping table into the input image set for screening to obtain the building interest point data of the target urban buildings in different building areas.

[0015] In this embodiment, determining the building image distribution domain of the urban building model in the climate environment simulation according to the building interest point data specifically includes:

[0016] Convert the building interest point data into building element points of the urban building model in the climate environment simulation;

[0017] Determine the building attribute elements of the urban building model in the climate environment simulation according to the building element points;

[0018] Determine the building image distribution domain of the urban building model in the climate environment simulation according to the building attribute elements.

[0019] In this embodiment, determining the building form index when constructing the urban building model through the corrected feature attributes specifically includes:

[0020] Determine the building category frequency density corresponding to the interest point data in each building according to the corrected feature attributes;

[0021] Determine the building type ratio corresponding to the interest point data in each building;

[0022] Determine the building form index when constructing the urban building model according to the building category frequency density and the building type ratio.

[0023] In this embodiment, determining the building category model corresponding to the target urban building complex in different climate environments specifically includes:

[0024] Determining the building attribute frequency density corresponding to the target urban building complex in different climate environments;

[0025] Determining the characteristic graphic description of the target urban building complex in different climate environments according to the building attribute frequency density;

[0026] Constructing the building category model corresponding to the target urban building complex in different climate environments according to the characteristic graphic description.

[0027] In this embodiment, determining the building form interest points of the urban building model from the gradient influence coefficient and the building image distribution domain specifically includes:

[0028] Determining the building climate zoning of the building according to the building image distribution domain;

[0029] Determining the building form interest points of the urban building model through the building climate zoning and the gradient influence coefficient.

[0030] In this embodiment, the gradient influence coefficient represents the degree of influence of a building on the climate environment in different building areas.

[0031] In this embodiment, the building geometric feature image refers to an image showing the geometric form and structural features of a building.

[0032] In this embodiment, the input image set refers to a high-resolution image data set of the target urban building in climate environment simulation.

[0033] In a second aspect, the present application provides a system for constructing an urban building model based on climate environment simulation, which is used to execute a method for constructing an urban building model based on climate environment simulation. The model construction system includes:

[0034] An image set acquisition module, configured to acquire an input image set of the target urban building from a climate environment simulation platform;

[0035] A building hierarchy analysis module, configured to perform a hierarchy analysis on the input image set to obtain building interest point data of the target urban building in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data;

[0036] A feature image filling module, which is used to obtain the building geometric feature image of the target urban building complex in climate environment simulation, fill in the missing parts of the building geometric feature image, obtain the corrected feature attributes of urban buildings under different climate environment categories, and then determine the building form indexes for building urban building models through the corrected feature attributes;

[0037] A building shape determination module, which is used to determine the building category models corresponding to the target urban building complex in different climate environments, add the building attribute label image of the target urban building to the building category models, obtain the gradient influence coefficients of the target urban building image in different building areas, and then determine the building shape interest points of the urban building model from the gradient influence coefficients and the building image distribution domain;

[0038] A building model construction module, which is used to classify and construct the urban building model according to the building form indexes and the building shape interest points, and perform batch automatic conversion.

[0039] The technical solutions provided by the disclosed embodiments of the present application have the following beneficial effects:

[0040] Obtain the input image set of the target urban building from the climate environment simulation platform; perform hierarchical analysis on the input image set to obtain the building interest point data of the target urban building in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data; obtain the building geometric feature image of the target urban building complex in climate environment simulation, fill in the missing parts of the building geometric feature image, obtain the corrected feature attributes of urban buildings under different climate environment categories, and then determine the building form indexes for building urban building models through the corrected feature attributes; determine the building category models corresponding to the target urban building complex in different climate environments, add the building attribute label image of the target urban building to the building category models, obtain the gradient influence coefficients of the target urban building image in different building areas, and then determine the building shape interest points of the urban building model from the gradient influence coefficients and the building image distribution domain; classify and construct the urban building model according to the building form indexes and the building shape interest points, and perform batch automatic conversion.

[0041] It can be seen that in the present application, it can effectively make up for the lack and incompleteness of model data attributes caused by excessive simplification of building forms. Among them, by obtaining high-quality building image data from the climate environment simulation platform, the accuracy and integrity of the building model can be ensured, providing a consistent and accurate data basis for subsequent building analysis and climate simulation, thereby improving the reliability of the simulation results. Hierarchical analysis can effectively extract the key features of the building, accurately identify the functional differences of different building areas, provide structured data for subsequent building attribute modeling and climate environment simulation, and thus improve the refinement degree of the model and the accuracy of the simulation. By complementing the missing building geometric feature images and correcting the building feature data, the integrity and accuracy of the building model can be ensured, thereby improving the adaptability of the building to different climate environments and enhancing the accuracy and practicality of climate simulation. By combining the building category model and the gradient influence coefficient, the influence of the building on the surrounding environment (such as wind field, thermal radiation, etc.) can be accurately described, making the building performance in climate simulation more in line with the actual situation, thereby improving the environmental adaptability and accuracy of the simulation. By automatically classifying and batch-converting building models, the efficiency of building data processing can be greatly improved, manual intervention can be reduced, and a large amount of building data can be ensured to efficiently and consistently adapt to the climate environment simulation system, significantly enhancing the intelligence and accuracy of urban building analysis.

[0042] In summary, the technical solution adopted in the present application can construct a building model under the influence of diverse urban climate environments to improve the accuracy of constructing a city building model based on climate simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0044] Figure 1 is a flowchart of a method for constructing a city building model based on climate environment simulation provided by the present application;

[0045] Figure 2 is a schematic flowchart of determining the corrected feature attributes provided by the present application;

[0046] Figure 3 is a schematic flowchart of determining the gradient influence coefficient provided by the present application;

[0047] Figure 4 is a module structure diagram of a system for constructing a city building model based on climate environment simulation provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] Next, in combination with the accompanying drawings in the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0049] The embodiments of the present application provide a method and system for constructing an urban building model based on climate environment simulation. The core is to obtain a set of input images of the target urban building from a climate environment simulation platform; perform hierarchical analysis on the set of input images to obtain building interest point data of the target urban building in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data; obtain the building geometric feature images of the target urban building complex in climate environment simulation, fill in the missing parts of the building geometric feature images to obtain the corrected feature attributes of urban buildings under different climate environment categories, and then determine the building form indicators when constructing the urban building model through the corrected feature attributes; determine the building category models corresponding to the target urban building complex in different climate environments, add the building attribute label images of the target urban building to the building category models to obtain the gradient influence coefficients of the target urban building images in different building areas, and then determine the building form interest points of the urban building model from the gradient influence coefficients and the building image distribution domain; classify and construct the urban building model according to the building form indicators and the building form interest points, and perform batch automatic conversion. By adopting the above solution, building model construction can be carried out under the influence of diverse urban climate environments to improve the accuracy of constructing urban building models based on climate simulation.

[0050] To better understand the above technical solutions, the above technical solutions will be described in detail below in combination with the accompanying drawings of the specification and specific implementation manners. Refer to Figure 1 As shown, this figure is an exemplary flowchart of a method for constructing an urban building model based on climate environment simulation according to an embodiment of the present application. The transaction protection method includes the following steps:

[0051] In step S1, a set of input images of the target urban building is obtained from a climate environment simulation platform.

[0052] In specific implementation, first, clarify the scope of the target city, define the research area in the GIS platform, and screen the POI (Point of Interest) and AOI (Area of Interest) data, as well as high-resolution satellite images or drone aerial images corresponding to the area. These data usually cover core information such as building locations, boundaries, and functional categories. Then, perform data cleaning on the initially collected data to eliminate records that are outside the research scope or do not belong to buildings. For example, use the geographical boundary intersection algorithm to delete non-building data and remap the POI and AOI categories based on classification standards (such as the "General Code for Civil Buildings"). For the POI and AOI data formats, convert them into GIS point features and polygon features respectively to achieve consistency with the image set. Subsequently, preprocess the image data. This mainly includes resolution unification, image cropping, denoising processing (such as Gaussian filtering), and color enhancement to make the images have sufficient analysis quality. Replace blurred or noisy images to ensure the coverage rate and data quality of the research area. Finally, store the preprocessed data in a unified format (such as GeoTIFF or NetCDF), which will not be elaborated here.

[0053] It should be noted that in this application, the input image set refers to the high-resolution image data set of the buildings in the target city in climate environment simulation. These images can include satellite images, drone aerial images, two-dimensional floor plans or three-dimensional models of buildings, etc.

[0054] In step S2, perform hierarchical analysis on the input image set to obtain the building interest point data of the target city buildings in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data.

[0055] In this embodiment, the hierarchical analysis of the input image set to obtain the building interest point data of the target city buildings in different building areas can be implemented by the following steps:

[0056] Extract features from multiple images in the input image set to obtain corresponding multiple feature images;

[0057] Perform category division according to all feature images to obtain a category mapping table corresponding to each category of feature images;

[0058] Project the category mapping table into the input image set for screening to obtain the building interest point data of the target city buildings in different building areas.

[0059] In specific implementation, first, preprocess the input image set to ensure data consistency. Common preprocessing techniques include image size normalization (unifying the resolution) and denoising (such as Gaussian filtering). For example, for satellite images and drone-captured images from different sources, uniformly adjust the image resolution to 0.5 meters and remove blurred or noisy areas. Then use computer vision algorithms (such as SIFT, SURF, ORB, etc.) to extract significant feature points in the images. The feature points can be the edges or corner points of important elements such as the outlines of buildings, windows, doors, roofs, etc. Combine the extracted feature points with the corresponding regional information to generate a feature image. The feature image is not simply the original image, but an image containing key information of the building structure marked by computer vision algorithms. Then, use machine learning algorithms (such as K-means clustering, support vector machine (SVM), or deep learning models) to classify the feature image. Each category represents a building type or building area. For example, one category of images may represent high-rise residential buildings, and another may represent commercial buildings. Generate a category mapping table based on the classification results, which lists the specific categories of each type of feature image and their corresponding building areas. For example, the category mapping table may contain "Category 1: High-rise residential buildings", "Category 2: Commercial buildings", "Category 3: Public facilities", etc., and indicate the set of feature images for each category. Finally, apply the generated category mapping table to the input image set for image screening. By matching the features of each image with the predefined categories in the category mapping table, filter out the images of interest points in each building area. For example, in the images of the residential area, the system will automatically identify the feature images belonging to the "residential building" category and extract the building interest points (such as windows, balconies, etc.) in that area. Extract building interest point data from the screened images. These interest points can be structural elements such as the edges of buildings, windows, doors, roofs, etc., representing the geometric form and functional layout of the building.

[0060] It should be noted that in this application, the feature image represents an image containing key structural information of the building; the category mapping table refers to a table or data structure that establishes a relationship between image features or data points and predefined categories; the building interest point data refers to the key positions or feature points extracted from building images or building geometric models.

[0061] In this embodiment, determining the building image distribution domain of the urban building model in climate environment simulation based on the building interest point data can be implemented by the following steps:

[0062] Convert the building interest point data into building element points of the urban building model in climate environment simulation;

[0063] Determine the building attribute elements of the urban building model in climate environment simulation according to the building element points;

[0064] Determine the building image distribution domain of the urban building model in the climate environment simulation according to the described building attribute elements.

[0065] When specifically implemented, first, standardize the building interest point data (such as windows, doors, roofs, walls, etc.) extracted from the building image to make it compatible with the model format required for climate environment simulation. These interest points usually include information such as the spatial coordinates, dimensions, and types of the building, and then convert the building interest point data into specific building element points. Building element points usually include the vertices, boundaries, and key structural points of the building, representing the geometric characteristics of the building. For example, the corner points of the building, the center points of the windows, etc. can all be regarded as building element points, and using the building interest point data, through geometric derivation (such as calculating the boundaries, facades, roof contours, etc.) of the building, detailed geometric element points are constructed for the building in the climate simulation model. For example, the volume, planar structure, and facade form of the building are derived from the characteristics of the windows, doors, and roofs. Then, according to the geometric information of the building element points (such as the height, width, surface area, etc. of the building), calculate the basic attributes of the building. The building attribute elements may include the type, form, material, facade characteristics, window-wall ratio, etc. of the building. By analyzing the building element points and combining the functional division of the building, infer the functional categories of the building (such as residential, commercial, public facilities, etc.). These categories can be further associated with the physical attributes of the building (such as window area, roof type, etc.) to determine the building attribute elements. And spatially associate the building attribute elements with the building element points, that is, determine the physical attributes represented by each building element point in the climate simulation. For example, the window-wall ratio of a residential building may be larger, while a commercial building may have more glass curtain walls, and the attribute elements will vary with the building category. Finally, use the building attribute elements to determine the specific distribution of each building in the simulation through spatial mapping techniques (such as rasterization or regional division). According to the geometric form and functional division of the building, determine its corresponding climate environment simulation area. For example, the building image distribution domain of a high-rise building may cover a larger spatial area, while that of a low-rise residential building is smaller. Based on the geometric characteristics, attribute elements, and spatial information of the building category of the building, generate the image distribution domain of the building.

[0066] It should be noted that in this application, the building element points represent the basic data units in the urban building model, representing the geometric key positions or feature points of the building, and are usually used to describe the spatial layout and structural form of the building; the building attribute elements represent the data units describing the physical characteristics of the building, including the facade, window-wall ratio, building materials, functional categories, etc. of the building; the building image distribution domain represents the spatial influence area of the building image in the climate simulation, and is usually used to simulate the impact of the building on the climate environment (such as heat radiation, wind speed change, etc.).

[0067] In step S3, obtain the building geometric feature image of the target urban building complex in the climate environment simulation, fill in the missing parts of the building geometric feature image, obtain the corrected feature attributes of urban buildings under different climate environment categories, and then determine the building form indicators for constructing the urban building model based on the corrected feature attributes.

[0068] Specifically, obtaining the building geometric feature image of the target urban building complex in the climate environment simulation can be achieved in the following way: First, use remote sensing image processing technology to extract the external shape features of buildings from satellite images or drone images. This process can be realized through image segmentation algorithms (such as semantic segmentation based on convolutional neural networks), separating the buildings from the background and extracting geometric information such as their contours and vertices. Then, combine LiDAR data for 3D reconstruction to further obtain three-dimensional information such as the height, volume, and roof shape of the buildings. Next, through geometric fitting technology, integrate the extracted building boundaries and 3D data to generate the building geometric feature image of the target urban building complex. These images can accurately reflect the external structure of the buildings, such as the facade, roof structure, and window positions, and thus provide accurate building geometric input data for climate environment simulation. Finally, uniformly standardize these building geometric feature images to ensure consistent data formats, which will not be elaborated here.

[0069] It should be noted that in this application, the building geometric feature image refers to an image that shows the geometric form and structural features of buildings. These images not only include the external contours of buildings but may also contain geometric features such as building facades, roof shapes, window distributions, and door positions.

[0070] Preferably, in this embodiment, to fill in the missing parts of the building geometric feature image and obtain the corrected feature attributes of urban buildings under different climate environment categories, refer to Figure 3 As shown in the figure, which is a schematic flowchart for determining the corrected feature attributes in some embodiments of this application. The determination of the corrected feature attributes in this embodiment can be achieved through the following steps:

[0071] In step S31, determine the building feature deviations of urban buildings under different climate environment categories based on the building geometric feature map;

[0072] In step S32, use a trust analysis model to analyze the building geometric feature map to obtain the trust coefficient corresponding to the building geometric feature map;

[0073] In step S33, determine the attribute occupancy information of urban buildings under different climate environment categories based on the building feature deviations and the trust coefficient;

[0074] In step S34, determine the corrected characteristic attributes of urban buildings under different climate environment categories according to the attribute occupancy information.

[0075] When specifically implemented, first, extract the geometric attributes of the building from the obtained building geometric feature images, such as shape, size, roof height, window-wall ratio, etc. For example, some buildings may have larger windows or special roof structures, and these geometric features may cause heat energy loss or wind flow changes in climate environment simulations. Then, associate the building with a specific climate environment category according to the climate type of the city (such as tropical, temperate or cold regions). For example, buildings in tropical regions usually have larger window areas to maximize natural lighting, while buildings in cold regions may have more insulation measures. And based on the comparison between the climate environment category and the building geometric features, calculate the possible deviations of the building under various climate environments. For example, the window-wall ratio may be higher in tropical regions and lower in cold regions. By comparing the target building with the standard values of similar buildings under similar climate conditions, determine the characteristic deviation. Next, based on the quality, source and applicability to climate simulation of the building geometric feature map, establish a confidence analysis model. The confidence model assigns a confidence coefficient to each feature image according to factors such as the source (such as satellite imagery, field measurement or 3D scan), accuracy and integrity of the building data. Use the model to calculate the confidence coefficient for each building geometric feature map. The value of the confidence coefficient usually ranges between 0 and 1, and the larger the value, the higher the data reliability of the feature map. Determine the confidence of each feature map by analyzing the data missing situation, noise and the consistency with other data sources. Then, determine the attribute occupancy information of the building under different climate environments. The attribute occupancy information represents the relationship between the design characteristics (such as window-wall ratio, facade type, etc.) of the building under a specific climate environment and its deviation. For example, for the tropical climate category, the building has a larger window-wall ratio, and the data with higher confidence will indicate that such a building has stronger adaptability to solar radiation. Low-confidence data may reflect that the design of the building does not meet the requirements of the tropical climate. And use a weighted algorithm to weight the influence of each building feature according to the confidence coefficient and calculate its attribute occupancy under different climate environments. The weighted result helps to determine the corrected building features. Finally, the geometric features of the building can be adjusted through mathematical modeling (such as linear regression, random forest, etc.). For example, if the window-wall ratio of a building in the tropical climate is too small, the corrected characteristic attribute may be to increase the window-wall ratio to improve the ventilation and lighting efficiency of the building. Finally, the corrected characteristic attributes are calculated and output. These corrected attributes will be used as input data in subsequent climate environment simulations to ensure the high accuracy and adaptability of the simulation results.

[0076] It should be noted that in this application, the building feature deviation refers to the difference between the actual geometric features of a building (such as height, window-wall ratio, etc.) and the expected features under standard or ideal conditions; the confidence coefficient represents a numerical value of the reliability of data or model output, and is usually used to evaluate the validity of input data; the attribute occupancy information is the relationship information describing the difference between the features that a building should have in a specific climate environment and its actual features, reflecting the adaptability of the building to different environments; the modified feature attribute represents the attribute after adjusting the building geometric features (such as window-wall ratio, exterior wall material, etc.).

[0077] In this embodiment, the building form index for constructing the urban building model can be determined by the modified feature attribute through the following steps:

[0078] Determine the building category frequency density corresponding to the point-of-interest data in each building according to the modified feature attribute;

[0079] Determine the building type ratio corresponding to the point-of-interest data in each building;

[0080] Determine the building form index for constructing the urban building model according to the building category frequency density and the building type ratio.

[0081] In specific implementation, first, according to the modified feature attributes (such as window-wall ratio, building height, building facade material, etc.), corresponding point-of-interest data is extracted from each building. These points of interest are usually the prominent features of the building (such as windows, doors, roof forms, etc.), which can help determine the distribution of building categories. Classify the point-of-interest data of each building to determine the occurrence frequency of each type of point of interest in the building. For example, in a high-rise commercial building, the frequency density of windows and facade curtain walls may be relatively high, while in a low-rise residential building, the occurrence frequencies of doors and windows are more prominent. Calculate the frequency density of points of interest of each building category within the building, representing the proportion of each type of point of interest in the building. For example, for a building, the frequency density of the point of interest of windows is 30%, that of the roof is 20%, and that of the door is 50%. Then, by classifying the types of building points of interest (such as "windows", "doors", "roofs", etc.), determine the building type to which each point of interest belongs (such as residential, commercial, public building, etc.). Based on the frequency density of points of interest, calculate the proportion of each building type (such as residential, commercial, public building, etc.) in the building. The calculation method is to compare the frequency density of points of interest of a certain building category with the total frequency density of points of interest of the building to obtain the proportion of the building category. For example: in a comprehensive building, assume that the points of interest of windows, doors, and roofs are 30%, 50%, and 20% respectively, and according to the data of each type of point of interest, the windows and doors may correspond to residential functions, while the roof corresponds to public functions, so as to calculate that the proportion of the residential function of this building is 60%, the commercial function is 30%, and the public function is 10%. Finally, building form indicators include the shape coefficient (the ratio of building volume to external surface area), window-wall ratio (the ratio of window area to external wall area), building plane shape index, etc. For example, through the frequency density of building categories and the proportion of building types, the window-wall ratio of the building can be calculated, and further the thermal performance of the building under different climate environments can be estimated. Through a weighted algorithm or a multi-objective optimization model, comprehensive building form indicators are calculated. According to the calculated building form indicators, the building model is further optimized. For example, if a certain building has a relatively high shape coefficient, it may indicate that its energy consumption is relatively large, and the ratio of building volume to surface area can be reduced through optimized design.

[0082] It should be noted that in this application, the frequency density of building categories refers to the frequency or density of the occurrence of different types of building points of interest (such as windows, doors, roofs, etc.) in a certain building, which is usually used to describe the structural characteristics and spatial layout of the building; the proportion of building types refers to the proportion of different building function types in a certain building, which is usually calculated based on the frequency density of various points of interest (such as windows, doors, etc.) in the building; building form indicators represent indicators that quantify the characteristics of the spatial layout, external shape structure, etc. of the building, usually including the shape coefficient, window-wall ratio, etc., and are used to analyze the adaptability of the building to the environment and energy efficiency.

[0083] In step S4, determine the building category model corresponding to the target urban building complex in different climate environments, add the building attribute label image of the target urban building to the building category model to obtain the gradient influence coefficient of the target urban building image in different building areas, and then determine the building form interest points of the urban building model from the gradient influence coefficient and the building image distribution domain.

[0084] In this embodiment, determining the building category model corresponding to the target urban building complex in different climate environments can be achieved by the following steps:

[0085] Determine the building attribute frequency density of the target urban building complex in different climate environments;

[0086] Determine the characteristic graphic description of the target urban building complex in different climate environments according to the building attribute frequency density;

[0087] Construct the building category model corresponding to the target urban building complex in different climate environments according to the characteristic graphic description.

[0088] In specific implementation, first, collect the building attribute data of the target urban building complex. These attributes may include the window-wall ratio, shape coefficient, exterior facade material, roof type, etc. of the building. Building attributes are important indicators reflecting the adaptability of building structures to the climate. Then, extract these building attribute data through remote sensing technology, LiDAR scanning, or Building Information Modeling (BIM) data. The attributes of each building include its geometric dimensions, functional zoning, thermal performance, etc. Classify the target urban building complex in different climate environments and calculate the frequency density of each building attribute in each climate environment. For example, the window-wall ratio of a building may be relatively large in tropical climates and relatively small in cold climates. By statistically analyzing the frequencies of various building attributes, obtain their densities in the building complex. For example, in a tropical climate, the frequency density of buildings with a window-wall ratio of 0.4 is 70%, while in a temperate climate, the frequency density of buildings with a window-wall ratio of 0.2 is 50%. Then, create a characteristic graphic description of the building complex. Characteristic graphics usually visualize the distribution of building attributes through statistical methods (such as histograms, scatter plots, or heat maps), reflecting the distribution trends of building characteristics in different climate environments. The characteristic graphic description can extract the main characteristics of the building complex through dimensionality reduction methods such as cluster analysis or principal component analysis (PCA). For example, in tropical climates, characteristic graphics may prominently display large windows, sunshade facilities, etc., while in cold climates, characteristic graphics may focus on smaller windows and thick exterior facades. Map the distribution of building attributes into an adaptability model for different climate environments. For example, tropical climates may favor buildings with larger windows, while cold climates are more suitable for buildings with lower window-wall ratios. By analyzing the changes in the frequency density of building attributes in different climate environments, create an adaptive characteristic graphic. For example, in a tropical climate region, the window-wall ratio and exterior facade material of buildings may exhibit a specific frequency density distribution, forming the characteristic graphic of this climate region. Finally, construct a building category model for the target urban building complex. The building category model divides building types (such as residential, commercial, public buildings, etc.) and their performances in different climate environments based on information such as the attribute characteristics and climate adaptability of buildings. For example, residential buildings may have a relatively large window-wall ratio and may include sunshade facilities in tropical climates; while in cold climates, the exterior facades of commercial buildings may use heat-insulating materials. Therefore, the building category model will generate a set of adaptive building types according to the requirements of different climate environments, optimize and verify the building category model to ensure that the model can accurately reflect the adaptability of building characteristics in different climate environments. Verify the accuracy of the model by comparing it with actual building data, make adjustments based on climate data and building performance feedback, and then classify the building complex through algorithms (such as K-means clustering or decision tree classification) to determine the optimal design solutions for each type of building in different climate environments and further refine the category characteristics of the buildings.

[0089] It should be noted that in this application, the building attribute frequency density refers to the frequency or density of a specific building attribute (such as window-wall ratio, shape factor, etc.) in a building group, usually indicating the distribution of building characteristics in different climate environments; the characteristic graph description is to describe the characteristics of the building attribute distribution of the building group in different climate environments; the building category model represents the characteristics of building categories under different climate conditions.

[0090] Preferably, in this embodiment, the building attribute label image of the target city building is added to the building category model to obtain the gradient influence coefficient of the target city building image in different building areas. Refer to Figure 4 As described, this figure is a schematic flowchart of determining the gradient influence coefficient in some embodiments of this application. The gradient influence coefficient in this embodiment can be implemented by the following steps:

[0091] In step S41, determine the building attribute label image of the target city building;

[0092] In step S42, determine the building shape index of the target city building according to the building attribute label image;

[0093] In step S43, use the building category to establish a random forest model to predict the missing building categories of the interest point data and regional interest data of the target city building;

[0094] In step S44, according to the building shape index and the missing building category, add them to the building category model to obtain the gradient influence coefficient of the target city building image in different building areas.

[0095] Specifically, when implemented, first, the building attribute label image of the target city building usually comes from building information model (BIM), satellite images, unmanned aerial vehicle (UAV) aerial images, or light detection and ranging (LiDAR) scan data. These images contain key attribute information of the building, such as window-wall ratio, exterior wall material, building height, roof structure, etc. Through image processing and data extraction technologies (such as image segmentation, object recognition), building attribute data is extracted from the original images. According to the extracted building attribute data, a building attribute label image is generated. Each label represents a building attribute category and is marked in the corresponding area of the building image. For example, the window area may be marked with the "window" label, and the exterior wall area is marked with the "exterior wall" label. Use image processing tools (such as image annotation tools) to correspond the building attribute data with the image. Then, the building shape index is an index used to measure the geometric complexity of a building. The commonly used calculation method is to use data such as the outer contour, area, and perimeter of the building, and calculate the shape index through a formula. For example, the building shape index can be calculated by the following formula: S i =P 2 / A, where P is the outer perimeter of the building and A is the base area of the building. The lower the shape index, the more regular the planar shape of the building indicates, and the higher it is, the more complex the building shape indicates. Then, according to the building attribute label image, geometric features of the building are extracted, such as the outer contour, window layout, etc., and the perimeter and area of the building are calculated. These geometric data are extracted through image processing algorithms (such as edge detection, region analysis). Calculate the shape index of the building to evaluate the geometric complexity of the building. Then, the random forest model is an ensemble learning method that classifies and predicts through multiple decision trees. In this step, the building category (such as residential, commercial, public building, etc.) is used as a label, and a random forest model is established based on building attribute data (such as window-wall ratio, facade material, shape index, etc.). The random forest learns patterns through training data (known building categories and feature data) and is used to predict missing data or building categories that cannot be directly obtained. The random forest model can handle the problem of missing data. For example, when the point-of-interest data or regional interest data of some buildings are missing, the model will predict these missing building categories based on known building features (such as shape index, window-wall ratio, etc.). Through the training of building features and categories, the model can fill in appropriate predicted categories for missing data. When calculating the gradient influence coefficient, consider the influence of the building's shape index and the missing building category on the environment (such as wind field, heat radiation, etc.). Buildings with a lower building shape index may have less influence on the wind field, while buildings with a more complex shape may generate stronger air disturbances and heat island effects. Therefore, according to the shape index and building category, the influence coefficient of the building in different regions is calculated. Finally, the building shape index and the predicted missing building category are added to the building category model to ensure that the influence coefficient of the building in different regions can be accurately reflected. In climate simulation, the gradient influence coefficient of the building can help simulate how the building changes the local climate (such as temperature, wind speed, etc.).

[0096] It should be noted that in this application, the building attribute label image represents building attribute data presented in image format, usually including key features such as the window-wall ratio, facade material, and roof shape of the building, and each region is marked with corresponding building attribute labels; the building shape index is an index that describes the complexity of the building's planar form, usually calculated from the outer perimeter and base area of the building; the missing building category refers to the situation where the category information of some buildings cannot be obtained due to data loss or incompleteness. Through model prediction, the missing category can be supplemented; the gradient influence coefficient represents the degree of influence of the building on the climate environment in different building regions, and is usually used in climate simulation to evaluate the physical effect of the building on the surrounding environment.

[0097] In this embodiment, the building shape interest points of the urban building model determined by the gradient influence coefficient and the building image distribution domain can be implemented by the following steps:

[0098] Determine the building climate zoning of the building according to the building image distribution domain;

[0099] Determine the building form interest points of the urban building model through the building climate zoning and the gradient influence coefficient.

[0100] In specific implementation, the building climate zoning is a region divided according to the climate characteristics of the location where the building is located. For example, different regions of a city may have different climate conditions, such as tropical climate, temperate climate, cold climate, etc. Each climate zoning will affect the requirements of building design (such as window-wall ratio, roof form, facade material, etc.). The building image distribution domain provides the spatial proportion and distribution of the building in the city. Combining the geometric shape of the building and the urban area where it is located, through a Geographic Information System (GIS) or spatial analysis tool, determine the climate zoning to which each building belongs. For example, a building located in the central area of the city may be affected by the urban heat island effect, while a building located in the suburbs may be more affected by the natural climate. According to the climate characteristics of the area where the building is located, map the building image distribution domain to the climate zoning to determine the climate zoning where each building is located. For example, a building located in the tropical area of the city may have its climate zoning marked as "tropical climate", while a building located in a high-latitude area may be marked as the "cold climate" zoning. When determining the building form interest points, first set the adaptability requirements according to the building's climate zoning. Different climate zonings have different standards for the thermal performance, ventilation requirements, sunlight requirements, etc. of the building. Then, weight each interest point of the building through the gradient influence coefficient to determine the location and influence range of the building form interest points. For example, in the tropical climate zone, the building's facade may have a significant impact on the local temperature and thermal radiation, while in the cold climate zone, the building form and window-wall ratio may have a greater impact on daylighting and heat retention. Finally, combine the climate zoning and the gradient influence coefficient to determine the location and influence range of the building form interest points. These interest points are usually the key areas where the building affects the environment (such as air flow, heat conduction, etc.). For example, in the tropical climate, the south facade of the building may become the building form interest point because the south facade absorbs more solar radiation; in the cold region, the north facade of the building may become the key interest point because the north facade is usually exposed to a low-temperature environment, affecting the heat loss of the building.

[0101] It should be noted that in this application, the building climate zoning represents the area where the climate characteristics of the area where the building is located divide the city, and is used to describe the adaptability requirements of the building to climate factors (such as temperature, humidity, precipitation, etc.); the building form interest point represents the key area or location where the building affects the environment (such as wind field, thermal radiation, etc.) in climate simulation.

[0102] In step S5, the urban building model is classified and constructed according to the building form index and the building form interest points, and batch automatic conversion is performed.

[0103] When specifically implemented, first, by calculating the building form indexes of buildings (such as shape coefficient, window-wall ratio, shape index, etc.), a set of structured attribute data is generated for each building. These indexes reflect the geometric characteristics of the buildings and their impacts on the environment (such as heat radiation, air flow, etc.). For example, buildings with a low building shape index usually have relatively regular shapes and can better adapt to the climate environment, while buildings with complex shapes may have a stronger impact on the surrounding environment. According to these form indexes, buildings can be initially classified into categories such as "regular buildings", "complex buildings", "high energy efficiency buildings", etc. Then, the building form interest points (such as the key functional areas, facades, windows, etc. of the building) are used as the basis for further classification during this process. These interest points reflect the energy efficiency, ventilation, lighting and other characteristics of the building. For example, in tropical climates, buildings with large windows usually have a higher heat load; while in cold climates, the building form interest points may include an efficient window-wall ratio and insulation layer. Therefore, by identifying the building form interest points, the classification of buildings can be further refined. After the building classification is completed, the next step is batch automatic conversion. This process relies on rule-based or machine learning algorithms to convert the building form data (such as form index, interest points, etc.) into a software input format suitable for climate simulation. Common conversion formats include BIM, GIS format or NetCDF format. Machine learning models (such as random forest or support vector machine) are used to automatically identify different building types and map them to the corresponding building category models.

[0104] It can be seen that in this application, it can effectively make up for the lack and incompleteness of model data attributes caused by excessive simplification of building forms. Among them, by obtaining high-quality building image data from the climate environment simulation platform, the accuracy and integrity of the building model can be ensured, providing a consistent and accurate data basis for subsequent building analysis and climate simulation, thereby improving the reliability of the simulation results. Hierarchical analysis can effectively extract the key features of buildings, accurately identify the functional differences in different building areas, provide structured data for subsequent building attribute modeling and climate environment simulation, and thus improve the refinement degree of the model and the accuracy of the simulation. By complementing the missing building geometric feature images and correcting the building feature data, the integrity and accuracy of the building model can be ensured, thereby improving the adaptability of the building to different climate environments and the accuracy and practicality of climate simulation. By combining the building category model and the gradient influence coefficient, the influence of the building on the surrounding environment (such as wind field, thermal radiation, etc.) can be accurately described, making the building performance in climate simulation more in line with the actual situation, and thus improving the environmental adaptability and accuracy of the simulation. By automatically classifying and batch-converting building models, the efficiency of building data processing can be greatly improved, manual intervention can be reduced, and a large amount of building data can be ensured to efficiently and consistently adapt to the climate environment simulation system, significantly improving the intelligence and accuracy of urban building analysis.

[0105] In summary, the technical solution adopted in this application can construct a building model under the influence of diverse urban climate environments to improve the accuracy of constructing an urban building model based on climate simulation.

[0106] This application provides a system for constructing an urban building model based on climate environment simulation, with reference to Figure 2 As shown in the figure, which is a schematic diagram of the system for constructing an urban building model based on climate environment simulation according to this embodiment of this application, the model construction system includes:

[0107] An image set acquisition module 100, configured to acquire an input image set of target urban buildings from a climate environment simulation platform;

[0108] A building hierarchical analysis module 200, configured to perform hierarchical analysis on the input image set to obtain building interest point data of the target urban building in different building areas, and determine the building image distribution domain of the urban building model in climate environment simulation according to the building interest point data;

[0109] A feature image filling module 300, configured to acquire building geometric feature images of the target urban building complex in climate environment simulation, perform missing filling on the building geometric feature images to obtain corrected feature attributes of urban buildings under different climate environment categories, and further determine the building form indicators when constructing the urban building model through the corrected feature attributes;

[0110] The building shape determination module 400 is used to determine the building category model corresponding to the target urban building complex in different climate environments, add the building attribute label image of the target urban building to the building category model to obtain the gradient influence coefficient of the target urban building image in different building areas, and then determine the building shape interest points of the urban building model from the gradient influence coefficient and the building image distribution domain;

[0111] The building model construction module 500 is used to classify and construct the urban building model according to the building form index and the building shape interest points, and perform batch automatic conversion.

[0112] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each process and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowcharts and / or block diagrams can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0113] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program. This program can be stored in a computer-readable storage medium, and the storage medium includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc memories, magnetic disk memories, tape memories, or any other medium that can be used to carry or store data and is computer-readable.

[0114] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

Claims

1. A method for constructing an urban building model based on climate environment simulation, characterized in that: The model building method comprises the following steps: Obtain a set of input images of buildings in the target city from a climate environment simulation platform; Performing hierarchical analysis on the input image set to obtain building interest point data of target urban buildings in different building areas, and determining the building image distribution domain of the urban building model in the climate environment simulation according to the building interest point data; Acquire the building geometric feature image of the target urban building complex in the climate environment simulation, perform missing filling on the building geometric feature image, obtain the corrected feature attributes of the urban buildings under different climate environment categories, and then determine the building form index when constructing the urban building model through the corrected feature attributes; Determine the building category model corresponding to the target city building complex in different climatic environments, add the building attribute label image of the target city building to the building category model, obtain the gradient influence coefficient of the target city building image in different building areas, and then determine the building shape interest point of the urban building model according to the gradient influence coefficient and the building image distribution domain; The urban building models are classified and constructed according to the building form indicators and the building shape interest points, and are automatically converted in batches.

2. A method for constructing an urban building model based on climate environment simulation as claimed in claim 1, characterized in that: The input image set is subjected to hierarchical analysis to obtain the building interest point data of the target city building in different building areas, specifically including: Extracting features from a plurality of images in the input image set to obtain a plurality of corresponding feature images; Classify all feature images into categories and obtain a category mapping table corresponding to each category of feature images; The category mapping table is projected onto the input image set for screening to obtain building interest point data of target city buildings in different building areas.

3. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: Determining the building image distribution domain of the urban building model in the climate environment simulation according to the building interest point data specifically includes: Converting the building interest point data into building element points of the urban building model in the climate environment simulation; Determine the building attribute elements of the urban building model in the climate environment simulation according to the building element points; The building image distribution domain of the urban building model in the climate environment simulation is determined according to the building attribute elements.

4. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: Determining the architectural form index when constructing the urban architectural model by modifying the characteristic attributes specifically includes: Determine the building category frequency density corresponding to the point of interest data in each building according to the modified characteristic attributes; Determine the proportion of building types corresponding to the point of interest data in each building; The building form index when constructing the urban building model is determined according to the building category frequency density and the building type ratio.

5. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: The building category models corresponding to the target urban building complexes in different climate environments specifically include: Determine the frequency density of building attributes corresponding to the target urban building complex in different climate environments; Determine characteristic graphic descriptions of target urban building complexes in different climate environments according to the frequency density of the building attributes; A building category model corresponding to the target city building complex in different climate environments is constructed according to the characteristic graphic description.

6. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: Determining the building shape interest points of the urban building model by the gradient influence coefficient and the building image distribution domain specifically includes: Determine the building climate zone of the building according to the building image distribution domain; The building shape interest points of the urban building model are determined by the building climate zoning and the gradient influence coefficient.

7. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: The gradient influence coefficient indicates the influence of a building on the climate environment in different building areas.

8. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: The architectural geometric feature image refers to an image showing the geometric form and structural features of a building.

9. The method for constructing an urban building model based on climate environment simulation according to claim 1, characterized in that: The input image set refers to a high-resolution image data set of target city buildings in climate environment simulation.

10. A system for constructing a city building model based on climate environment simulation, used to execute a method for constructing a city building model based on climate environment simulation as claimed in any one of claims 1 to 9, characterized in that: The model building system comprises: An image set acquisition module is used to acquire an input image set of target city buildings from a climate environment simulation platform; A building hierarchy analysis module is used to perform a hierarchy analysis on the input image set to obtain building interest point data of target city buildings in different building areas, and determine the building image distribution domain of the urban building model in the climate environment simulation according to the building interest point data; The feature image filling module is used to obtain the building geometric feature image of the target urban building complex in the climate environment simulation, fill in the missing parts of the building geometric feature image, obtain the corrected feature attributes of the urban buildings under different climate environment categories, and then determine the building form index when constructing the urban building model through the corrected feature attributes; A building shape determination module is used to determine the building category model corresponding to the target city building complex in different climatic environments, add the building attribute label image of the target city building to the building category model, obtain the gradient influence coefficient of the target city building image in different building areas, and then determine the building shape interest point of the urban building model based on the gradient influence coefficient and the building image distribution domain; The building model construction module is used to classify and construct urban building models according to the building form indicators and the building shape interest points, and automatically convert them in batches.

Images