18 results about "3d building models" patented technology

This invention relates to the field of intelligent reconstruction technology for 3D models, specifically a method and system for intelligent reconstruction of 3D building models based on deep learning. The method includes acquiring multi-source data of the target ancient building; generating a dense point cloud through motion reconstruction and multi-view stereo matching; identifying structural and non-structural components through semantic segmentation and reconstructing an initial mesh model with semantically labeled information; using a graph neural network to analyze the spatial topological relationships of structural components, and combining this with a historical construction knowledge base to generate hierarchical combination methods and connection constraint rules; adaptively adjusting the pose of mesh nodes according to the connection constraint rules to drive the precise assembly of parameterized components, and fusing them to obtain a structured 3D model of the target ancient building. This invention can restore the structural component connection relationships in hidden areas of ancient buildings under non-contact acquisition conditions, significantly improving the structural rationality and geometric accuracy of the 3D model.

The invention discloses a 3D building model modeling method which comprises the following steps: S1, acquiring basic parameters of a target building, and generating a parameterized initial 3D building model based on the basic parameters; s2, parameter space sampling is carried out based on a preset parameter value range, and multiple building parameter combinations are generated; s3, inputting the multiple groups of building parameter combinations into a building energy consumption simulation engine for simulation calculation to obtain building energy consumption simulation results corresponding to each group of parameter combinations; s4, the multiple building parameter combinations and the corresponding building energy consumption simulation results serve as a training sample set, an AI agent model is trained, and the AI agent model is used for establishing a mapping relation from building parameters to energy consumption indexes. Physical simulation, artificial intelligence, an optimization algorithm and the Internet of Things technology are deeply fused, and the core pain points of low efficiency, insufficient precision, disconnection with operation and the like in traditional building performance simulation are successfully solved.

The invention relates to the technical field of computer vision, in particular to a personnel positioning method based on a monocular camera and a building three-dimensional model. The personnel positioning method comprises the following steps: loading a three-dimensional model of a building and establishing a reference coordinate system, calibrating position coordinates and attitude parameters of a monocular camera in the reference coordinate system, and calibrating internal parameters and external parameters of the monocular camera; performing Gaussian filtering denoising, illumination compensation and distortion correction preprocessing on the original image acquired by the monocular camera in sequence; performing personnel target real-time detection by adopting a convolutional neural network target detection model, and outputting personnel bounding box parameters and confidence; and mapping the two-dimensional person bounding box into a three-dimensional space by using a camera pose matrix, and calculating the three-dimensional position coordinates of the person by combining a building three-dimensional model as a boundary constraint condition. The personnel positioning method is low in deployment cost and high in positioning precision.

This invention provides a method and apparatus for 3D reconstruction of buildings using dual-view satellite imagery, relating to the field of image processing technology. The method includes: acquiring left-view and right-view satellite images of the same region to construct a stereo image pair; processing the stereo image pair using a 3D building reconstruction network to obtain a final disparity map; wherein the 3D building reconstruction network includes at least a visual transformer network, a multi-scale edge aggregation module, and a semantic contour disparity upsampling module; and generating a 3D building model based on the final disparity map. This invention can effectively improve the reconstruction accuracy of 3D buildings.

The invention discloses a monocular remote sensing image three-dimensional building modeling method based on multi-task learning, which relates to the technical field of three-dimensional modeling, and comprises the following steps: constructing an instance-level multi-task learning model, and optimizing the positioning of an overall bounding box of a building instance by using a cascade detection module; by adopting a roof frame extraction branch, a roof mask extraction branch, a house bottom mask extraction module based on geometrical relationship reasoning and a height predictor based on a binning mechanism, the problem of house bottom information loss caused by facade shielding and the problem of high height estimation difficulty caused by large height difference of various buildings can be effectively relieved; using a multi-task joint loss function to train an instance-level multi-task learning model; and inputting a monocular image into the trained instance-level multi-task learning model to obtain a house bottom mask and a height value, and further constructing an LOD 1 (Level of Detail 1) building three-dimensional model. According to the invention, robust three-dimensional building modeling can be carried out on a monocular image in a city scene under a non-orthographic condition.

This invention discloses a method, apparatus, device, and storage medium for identifying the roof surface of a three-dimensional building model. The method includes: responding to a target three-dimensional building model uploaded by a target user via a front-end interactive interface, and performing model analysis on the target three-dimensional building model to obtain several triangular faces; determining the target roof surface merging method based on the coordinates of each vertex corresponding to each triangular face and the number of triangular faces corresponding to each triangular face; if the target roof surface merging method is automatic merging, determining at least one pair of adjacent triangular faces based on the coordinates of each vertex corresponding to each triangular face; determining the triangular faces to be merged based on the coordinates of each vertex corresponding to each pair of adjacent triangular faces; merging the triangular faces to be merged to obtain a target merged face; determining the area of ​​the merged face corresponding to the target merged face; if the area of ​​the merged face is greater than a preset area threshold, then determining the target merged face as the roof surface of the target three-dimensional building model.

The invention discloses a three-dimensional model generation method, device and equipment. The method comprises the steps that at least one three-dimensional basic model corresponding to the current environment of a vehicle and a basic layout rule of the at least one three-dimensional basic model are acquired; in response to received style description content input by a user, determining a target style of at least one three-dimensional fundamental model; based on the target style, determining at least one group of mapping files matched with the target style, and respectively applying the at least one group of mapping files to the corresponding three-dimensional basic models in the at least one three-dimensional basic model to obtain at least one target three-dimensional model of the target style; arranging the at least one target three-dimensional model based on a basic layout rule; and rendering and displaying the arranged at least one target three-dimensional model in real time. Compared with a manual modeling mode in the traditional technology, the scheme can automatically generate the 3D building model meeting the personalized style requirement of the user, so that the generation efficiency of the stylized three-dimensional model is improved.

Disclosed are systems and method for determining information related to building materials based on determined measurements from a multi-dimensional building model comprising features and elements embodying such materials and measurements. The multi-dimensional model may be based on a plurality of received images, such as ground-based images of a building. The multi-dimensional model may be scaled, or a scale extracted based on data of the model. The multi-dimensional model may comprise architectural elements, and the scale used to determine measurements of such architectural elements. With the scaled measurements of the architectural elements in the model, product information related to multi-dimensional model and its elements may be derived and combined in alternative means.

This invention belongs to the field of geographic information system technology and discloses a method, system, terminal, and storage medium for texture restoration of 3D building models based on a bimodal collaborative restoration network. The method includes: acquiring an image of the building to be restored; performing occlusion detection and segmentation on the image of the building to be restored based on a semantically guided visual model to obtain an occlusion region mask; inputting the image of the building to be restored and the occlusion region mask into a bimodal collaborative restoration network for texture restoration; wherein the bimodal collaborative restoration network is an image restoration network employing a symmetrical U-shaped encoding-decoding hierarchical structure fused state space model; and outputting a texture-restored building image. This invention significantly reduces the computational complexity of the model while ensuring restoration quality, is suitable for high-resolution, large-scale urban scenes, and can effectively improve the integrity, structural rationality, and visual realism of 3D building model textures.

The invention discloses an in-city village property right safety and space reconstruction collaborative optimization method and device, and relates to the technical field of city updating and space governance. The method is used for solving the problems that property conflict overlying is complex, the extrusion risk of low-income tenants is high, and space schemes are difficult to plan as a whole. The method comprises the following steps: firstly, integrating cadastral data, a three-dimensional building model and income distribution information to generate a property right conflict distribution map; then, simulating a multi-class property right disposal path and spatial constraints thereof based on the conflict graph, calculating a demolition release space and plot ratio compensation demand of the illegal building, and outputting a spatial decoupling graph; further, inputting the decoupling graph into a risk model, predicting a tenant loss scale and a social network fracture risk, and generating a risk early warning graph; and finally, constructing a government, villager and developer three-party income distribution model, dynamically adjusting a guarantee house proportion, iteratively optimizing and updating a scheme, and forming a collaborative scheme package considering fairness and space efficiency.

The invention discloses a display method and system based on a building three-dimensional model at a mobile terminal, and belongs to the technical field of real estate digitization and three-dimensional visualization. The method comprises the following steps: a model hierarchical construction step: constructing a three-dimensional model of a ''building-storey-household'' hierarchical structure based on two-dimensional data; a model optimization and interactive packaging step: importing the model into a Unity 3D engine, converting the model into an AssetBundle format, compressing the AssetBundle format, developing an interactive function, and packaging the model into a WebGL application; a private deployment and identifier generation step: deploying the application to a private server, configuring a reverse proxy through Nginx, and generating an associated two-dimensional code; and a mobile terminal asynchronous calling and displaying step: in response to the code scanning request, the server asynchronously transmits the application to a mobile terminal browser for rendering and displaying. According to the method, automatic association of the three-dimensional model and the service attributes, high-performance rendering of the mobile terminal and low-threshold code scanning calling are realized, and the technical problems of information fault, performance bottleneck and high distribution threshold are solved.

This invention relates to the field of electronic digital data processing, and more particularly to a method and apparatus for material segmentation of building facades. The method includes: establishing a 3D building model based on the building project's facade drawings, comprising multiple areas requiring material segmentation, where each material segmentation area is a closed shape enclosed by line segments; inputting parameterized segmentation material rules into the cells, combining them with the closed shapes to obtain segmentation lines; combining all cells to generate a division scheme for the closed shapes of the material segmentation areas; inputting the division scheme into the 3D building model; performing load verification on the obtained 3D building model; obtaining the verification result; and if the verification result meets preset conditions, determining the division scheme as the target division scheme. This invention can solve the problems of low efficiency, difficulty in adapting to complex scenarios, and delays in designing and verifying segmentation and safety verification in traditional material segmentation design.

The application provides a 3D scene generation method and device based on structure control information guidance, equipment and medium, belonging to the technical field of image processing, based on the scene generation requirement, the 3D building model of target building structure is constructed;Based on the 3D building model, the wireframe structure diagram and the depth map of multiple perspectives are generated;The wireframe structure diagram and the depth map of multiple perspectives are used as structure control information, and the pre-trained image generation model is guided to generate multiple target style images under different perspectives in combination with target 3D scene text prompt words;The generated multiple target style images under different perspectives are used as texture information, and are mapped to the 3D building model to obtain a target 3D scene model.The application ensures the consistency of scene style according to the structured controllable 3D scene model.Meanwhile, the target 3D scene text prompt word is guided to generate, and the scene homogenization problem caused by asset reuse in the existing digital asset library technology is solved.

This disclosure provides a method, apparatus, and medium for separating vector data from a 3D building model, relating to the field of 3D technology. The method includes: acquiring first vector data of a 3D building model, the first vector data including node coordinates and connection order of multiple 3D building models, where three nodes in the connection order form a triangular plane, and multiple triangular planes form the surface of the 3D model; reading the coordinates of three nodes in the connection order from the first vector data, and determining whether the triangular plane formed by the next set of three node coordinates intersects with the triangular plane formed by the previous set of three node coordinates; if so, assigning the next set of three node coordinates and the previous set of three node coordinates to the same 3D building model, until all the first vector data is divided into multiple 3D building models. This disclosure achieves the splitting of the building model by determining whether the nodes forming the triangular planes intersect and thus whether they belong to the same building.

The invention provides a construction method of a 4D building model, and relates to the technical field of construction progress management. The method comprises the steps of obtaining construction plan data of a building project, and generating reference plan characteristics and plan arrangement based on the construction plan data; carrying out lightweight processing on the 3D building model based on the reference plan features; spatial reconstruction is carried out on the 3D building model after lightweight processing based on the reference plan features, and sub-models corresponding to all the components in the 3D building model are obtained; for the sub-model of any component, binding the sub-model with a time axis and a key frame respectively; and driving each sub-model based on the planned arrangement to obtain the 4D building model. The method can improve the management efficiency of the construction progress.

This invention relates to the field of machine learning, specifically to a domain-adaptive method and system for 3D building reconstruction from monocular remote sensing images. The method includes: constructing and initializing a dual-model architecture consisting of a student model and a teacher model; training the student model using a source domain dataset, calculating semantic segmentation loss, height prediction loss, and offset prediction loss, and updating the teacher model; performing structured cross-domain mixing of source and target domain data to generate mixed data; calling the teacher model to generate high-confidence pixel category pseudo-labels for the mixed data and target domain data; constructing positive and negative sample pairs, introducing a contrastive learning mechanism to train the student model, and obtaining the target model; inputting the monocular remote sensing image to be processed into the target model, outputting building roof masks, roof footprint masks, building height, and offset vectors between the roof and footprints, constructing three-dimensional coordinates, and generating a 3D building model. This invention solves the domain offset problem caused by differences in imaging conditions across regions.

This invention discloses a method for reconstructing 3D buildings using radar point clouds. The invention includes the following steps: First, the timestamps of the point cloud and the image are aligned using visual offsets to obtain a mapping relationship between the point cloud and the image; next, the point cloud with continuous time trajectories is stitched and cropped according to pose to obtain the final building point cloud; the point cloud is filtered using a reflectivity threshold, and then architectural semantic features are extracted and distinguished using various architectural prior knowledge to generate a structured facade information representation; finally, the facade information is converted into an architectural information model, and texture features are projected to ultimately form a textured 3D building model. This invention enables the direct generation of 3D building models from point clouds, combining real-time processing capabilities with innovative processing strategies. It is widely applicable to multiple fields such as architectural design, urban planning, and cultural heritage protection, providing a new, efficient, and accurate approach for 3D building reconstruction.

This invention discloses a building measurement method based on UAV remote sensing. In the field of building measurement technology, the method utilizes GPS to acquire the initial position information of the UAV, measures surrounding building information, constructs an initial 3D building model, records the UAV's movement trajectory, randomly selects trajectory points from the trajectory, and continues until all layers of building information in a multi-layer framework are measured, outputting the building model. The advantages of this invention are: by extracting trajectory points and calculating the distance to adjacent trajectories, the coverage integrity of the current layer can be verified, avoiding repeated or missed measurements; data is extracted according to a preset cycle, avoiding redundant sampling, reducing data volume while ensuring coverage of key points, and shortening the overall measurement time; the measurement methods of LiDAR, millimeter-wave radar, and ultrasonic radar are broken down into measurement units, verification units, and auxiliary measurement methods; and abnormal data is identified and eliminated through data comparison of different combinations.