Vector map and three-dimensional model fusion method and system

Abstract

The invention relates to a fusion method and system of a vector map and a three-dimensional model, and relates to the field of map modeling, and the method comprises the steps: selecting a vector map fusion region through a user side, extracting and classifying two types of element types, calibrating the three-dimensional model and size constraint of two types of elements through an element type library, and constructing the distribution space constraint of the three-dimensional model; and finally, on the basis of the information, the two types of elements are respectively modeled and fused to generate a geographic information three-dimensional model, so that the technical problem of relatively poor fusion precision of small-scale elements caused by limited information precision of the vector map when the vector map and the three-dimensional model are fused is solved. The precision and effect of the vector map and the three-dimensional model during small-scale element fusion are effectively improved, and the reality sense and availability of fusion display are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of map modeling, and in particular to a method and system for fusing vector maps with three-dimensional models. Background Art

[0002] With the rapid development of geographic information systems (GIS) and 3D visualization technology, the integrated display of vector maps and 3D models has become an important research direction. Vector maps, with their precise geographic location information and rich feature types, provide the foundation for the expression of geographic information, while 3D models, with their intuitive and three-dimensional display effects, provide a more vivid form of visualization for geographic information.

[0003] However, in existing fusion displays of vector maps and 3D models, while vector maps can provide geographic location information and feature types, the accuracy of the information they can provide is limited. In particular, for map elements with a scale below a certain threshold, vector maps can often only display their feature type and approximate location without providing sufficient detailed information, which is crucial for the construction and display of stereoscopic 3D scenes. For example, the specific shape and size of buildings, as well as tiny objects in the surrounding environment, are crucial. Due to this limitation in information accuracy, existing fusion methods often struggle to achieve ideal fusion accuracy when dealing with small-scale features. This results in the inability to accurately display some important details in stereoscopic 3D scenes, affecting the realism and usability of the entire scene. Therefore, how to improve the accuracy of the fusion of vector maps and 3D models for small-scale features has become a technical problem that needs to be urgently addressed in the current fields of geographic information systems and 3D visualization technology. Summary of the Invention

[0004] The present invention aims to solve the technical problem in the prior art that when a vector map is fused with a three-dimensional model, the fusion accuracy of small-scale elements is poor due to the limited accuracy of the vector map information. A method and system for fusing vector maps and three-dimensional models are provided to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for fusing a vector map with a three-dimensional model, comprising: selecting a fusion area on a vector map through a user terminal to obtain a target area; extracting a first-class feature type and a second-class feature type within the target area, wherein the first-class feature type has a first feature position identifier and a first feature size identifier, and the second-class feature type has a second feature position identifier; inputting the second-class feature type into a feature type library, calibrating the feature three-dimensional model and feature size constraints; constructing a distribution space constraint for the second-class feature type based on the second feature position identifier; performing feature size optimization based on the feature three-dimensional model and the feature size constraints, combined with the distribution space constraints of the second-class feature type, to obtain a predicted feature size identifier; modeling the first-class feature type based on the first feature position identifier and the first feature size identifier, and modeling the second-class feature type based on the second feature position identifier and the predicted feature size identifier, generating a target area geographic information stereo model, and performing fusion of the vector map and the three-dimensional model.

[0007] In a second aspect, the present invention provides a fusion system for a vector map and a three-dimensional model, the system comprising: an area selection module for performing a fusion area selection on a vector map through a user terminal to obtain a target area; an element extraction module for extracting a first type of element and a second type of element within the target area, wherein the first type of element has a first element position identifier and a first element size identifier, and the second type of element has a second element position identifier; an information calibration module for inputting the second type of element into an element type library, calibrating the element three-dimensional model and the element size constraint; a constraint construction module for constructing a distribution space constraint for the second type of element based on the second element position identifier; a size optimization module for performing element size optimization based on the element three-dimensional model and the element size constraint, combined with the distribution space constraint of the second type of element, to obtain a predicted element size identifier; a fusion execution module for modeling the first type of element based on the first element position identifier and the first element size identifier, and modeling the second type of element based on the second element position identifier and the predicted element size identifier, generating a three-dimensional model of geographic information of the target area, and executing the fusion of the vector map and the three-dimensional model.

[0008] The beneficial effects of the present invention are: by selecting the vector map fusion area on the user side, extracting and classifying two types of feature types, using the feature type library to calibrate the three-dimensional models and size constraints of the two types of features, constructing their distribution space constraints, and obtaining predicted sizes through size optimization, and finally modeling the two types of features separately based on this information and fusion to generate a geographic information three-dimensional model, effectively improving the accuracy and effect of the vector map and the three-dimensional model when fusing small-scale features, and enhancing the realism and usability of the fusion display. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 This is a flow chart of the method for fusing a vector map and a three-dimensional model provided by the present invention.

[0010] Figure 2 This is a structural diagram of the vector map and three-dimensional model fusion system provided by the present invention.

[0011] Explanation of the accompanying symbols: area selection module 11, feature extraction module 12, information calibration module 13, constraint construction module 14, size optimization module 15, fusion execution module 16. DETAILED DESCRIPTION

[0012] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0013] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0014] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0015] Example 1:

[0016] like Figure 1 As shown, an embodiment of the present invention provides a method for fusing a vector map with a three-dimensional model, including:

[0017] S10: Through the user end, the vector map is framed to select the fusion area to obtain the target area.

[0018] For example, the scene requirements for the integration of vector maps and three-dimensional models are mainly reflected in the need to accurately and intuitively display geographic spatial information, especially in the fields of urban planning, architectural design, disaster simulation, virtual tourism, etc., which require both accurate presentation of geographic location and feature types, and the ability to display the three-dimensional form and detailed information of the scene through three-dimensional models to meet the needs of multi-dimensional, multi-level analysis, display and application of geographic spatial information.

[0019] Among them, vector map refers to a form of map related to the fusion display of three-dimensional models. Vector maps store and express map information in the form of vector data, and depict various elements on the map, such as points, lines, and surfaces, by recording geometric features such as coordinates, directions, and angles. Unlike pixel-based raster maps, vector maps do not distort when scaled and can maintain the clarity and accuracy of graphics. However, when vector maps are fused with three-dimensional models for display, they face the problem of poor fusion accuracy of small-scale elements. For example, in city maps, roads, building outlines, etc. can be accurately represented by vector data. However, when some small street signs or small building details are displayed in combination with three-dimensional models, the accuracy is limited.

[0020] This solution proposes a method for fusing vector maps with 3D models. First, the user interface provides an intuitive and easy-to-use selection tool. This tool features precise coordinate positioning and region selection, allowing users to freely select specific regions on the vector map for fusion based on their needs. This is known as user-side selection of the fusion region. During selection, the user interface displays the bounding coordinates and area information of the selected region in real time, ensuring the accuracy and traceability of the selected region. For example, in an urban planning scenario, a user may wish to select an area containing multiple buildings and roads for 3D display. The selection tool provided by the user interface can meet this requirement. After selection, the system automatically extracts the vector map data within the selected region and identifies it as the target area. This target area not only contains geographic information but also defines the specific scope of the subsequent fusion processing, laying the foundation for feature extraction, 3D modeling, and fusion display. This step enables precise positioning and data processing of specific areas, effectively improving the targetedness and efficiency of the fusion process while also enhancing the realism and usability of the final display.

[0021] S20: Extracting a first-category element type and a second-category element type in the target area, wherein the first-category element type has a first element position identifier and a first element size identifier, and the second-category element type has a second element position identifier.

[0022] Optionally, feature extraction operations are further performed on the selected target area, which is specifically divided into identification and classification of first-class feature types and second-class feature types.

[0023] The first type of feature usually refers to large-scale features, such as buildings and main roads. These features have clear geometric form and size information in the vector map, such as feature identifiers such as building outline polygons and road width attributes. The system will completely retain their first element location identifier (that is, specific geographic location coordinates) and first element size identifier (such as length, width, height and other precise size data) to facilitate subsequent accurate 3D modeling. For example, when processing a vector map of an urban area, the system can identify and extract the outline polygons of high-rise buildings and their specific dimensions, ensuring that the building's form can be faithfully restored in the 3D model.

[0024] Category II features typically include small-scale features, such as streetlights and manhole covers. If the size of these features in the vector map is below a preset threshold, the system will simplify them according to rules. These features may be reduced to point features, retaining only their secondary feature location identifier (i.e., location coordinates) and feature type information, or assigned a default size, such as simplifying streetlights into circular point features with a uniform radius of 1 meter. The preset threshold is a metric for determining size and scale, determining whether simplification should be performed. This value is determined based on actual conditions and is not specifically limited.

[0025] Through feature classification and processing, we can not only ensure the accurate presentation of large-scale features in the three-dimensional model, but also reasonably simplify small-scale features, reduce the amount of data processing, and improve the efficiency of fusion processing, while maintaining the realism and usability of the overall scene, providing a high-quality data foundation for subsequent three-dimensional modeling and fusion display.

[0026] S30: Input the two types of feature into a feature type library, and calibrate the feature three-dimensional model and feature size constraints.

[0027] Specifically, further calibration operations are performed on the extracted second-category feature types, which cover small-scale features such as street lamps and manhole covers. The system inputs these second-category feature types into a pre-built feature type library, which stores the mapping relationship between various features and corresponding three-dimensional models and size constraints. The feature type library includes feature type identifiers, three-dimensional model templates, and size constraint ranges. The system searches the feature type library, matches and calibrates the corresponding three-dimensional models and feature size constraints based on the specific identifiers of the second-category feature types. For example, for the second-category feature type of street lamps, the system finds its corresponding three-dimensional model template in the feature type library, such as a model of a cylinder and a lamp head, and determines its size constraint range, such as the height range, lamp head diameter, etc.

[0028] Through this process, the system can accurately determine the three-dimensional representation and size range for each type of second-category feature, providing an accurate and consistent model and size basis for subsequent three-dimensional modeling, thereby effectively improving the realism and accuracy of the three-dimensional model during the fusion process, and ensuring the reasonable display of small-scale features in the fusion scene.

[0029] S40: Constructing a second-category feature type distribution spatial constraint based on the second feature location identifier.

[0030] Furthermore, based on the extracted second element position identifier of the second type of elements, that is, the precise coordinate information of the elements in the vector map, the distribution space constraint is constructed.

[0031] During the construction process, the system first extracts the boundary locations of the target area from the vector map to construct a basic spatial constraint framework for the target area's distribution. Then, based on the first element location and first element size identifiers of a feature type (such as large-scale features like buildings and main roads), the system deploys them within this framework to determine the spatial distribution of a feature type.

[0032] Next, the system will remove the space occupied by the first type of feature from the basic target area distribution space constraints, and obtain several mutually unconnected redundant distribution spaces. These spaces are the areas where the second type of feature may exist. For example, on a street, the system will first determine the location and size of buildings and roads, and then exclude these areas from the overall space of the street. The remaining space is where small-scale features such as street lights and manhole covers may be distributed. Further, based on the second element position identifier of the second type of feature, these small-scale features are allocated to the corresponding redundant distribution space, and specific distribution constraints are constructed for each space, such as the minimum spacing between features, the relative position relationship with roads or buildings, etc.

[0033] By constructing such distribution space constraints, the system can ensure that the distribution of the two types of elements in the fused three-dimensional scene is reasonable and realistic, effectively improving the realism and usability of the fused scene, and providing a strong spatial layout basis for subsequent three-dimensional modeling and fused display.

[0034] S50: Based on the feature three-dimensional model and the feature size constraint, combined with the two-category feature type distribution space constraint, perform feature size optimization to obtain a predicted feature size identifier.

[0035] In detail, in order to ensure the reasonable presentation of the second type of feature, that is, small-scale features such as street lights and manhole covers in the three-dimensional scene, the feature size optimization operation is performed based on the calibrated feature three-dimensional model and feature size constraints, and combined with the distribution space constraints of the second type of feature.

[0036] Specifically, the system first obtains a 3D model corresponding to each second-category feature type. This model defines the feature's basic morphology and structural characteristics. It also obtains the size constraints set for that feature type, such as minimum and maximum size limits. These form the basic boundaries for feature resizing. Then, combined with the spatial constraints of the second-category feature type's distribution, which define the feature's possible spatial distribution and its relative positional relationship with other features, an algorithm analyzes how to adjust the feature's size to make it more visually harmonious and consistent with the actual scene while satisfying the spatial constraints. For example, on a narrow street, a streetlight cannot be too large, otherwise it will appear abrupt and unrealistic. The system uses a size optimization algorithm to find the optimal size that satisfies both the size constraints and the spatial distribution requirements, taking into account factors such as the street width, the height of surrounding buildings, and the spacing between streetlights. Finally, through a series of calculations and analyses, a predicted feature size identifier is obtained. This identifier contains the specific size parameters of the feature in the 3D scene, providing an accurate sizing basis for subsequent 3D modeling and integrated display.

[0037] By optimizing the element size, the system can ensure that the size of the second type of elements in the three-dimensional scene is both practical and beautiful and coordinated, effectively improving the realism and visual effects of the fusion scene.

[0038] S60: Modeling the first type of feature based on the first feature position identifier and the first feature size identifier, modeling the second type of feature based on the second feature position identifier and the predicted feature size identifier, generating a three-dimensional model of geographic information of the target area, and performing fusion of the vector map and the three-dimensional model.

[0039] Specifically, in the process of integrating vector maps with three-dimensional models, three-dimensional modeling is performed based on the first element position identifier and first element size identifier of a type of element (such as large-scale elements such as buildings and main roads). The first element position identifier clarifies the precise coordinates of the element in the vector map, while the first element size identifier provides detailed size information of the element, such as the length, width, and height of a building or the width of a road. The system uses this identification information to call the corresponding three-dimensional modeling algorithm to generate a three-dimensional model corresponding to a type of element in the vector map, ensuring that the model is consistent with the original vector map data in terms of spatial position and size.

[0040] For the second type of feature (small-scale features such as street lights and manhole covers), the system models them based on the second feature position identifier and the predicted feature size identifier obtained through size optimization. The second feature position identifier determines the specific location of the feature in the vector map, while the predicted feature size identifier provides the feature with a reasonable size in the three-dimensional scene. Based on these identifiers, the system generates three-dimensional models for the second type of feature. These models are more consistent with the actual scene in size and are coordinated with the surrounding environment in spatial distribution.

[0041] Through the above steps, a three-dimensional geographic information model of the target area is generated. This model not only contains the original feature information in the vector map, but also compensates for the features that do not exist in the original vector map (such as the three-dimensional model and size information of small-scale features), making the model more complete and realistic. Finally, the system performs a fusion operation of the vector map and the three-dimensional model, superimposing the generated three-dimensional model on the original vector map to achieve seamless connection between the two. This fusion process not only improves the visualization effect of geographic information, but also enables users to more intuitively understand the spatial structure and feature distribution of the target area, providing more accurate and comprehensive data support for urban planning, architectural design, disaster simulation and other fields.

[0042] In a preferred embodiment, the two types of elements are input into an element type library, and the element three-dimensional model and element size constraints are calibrated, including:

[0043] Two types of first element types are extracted from the two types of element types.

[0044] The second category of first feature types are input into the feature type library and matched with a historical deployment parameter set, wherein the feature type library is used to store one-to-one correspondence between feature types and deployment parameters.

[0045] Representative parameters are extracted from the historical deployment parameter set to obtain a three-dimensional model of the element and element size constraints.

[0046] Specifically, for processing Class II feature types (small-scale features like streetlights and manhole covers), Class II primary feature types must be extracted from the feature types. This step aims to clarify the specific feature categories being processed. These Class II primary feature types are then entered into a pre-built feature type library. The core function of this feature type library is to store and manage the mapping between various feature types and their corresponding deployment parameter sets. This deployment parameter set contains key information such as the feature's 3D model and feature size constraints.

[0047] When a second-category first feature type is entered, the system performs a matching operation in the feature type library to find the corresponding historical deployment parameter set. For example, if the current second-category first feature type is a streetlight, the system will search the feature type library for all the corresponding historical deployment parameters. These parameters cover the 3D models of different streetlight models and their respective dimensional constraints.

[0048] Next, representative parameter extraction is performed on the matched historical deployment parameter set. This process uses algorithmic analysis to select the most representative and universally applicable 3D feature models and feature size constraints from the numerous historical parameters. For example, for a streetlight feature type, the system might select a widely used and visually appealing 3D streetlight model based on historical deployment data and determine its typical size range as the feature size constraint.

[0049] Through this series of operations, the three-dimensional model and feature size constraints of the second-category first feature type can be calibrated efficiently and accurately, providing a reliable data basis for subsequent three-dimensional modeling and fusion processing, effectively improving the realism and accuracy of the fusion scene, and reducing the errors and time consumption caused by manual parameter setting.

[0050] In a preferred embodiment, extracting representative parameters from the historical deployment parameter set to obtain a feature three-dimensional model and feature size constraints includes:

[0051] A historical deployment three-dimensional model set and a historical deployment size set are extracted from the historical deployment parameter set.

[0052] Based on a predefined shape similarity threshold, clustering is performed on the set of historical deployment three-dimensional models to obtain multiple clusters of historical deployment three-dimensional models, wherein the multiple clusters of historical deployment three-dimensional models have multiple centroid historical deployment three-dimensional models.

[0053] The multi-cluster historical deployment three-dimensional model is traversed, size boundary statistics are performed on the historical deployment size set, and the element size constraint is obtained.

[0054] Optionally, when processing the historical deployment parameter set to extract representative parameters, a historical deployment 3D model set and a historical deployment dimension set are first separated from the set. The historical deployment 3D model set contains the 3D model data used for various elements (such as streetlights and manhole covers) in past deployments, while the historical deployment dimension set records the specific dimensions of these elements at the time of deployment.

[0055] Subsequently, a clustering operation is performed on the collection of historical deployment 3D models using a predefined shape similarity threshold. This step calculates the shape similarity between models and groups 3D models with similarities above the threshold into the same cluster, resulting in multiple clusters of historical deployment 3D models. Each cluster contains several 3D models with similar shapes, and each cluster has a centroid historical deployment 3D model that represents the overall shape characteristics of the cluster model. For example, in a collection of historical deployment 3D models of street lamps, the system may cluster them based on factors such as lamp pole shape and lamp head design, resulting in several clusters of street lamp 3D models with different design styles.

[0056] Next, the system traverses multiple clusters of historically deployed 3D models and, for each cluster, calculates the size boundaries of the corresponding dimensional data in the historical deployment size set. This step aims to identify the reasonable size range for each cluster model, namely, the element size constraints. For example, for a cluster of 3D streetlight models, the system might calculate the minimum and maximum heights of the lampposts within that cluster, as well as the common size range, to determine the size constraints for that cluster.

[0057] This series of operations efficiently and accurately extracts feature 3D models and feature size constraints from the historical deployment parameter set. These representative parameters not only reflect the actual conditions of past deployments but also provide reliable data support for subsequent 3D modeling and fusion processing. Through cluster analysis and dimension boundary statistics, the system ensures that the extracted feature 3D models are representative and that the feature size constraints meet actual deployment requirements, effectively improving the realism and accuracy of the fusion scene.

[0058] In a preferred embodiment, clustering is performed on the set of historical deployment 3D models based on a predefined shape similarity threshold to obtain a multi-cluster historical deployment 3D model, wherein the multi-cluster historical deployment 3D model has a plurality of centroid historical deployment 3D models, including:

[0059] Perform shape similarity analysis on the historical deployment three-dimensional model set to obtain a historical deployment three-dimensional model similarity set.

[0060] Based on a predefined shape similarity threshold and in combination with the historical deployment three-dimensional model similarity set, clustering is performed on the historical deployment three-dimensional model set to obtain multiple clusters of initial historical deployment three-dimensional models.

[0061] Clusters whose number of historical deployment three-dimensional models within the cluster is less than or equal to the cluster number threshold are deleted to obtain the multiple cluster historical deployment three-dimensional models.

[0062] The multi-cluster historical deployment three-dimensional models are traversed, and LOF outlier factor analysis is performed in combination with the historical deployment three-dimensional model similarity set to obtain a multi-cluster LOF outlier factor set.

[0063] From the historical deployment three-dimensional model set, multiple minimum historical deployment three-dimensional models of the multi-cluster LOF outlier factor set are respectively extracted and set as the multiple centroid historical deployment three-dimensional models.

[0064] In detail, when processing a set of historically deployed three-dimensional models, a shape similarity analysis is performed on each model in the set. By calculating the geometric feature similarity between models, such as surface curvature, edge length, volume ratio, etc., a historically deployed three-dimensional model similarity set is generated, which records the similarity values ​​between all models. Subsequently, based on a predefined shape similarity threshold and in combination with the similarity set, a clustering algorithm (such as hierarchical clustering, DBSCAN, etc.) is used to cluster the set of historically deployed three-dimensional models, and models with a similarity higher than the threshold are grouped into the same cluster, thereby obtaining multiple clusters of initial historically deployed three-dimensional models. For example, if the shape similarity threshold is set to 0.8, all models with a similarity higher than 0.8 will be clustered into one cluster.

[0065] To remove noise and outliers, the system further removes clusters where the number of historical deployment 3D models within a cluster is less than or equal to a preset cluster threshold, retaining clusters with a sufficient number of models. Ultimately, a multi-cluster historical deployment 3D model is obtained. Furthermore, to further optimize the clustering results, the system traverses these clusters and performs LOF (Local Outlier Factor) analysis on the models within each cluster, combined with the historical deployment 3D model similarity set. Taking the first historical deployment 3D model as an example, a value k is set as the neighborhood size (e.g., k = 5). K historical deployment 3D models are selected from the largest to the smallest similarity to construct a neighborhood, and the first mean similarity of the models within the neighborhood is calculated. The mean of the similarity means of all historical deployment 3D models is then calculated, setting this as the global mean similarity. Finally, the ratio of the global mean similarity to the first mean similarity is calculated, setting this as the first LOF outlier factor. If a model's LOF outlier factor is significantly higher than that of other models, it indicates that the model may be an outlier. By traversing all clusters and calculating the LOF outlier factor for each model, the system obtains a multi-cluster LOF outlier factor set.

[0066] Then, based on the LOF outlier factor set, the system extracts the model with the smallest LOF outlier factor from the set of historically deployed 3D models within each cluster. These models are then assigned to multiple centroid historically deployed 3D models. The centroid model represents the overall characteristics of the cluster's models and has a low degree of outliers, better reflecting the typical shapes of the models within the cluster. For example, in a cluster containing multiple 3D streetlight models, the system will select the streetlight model with the most typical shape and the highest similarity to the other models as the centroid model through LOF outlier factor analysis.

[0067] Through this series of operations, the system can efficiently and accurately extract multiple clusters of historical deployment 3D models and their centroid models from the historical deployment 3D model set, providing a reliable data foundation for subsequent 3D modeling and fusion processing, and effectively improving the realism and accuracy of the fusion scene.

[0068] In a preferred embodiment, based on the second element location identifier, a second type of element type distribution spatial constraint is constructed, including:

[0069] The boundary position of the target area is extracted from the vector map, and the target area distribution space constraint is constructed.

[0070] Based on the first element position identifier and the first element size identifier, deployment is performed with spatial constraints distributed in the target area to obtain a spatial distribution of a type of element.

[0071] The spatial distribution of the type of element is deleted from the target area distribution spatial constraint to obtain a first redundant distribution space to an Nth redundant distribution space, wherein any two redundant distribution spaces are not interconnected.

[0072] Extract the second feature type corresponding to the feature location identifier belonging to the first redundant distribution space from the second feature location identifier and set it as a first group of second feature types, wherein the distribution space constraint of the first group of second feature types is the first redundant distribution space.

[0073] Until the second type of element type corresponding to the element position identifier belonging to the first redundant distribution space is extracted from the second element position identifier, it is set as the Nth group of second type of element type, wherein the distribution space constraint of the Nth group of second type of element type is the Nth redundant distribution space.

[0074] Furthermore, in the process of constructing the spatial constraints for the distribution of the second type of feature, the boundary position information of the target area is extracted from the vector map, and a spatial constraint framework for the distribution of the target area is constructed based on this information, which clarifies the overall spatial scope of the target area.

[0075] Subsequently, based on the first element position identifier and first element size identifier of a type of feature (such as buildings, main roads, etc.), deployment operations are performed within the spatial constraint framework of the target area distribution. By accurately calculating the spatial occupancy of each type of feature, the spatial distribution results of the type of feature are obtained. This result describes in detail the specific location and size of the type of feature in the target area.

[0076] Next, to determine the distribution space for the second-class feature types (such as streetlights and manhole covers), the system removes the spatial distribution of the first-class feature types from the target area's distribution space constraints. This results in several mutually exclusive redundant distribution spaces, denoted as the first redundant distribution space through the Nth redundant distribution space. These redundant distribution spaces represent the spatial regions unoccupied by the first-class feature types and are potential locations for the second-class feature types.

[0077] The system then extracts the feature location identifiers belonging to each redundant distribution space from the second feature location identifiers and determines the corresponding second-class feature types. For example, the corresponding second-class feature types are extracted from the feature location identifiers belonging to the first redundant distribution space and set as the first group of second-class feature types. The distribution space constraints of this group of second-class feature types are the first redundant distribution space.

[0078] Similarly, the system will process all redundant distribution spaces in sequence, and obtain the second-class feature type corresponding to the feature location identifier extracted from the second feature location identifier and belonging to the Nth redundant distribution space, set it as the Nth group of second-class feature types, and the distribution space constraint of the Nth group of second-class feature types is the Nth redundant distribution space.

[0079] Through this series of operations, the system accurately assigns spatial constraints to each type of secondary feature, ensuring a rational and efficient distribution of secondary features within the target area, avoiding conflicts with primary features while improving spatial utilization. This, in turn, is crucial for subsequent 3D modeling and fusion processing, ensuring the realism and accuracy of the fused scene and providing powerful data support for fields such as urban planning and architectural design.

[0080] In a preferred embodiment, based on the feature three-dimensional model and the feature size constraint, combined with the two-category feature type distribution space constraint, feature size optimization is performed to obtain a predicted feature size identifier, including:

[0081] From the second-category feature type distribution spatial constraints, extract the i-th group of second-category feature types and the i-th group of second-category feature type distribution spatial constraints, N≥i≥1.

[0082] Taking the spatial distribution constraint of the i-th group of second-category feature types as a restriction, based on the feature three-dimensional model and the feature size constraint, the i-th group of second-category feature types are randomly combined and deployed to obtain several i-th group of second-category feature type deployment schemes.

[0083] Traverse the plurality of i-th group second-category feature type deployment schemes, perform frequency mining of combination schemes, and obtain a plurality of support degrees.

[0084] Based on the several supports, the maximum support value of the several i-th group of second-category feature type deployment plans is extracted to obtain the target i-th group of second-category feature type deployment plan, and the i-th group of second-category feature type prediction feature size identifier is configured.

[0085] Preferably, during the feature size optimization process, the i-th group of second-class feature types and their corresponding distribution spatial constraints are extracted from the second-class feature type distribution spatial constraints, where i ranges from 1 to N, and N represents the number of groups of second-class feature types. This step aims to clarify the specific second-class feature type to be processed and its spatial distribution range.

[0086] Subsequently, the system randomly deploys the i-th group of second-category feature types, using the spatial distribution constraints of the i-th group of second-category feature types as constraints, combined with pre-calibrated feature 3D models and feature size constraints. During this process, the system generates several deployment plans for the i-th group of second-category feature types, each of which contains the distribution of a specific number and type of second-category features within a specified space. For example, if the i-th group of second-category feature types is streetlights, the system will randomly generate multiple streetlight deployment plans within the spatial range of their distributable range. The number, location, and size of streetlights in each plan may vary.

[0087] Next, the system traverses these deployment plans and uses a combination plan frequency mining algorithm (such as the Apriori algorithm) to analyze the frequency of occurrence of the element combination in each plan, thereby obtaining a number of support degrees. The support degree reflects the frequency with which a specific element combination appears in all deployment plans and is an important indicator for measuring the quality of the plan. Based on these support degrees, the system further extracts the deployment plan corresponding to the maximum support value and sets it as the target i-th group of second-category element type deployment plan. This plan achieves the optimal distribution of element combinations while satisfying spatial and size constraints. Finally, the system configures predicted element size identifiers for the second-category element types in the target deployment plan. These identifiers contain the specific size parameters of the elements in the optimal deployment plan.

[0088] Through this series of operations, the system can efficiently and accurately find the optimal size and distribution scheme for the second type of features, effectively improving the realism and rationality of the fusion scene, and providing accurate data support for subsequent three-dimensional modeling and fusion display.

[0089] In a preferred embodiment, the plurality of i-th group second-category feature type deployment schemes are traversed to perform frequency mining of combination schemes to obtain a plurality of support degrees, including:

[0090] Randomly select from the said several i-th group of second-category element type deployment plans, and extract the selected i-th group of second-category element type deployment plans.

[0091] Perform j-item frequency mining on the selected i-th group of second-category feature type deployment schemes to obtain multiple j-item support, where the initial value of j is equal to 2, j is an integer, and q≥j≥2.

[0092] When any one of the multiple j support items is less than or equal to the support threshold, or j is equal to q, the support of the selected i-th group of second-category element type deployment plan is configured to be equal to the product of the average of the multiple j support items and the ratio of j to q, and the support of the selected i-th group of second-category element type deployment plan is added to the several support levels.

[0093] Otherwise, j is incremented by one and the loop is executed.

[0094] Specifically, in the process of traversing several i-th group of second-category feature type deployment schemes to perform combination scheme frequency mining, the system first randomly selects one from these schemes and extracts the selected i-th group of second-category feature type deployment scheme. Subsequently, j-item frequency mining operations are performed on the selected scheme, where the initial value of j is set to 2, and j is an integer, and q is the preset maximum number of items threshold, q≥j≥2. This step aims to analyze the frequency of occurrence of different numbers of feature combinations in the scheme. For example, when j=2, the system will mine the occurrence of all two-item combinations in the scheme and calculate the corresponding 2-item support; when j=3, it will mine the occurrence of all three-item combinations and calculate the 3-item support, and so on.

[0095] During the mining process, the system will obtain multiple j-item support. If any of these j-item support is less than or equal to the preset support threshold, or j is equal to q (that is, the maximum number of items has been mined), the system will configure the support of the selected i-th group of second-category feature type deployment plan. The support is equal to the product of the mean of multiple j-item support and the ratio of j to q. This calculation method takes into account the weights of support of different numbers of items, because the more items the support has, the more frequently it has been jointly triggered in history, and the greater the possibility of occurrence, so its weight is correspondingly greater. Through this calculation, the system can more comprehensively evaluate the frequency and rationality of the selected deployment plan.

[0096] Finally, the system adds the calculated support of the selected second-category feature type deployment plan of the i-th group to several support sets. If the above conditions are not met (i.e., the support of all j items is greater than the support threshold and j is less than q), j is increased by 1, and the system continues to execute the loop until the termination condition is met.

[0097] Through this series of operations, the support of several i-th group of second-category feature type deployment plans can be efficiently and accurately mined, providing strong data support for subsequent feature size optimization and helping to improve the realism and rationality of the fusion scene.

[0098] The method for fusing a vector map with a three-dimensional model provided by the embodiment of the present invention has at least the following technical effects:

[0099] 1. By inputting the second-category feature type into the feature type library and calibrating the feature three-dimensional model and size constraints, and combining the distribution space constraints constructed based on the second feature position identifier to optimize the feature size, the reasonable size of the second-category feature in the three-dimensional scene can be determined quickly and accurately, avoiding the subjectivity and errors of manual setting, improving the accuracy and efficiency of feature modeling, and providing a reliable foundation for the subsequent fusion generation of the target area geographic information stereo model.

[0100] 2. Based on the vector map, the boundary position of the target area is extracted to construct distribution space constraints. After determining the spatial distribution of the first-type feature type by combining its position and size identifiers, the redundant distribution space is deleted. Then, corresponding distribution space constraints are assigned to the second-type feature type. This can intelligently divide the space to ensure that the distribution of the second-type features in the target area is consistent with the actual scenario while avoiding conflicts with the first-type features. This optimizes the spatial layout of the features and enhances the realism and rationality of the fusion scene.

[0101] 3. In the process of optimizing feature size, the frequency of combination schemes of the second-category feature type deployment schemes is mined, the support of different numbers of items is calculated, and the support of the final deployment scheme is determined based on the support. The scheme with the maximum support is then extracted as the target deployment scheme. The optimal scheme can be screened out from multiple deployment schemes, taking into account the frequency and rationality of different feature combinations, improving the quality of the fusion scheme, and making the fusion of vector maps and 3D models more accurate and practical.

[0102] Example 2:

[0103] like Figure 2 As shown, based on the same inventive concept as the method for fusing a vector map and a three-dimensional model provided in the first embodiment, an embodiment of the present invention further provides a system for fusing a vector map and a three-dimensional model, the system comprising:

[0104] The area selection module 11 is used to perform fusion area selection on the vector map through the user terminal to obtain the target area.

[0105] The feature extraction module 12 is configured to extract a first-class feature type and a second-class feature type in the target area, wherein the first-class feature type has a first feature position identifier and a first feature size identifier, and the second-class feature type has a second feature position identifier.

[0106] The information calibration module 13 is used to input the two types of feature into the feature type library, and calibrate the feature three-dimensional model and feature size constraints.

[0107] The constraint construction module 14 is used to construct a second-category feature type distribution spatial constraint based on the second feature location identifier.

[0108] The size optimization module 15 is used to optimize the size of the element based on the three-dimensional model of the element and the element size constraint, combined with the spatial constraint of the two types of element type distribution, to obtain a predicted element size identifier.

[0109] The fusion execution module 16 is used to model the first type of feature based on the first feature position identifier and the first feature size identifier, and to model the second type of feature based on the second feature position identifier and the predicted feature size identifier, to generate a three-dimensional model of geographic information of the target area, and to perform the fusion of the vector map and the three-dimensional model.

[0110] Furthermore, the information calibration module 13 is further configured to perform the following steps:

[0111] Extracting two types of first feature types from the two types of feature types; inputting the two types of first feature types into the feature type library and matching them with a set of historical deployment parameters, wherein the feature type library is used to store one-to-one correspondence between feature types and deployment parameters; extracting representative parameters from the set of historical deployment parameters to obtain a three-dimensional feature model and feature size constraints.

[0112] Furthermore, the information calibration module 13 is further configured to perform the following steps:

[0113] A historical deployment three-dimensional model set and a historical deployment size set are extracted from the historical deployment parameter set; clustering is performed on the historical deployment three-dimensional model set based on a predefined shape similarity threshold to obtain a multi-cluster historical deployment three-dimensional model, wherein the multi-cluster historical deployment three-dimensional model has multiple centroid historical deployment three-dimensional models; the multi-cluster historical deployment three-dimensional model is traversed, size boundary statistics are performed on the historical deployment size set, and the element size constraint is obtained.

[0114] Furthermore, the information calibration module 13 is further configured to perform the following steps:

[0115] Perform shape similarity analysis on the historical deployment three-dimensional model set to obtain a historical deployment three-dimensional model similarity set; based on a predefined shape similarity threshold, in combination with the historical deployment three-dimensional model similarity set, perform clustering on the historical deployment three-dimensional model set to obtain multiple clusters of initial historical deployment three-dimensional models; delete clusters in which the number of historical deployment three-dimensional models in the cluster is less than or equal to the cluster number threshold to obtain the multiple clusters of historical deployment three-dimensional models; traverse the multiple clusters of historical deployment three-dimensional models, and perform LOF outlier factor analysis in combination with the historical deployment three-dimensional model similarity set to obtain a multiple cluster LOF outlier factor set; from the historical deployment three-dimensional model set, extract multiple minimum historical deployment three-dimensional models of the multiple cluster LOF outlier factor set, and set them as the multiple centroid historical deployment three-dimensional models.

[0116] Furthermore, the constraint construction module 14 is further configured to perform the following steps:

[0117] Extract the boundary position of the target area from the vector map and construct the target area distribution space constraint; deploy in the target area distribution space constraint based on the first element position identifier and the first element size identifier to obtain a type of element spatial distribution; delete the type of element spatial distribution from the target area distribution space constraint to obtain a first redundant distribution space up to the Nth redundant distribution space, wherein any two redundant distribution spaces are not interconnected; extract the second type of element corresponding to the element position identifier belonging to the first redundant distribution space from the second element position identifier, and set it as the first group of type two element, wherein the distribution space constraint of the first group of type two element is the first redundant distribution space; until the second type of element corresponding to the element position identifier belonging to the first redundant distribution space is extracted from the second element position identifier, and set it as the Nth group of type two element, wherein the distribution space constraint of the Nth group of type two element is the Nth redundant distribution space.

[0118] Furthermore, the size optimization module 15 is further configured to perform the following steps:

[0119] From the spatial distribution constraints of the two-category feature types, extract the i-th group of second-category feature types and the i-th group of second-category feature type distribution spatial constraints, N≥i≥1; with the i-th group of second-category feature type distribution spatial constraints as restrictions, based on the feature three-dimensional model and the feature size constraints, randomly combine and deploy the i-th group of second-category feature types to obtain several i-th group of second-category feature type deployment schemes; traverse the several i-th group of second-category feature type deployment schemes, perform combination scheme frequency mining, and obtain several support degrees; based on the several support degrees, extract the maximum support degree of the several i-th group of second-category feature type deployment schemes to obtain the target i-th group of second-category feature type deployment scheme, and configure the i-th group of second-category feature type prediction feature size identifier.

[0120] Furthermore, the size optimization module 15 is further configured to perform the following steps:

[0121] Randomly select from the several i-th group of second-category feature type deployment schemes and extract the selected i-th group of second-category feature type deployment schemes; perform j-item frequency mining on the selected i-th group of second-category feature type deployment schemes to obtain multiple j-item supports, where the initial value of j is equal to 2, j is an integer, and q≥j≥2; when any j-item support of the multiple j-item support is less than or equal to the support threshold, or j is equal to q, configure the support of the selected i-th group of second-category feature type deployment scheme to be equal to the product of the average of the multiple j-item support and the ratio of j to q, and add the support of the selected i-th group of second-category feature type deployment scheme to the several supports; otherwise, add one to j and execute the loop.

[0122] Through the detailed description of the method for fusing a vector map with a three-dimensional model in the foregoing description, those skilled in the art can clearly understand the system for fusing a vector map with a three-dimensional model in this embodiment. As for the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0123] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for fusing a vector map with a three-dimensional model, characterized in that: include: Through the user end, select the fusion area of ​​the vector map to obtain the target area; Extracting a first-category element type and a second-category element type in the target area, wherein the first-category element type has a first element position identifier and a first element size identifier, and the second-category element type has a second element position identifier; Inputting the two types of feature into a feature type library, calibrating feature three-dimensional models and feature size constraints; Based on the second element location identifier, constructing a second type of element type distribution spatial constraint; Based on the three-dimensional model of the element and the element size constraint, combined with the spatial constraints of the two types of element type distribution, the element size is optimized to obtain a predicted element size identifier; The first type of feature is modeled based on the first feature position identifier and the first feature size identifier, and the second type of feature is modeled based on the second feature position identifier and the predicted feature size identifier, a three-dimensional model of geographic information of the target area is generated, and the fusion of the vector map and the three-dimensional model is performed.

2. The method according to claim 1, wherein Input the two types of elements into the element type library, calibrate the element 3D model and element size constraints, including: Extracting two types of first element types from the two types of element types; Inputting the second type of first element into the element type library and matching it with a historical deployment parameter set, wherein the element type library is used to store one-to-one correspondence between element types and deployment parameters; Representative parameters are extracted from the historical deployment parameter set to obtain a three-dimensional model of the element and element size constraints.

3. The method according to claim 2, wherein Extracting representative parameters from the historical deployment parameter set to obtain a three-dimensional model of the element and element size constraints includes: Extracting a historical deployment three-dimensional model set and a historical deployment size set from the historical deployment parameter set; Based on a predefined shape similarity threshold, clustering is performed on the set of historical deployment three-dimensional models to obtain a multi-cluster historical deployment three-dimensional model, wherein the multi-cluster historical deployment three-dimensional model has a plurality of centroid historical deployment three-dimensional models; The multi-cluster historical deployment three-dimensional model is traversed, size boundary statistics are performed on the historical deployment size set, and the element size constraint is obtained.

4. The method according to claim 3, wherein Based on a predefined shape similarity threshold, clustering is performed on the set of historical deployment three-dimensional models to obtain a multi-cluster historical deployment three-dimensional model, wherein the multi-cluster historical deployment three-dimensional model has a plurality of centroid historical deployment three-dimensional models, including: Performing shape similarity analysis on the historical deployment three-dimensional model set to obtain a historical deployment three-dimensional model similarity set; Based on a predefined shape similarity threshold and in combination with the historical deployment three-dimensional model similarity set, clustering the historical deployment three-dimensional model set is performed to obtain multiple clusters of initial historical deployment three-dimensional models; Deleting clusters whose number of historical deployment three-dimensional models is less than or equal to a cluster number threshold, and obtaining the multiple cluster historical deployment three-dimensional models; Traversing the multi-cluster historical deployment three-dimensional models, performing LOF outlier factor analysis based on the historical deployment three-dimensional model similarity set, and obtaining a multi-cluster LOF outlier factor set; From the historical deployment three-dimensional model set, multiple minimum historical deployment three-dimensional models of the multi-cluster LOF outlier factor set are respectively extracted and set as the multiple centroid historical deployment three-dimensional models.

5. The method according to claim 1, wherein Based on the second element location identifier, a second type of element type distribution spatial constraint is constructed, including: Extracting the boundary position of the target area from the vector map and constructing the target area distribution space constraint; Based on the first element position identifier and the first element size identifier, a spatial distribution of a type of element is obtained by performing deployment in the target area distribution spatial constraint; Deleting the spatial distribution of the type of element from the target area distribution spatial constraint to obtain a first redundant distribution space up to an Nth redundant distribution space, wherein any two redundant distribution spaces are mutually incommunicable; Extracting, from the second feature location identifier, two types of feature types corresponding to feature location identifiers belonging to the first redundant distribution space, and setting them as a first group of two types of feature types, wherein the distribution space constraint of the first group of two types of feature types is the first redundant distribution space; Until the second type of element type corresponding to the element position identifier belonging to the first redundant distribution space is extracted from the second element position identifier, it is set as the Nth group of second type of element type, wherein the distribution space constraint of the Nth group of second type of element type is the Nth redundant distribution space.

6. The method according to claim 5, wherein Based on the feature three-dimensional model and the feature size constraint, combined with the two-category feature type distribution space constraint, feature size optimization is performed to obtain a predicted feature size identifier, including: Extracting the i-th group of second-category feature types and the i-th group of second-category feature type distribution spatial constraints from the second-category feature type distribution spatial constraints, N≥i≥1; Taking the spatial distribution constraint of the i-th group of second-category feature types as a restriction, based on the feature three-dimensional model and the feature size constraint, randomly combining and deploying the i-th group of second-category feature types to obtain several i-th group of second-category feature type deployment schemes; Traversing the plurality of i-th group second-category feature type deployment schemes, performing frequency mining of combination schemes, and obtaining a plurality of support degrees; Based on the several supports, the maximum support value of the several i-th group of second-category feature type deployment plans is extracted to obtain the target i-th group of second-category feature type deployment plan, and the i-th group of second-category feature type prediction feature size identifier is configured.

7. The method according to claim 6, wherein Traverse the plurality of i-th group second-category feature type deployment schemes, perform frequency mining of combination schemes, and obtain a plurality of support degrees, including: Randomly select from the plurality of i-th group of second-category feature type deployment plans, extracting the i-th group of second-category feature type deployment plans; Perform j-item frequency mining on the selected i-th group of second-category feature type deployment schemes to obtain multiple j-item support, where the initial value of j is equal to 2, j is an integer, and q≥j≥2; When any one of the multiple j support items is less than or equal to the support threshold, or j is equal to q, the support of the selected i-th group of second-category element type deployment plan is configured to be equal to the product of the average of the multiple j support items and the ratio of j to q, and the support of the selected i-th group of second-category element type deployment plan is added to the multiple support items; Otherwise, j is incremented by one and the loop is executed.

8. The fusion system of vector map and 3D model is characterized by: A method for fusing a vector map with a three-dimensional model according to any one of claims 1 to 7, the system comprising: The region selection module is used to select the fused region on the vector map through the user terminal to obtain the target region; an element extraction module, configured to extract a first element type and a second element type in the target area, wherein the first element type has a first element position identifier and a first element size identifier, and the second element type has a second element position identifier; An information calibration module, configured to input the two types of feature into a feature type library, and calibrate feature three-dimensional models and feature size constraints; A constraint construction module, configured to construct a spatial constraint on the distribution of the second type of element based on the second element position identifier; A size optimization module is used to optimize the size of the element based on the three-dimensional model of the element and the element size constraint, combined with the spatial distribution constraint of the two types of element types, to obtain a predicted element size identifier; A fusion execution module is used to model the first type of feature based on the first feature position identifier and the first feature size identifier, and to model the second type of feature based on the second feature position identifier and the predicted feature size identifier, to generate a three-dimensional model of geographic information of the target area, and to perform the fusion of the vector map and the three-dimensional model.

Images