Surveying and mapping information intelligent analysis and visualization system based on Internet of Things

Abstract

The invention relates to the technical field of surveying and mapping information processing, and discloses a surveying and mapping information intelligent analysis and visualization system based on the Internet of Things. According to the system, a building space data acquisition terminal acquires geometric feature point cloud and environmental parameters of the surface of a building structure in real time through distributed Internet of Things nodes; the multi-source heterogeneous data fusion engine performs space-time alignment processing on the geometric feature point cloud to generate a building total-factor topological skeleton with an absolute coordinate system; the three-dimensional topological relation reconstruction module is used for extracting a spatial constraint relation of building components according to the topological skeleton and establishing a hierarchical building information model capable of being edited in a parameterized manner; and the dynamic visual rendering platform is used for generating a multi-scale spatial index based on the hierarchical building information model, and driving real-time light and shadow effects and interactive sectioning analysis. According to the system, full-process automatic processing and dynamic visualization of building surveying and mapping data are realized.

Description

Technical Field

[0001] This invention relates to the field of surveying and mapping information processing technology, specifically to an intelligent analysis and visualization system for surveying and mapping information based on the Internet of Things. Background Technology

[0002] Architectural surveying is a fundamental task in urban planning, engineering construction, and cultural relic protection. Its core objective is to obtain accurate spatial geometric information and topological relationships of buildings. Traditional architectural surveying mainly relies on optical instruments such as total stations and theodolites for single-point measurements, or uses close-range photogrammetry to obtain image data from limited perspectives. These methods have limitations such as low data acquisition efficiency, long operation cycles, and susceptibility to environmental interference. Moreover, the acquired data is mostly discrete points or two-dimensional images, making it difficult to directly construct a three-dimensional model that expresses the overall spatial relationships of the building. With the development of 3D laser scanning technology, massive amounts of 3D point cloud data of building surfaces can be quickly acquired through ground-based or airborne lidar, providing a data foundation for high-precision 3D modeling. However, the raw point cloud data only contains spatial coordinates and reflection intensity information, lacking semantic recognition of building components and their topological connections. Manually extracting building elements from disordered point clouds and reconstructing their 3D models remains a time-consuming, labor-intensive task that relies heavily on the experience of professionals.

[0003] Modern architectural surveying often requires the integration of data from multiple sensors, such as laser scanning point clouds, oblique photogrammetry results, indoor positioning trajectories, and environmental monitoring data like temperature and humidity. These data differ in acquisition time, spatial reference, and accuracy scale, forming multi-source, heterogeneous data silos. Current technologies lack effective multi-source data fusion mechanisms, making it difficult to unify data from different sources and with different properties under a single spatiotemporal reference for integrated analysis and representation. In terms of visualization, traditional 3D model displays primarily focus on static rendering, making it difficult to support real-time, multi-scale interactive operations and dynamic cross-sectional analysis based on different application needs. Therefore, there is an urgent need for an intelligent surveying information processing system capable of achieving seamless integration across the entire chain, from automated data acquisition and intelligent fusion processing to parametric modeling and interactive visualization. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent analysis and visualization system for surveying and mapping information based on the Internet of Things, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an intelligent analysis and visualization system for surveying and mapping information based on the Internet of Things, the system comprising: Building space data acquisition terminal, multi-source heterogeneous data fusion engine, three-dimensional topology reconstruction module and dynamic visualization rendering platform; The building space data acquisition terminal acquires the point cloud of geometric features on the surface of the building structure and environmental parameters in real time through distributed Internet of Things nodes. The multi-source heterogeneous data fusion engine performs spatiotemporal alignment processing on geometric feature point clouds to generate a topological skeleton of all building elements with an absolute coordinate system. The 3D topology reconstruction module extracts the spatial constraint relationships of building components based on the topological skeleton and establishes a hierarchical building information model that can be parametrically edited. The dynamic visualization rendering platform generates multi-scale spatial indexes based on hierarchical building information models, driving real-time lighting effects and interactive cross-sectional analysis.

[0006] Preferably, the building space data acquisition terminal includes a lidar array and a multispectral sensor group; The lidar array captures the three-dimensional coordinate sequence of the building facade with millimeter-level precision using a non-contact scanning method; The multispectral sensor array synchronously collects reflectivity data of building surface materials and establishes a pixel-level mapping relationship with the three-dimensional coordinate sequence.

[0007] Preferably, the operations performed by the multi-source heterogeneous data fusion engine include: Motion distortion compensation is performed on the discrete point cloud data acquired by the lidar array according to the scanning time sequence; The feature descriptor matching algorithm is used to eliminate the observation parallax between the multispectral sensor group and the lidar array; By using ground control points to perform coordinate system transformation on the compensated point cloud data, a spatial reference frame conforming to international surveying and mapping standards is generated.

[0008] Preferably, the three-dimensional topological relationship reconstruction module includes a component semantic segmentation unit and a spatial relationship reasoning unit; The component semantic segmentation unit identifies the physical boundaries of building exterior walls, windows, and load-bearing columns based on material reflectivity data; The spatial relationship reasoning unit calculates the geometric constraints of perpendicularity, parallelism, and coplanarity between components based on physical boundaries.

[0009] Preferably, the generation process of a hierarchical building information model includes: Abstract building components that meet geometric constraints into parametric template instances; Establish parent-child cascading references between template instances based on the spatial topological relationships between components; Versioning of cascading reference structures supports tracing historical modifications.

[0010] Preferably, the dynamic visualization rendering platform includes a real-time computing unit and an interactive operation unit; The real-time computing unit dynamically loads building component mesh data of varying precision based on the viewing angle. When the interactive operation unit responds to a spatial sectioning command, it automatically reconstructs the topological connection relationships between components.

[0011] Preferably, the operating modes of the building space data acquisition terminal include: The overall building deformation monitoring data is updated periodically under a fixed monitoring mode; In mobile mapping mode, the coverage of indoor spatial point cloud is incrementally expanded through synchronous positioning and mapping technology.

[0012] Preferably, the multi-source heterogeneous data fusion engine also includes a data quality assessment unit; The data quality assessment unit generates data credibility indicators based on point cloud density, spectral resolution, and coordinate system transformation residuals. The 3D topology reconstruction module dynamically adjusts the decision threshold for semantic segmentation of components based on the credibility index.

[0013] Preferably, the operating protocols supported by the dynamic visualization rendering platform include: Publish tile map services for building information models using the Open Geospatial Consortium standard protocol; This implementation uses a network graphics library protocol to achieve 3D spatial measurement and analysis functions on the browser side.

[0014] Preferably, the system further includes a building code verification module; The building code verification module logically matches the hierarchical building information model with the current building design code clauses; The 3D topology reconstruction module automatically marks the points of spatial topology conflict that violate the specifications based on the matching results.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention significantly improves the efficiency of building surveying information acquisition, the level of intelligent processing, and the application value by constructing an integrated data processing and visualization pipeline. The building space data acquisition terminal utilizes distributed IoT nodes for synchronous data acquisition, expanding data coverage, improving acquisition efficiency, and achieving simultaneous acquisition of environmental parameters and geometric data, providing richer contextual information for subsequent analysis. The core function of the multi-source heterogeneous data fusion engine is to solve the integration problem of multi-source data. Through spatiotemporal alignment processing, it unifies geometric feature point clouds acquired at different times and from different sensors into an absolute coordinate system, generating a topological skeleton of all building elements. This step lays a solid foundation for the transformation from raw data to a structured model, enabling discrete point cloud data to begin to exhibit spatial relationships between building elements.

[0016] The 3D topology reconstruction module, based on the topological skeleton, further intelligently identifies and extracts the spatial constraints between building components such as walls and columns, and establishes a parameterized and editable hierarchical building information model. This model not only contains geometric information but also embodies the logical relationships and hierarchical structure between components, giving it excellent editability and information carrying capacity, facilitating subsequent modifications, updates, and information association. The dynamic visualization rendering platform generates multi-scale spatial indexes based on the hierarchical model, enabling rapid and smooth scaling and browsing from the overall building to local components. Driving real-time lighting effects makes the model display more realistic, helping to intuitively assess the building's spatial sense and lighting conditions. The interactive sectioning analysis function gives users powerful exploration capabilities, allowing dynamic sectioning of the building model from any angle and position, intuitively observing the internal structure and spatial relationships, greatly facilitating professional applications such as design review, facility management, and emergency planning. This system shifts building surveying from a traditional data acquisition-centric model to a modern model centered on intelligent analysis and interactive applications, enhancing the depth and breadth of surveying results utilization. Attached Figure Description

[0017] Figure 1 For building space data acquisition and multi-source fusion analysis; Figure 2 A flowchart illustrating the working principle of a building space data acquisition terminal; Figure 3 A flowchart illustrating the operation of a multi-source heterogeneous data fusion engine; Figure 4 This is a diagram for reconstructing three-dimensional topological relationships and analyzing components. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 This invention provides an intelligent analysis and visualization system for surveying and mapping information based on the Internet of Things, the system comprising: The collaborative operation of the building space data acquisition terminal, the multi-source heterogeneous data fusion engine, the 3D topology reconstruction module, and the dynamic visualization rendering platform is crucial. The building space data acquisition terminal acquires the geometric feature point cloud and environmental parameters of the building structure surface in real time through distributed IoT nodes. The distributed IoT nodes are deployed in key locations around and inside the building to capture spatial data at a high sampling rate. The multi-source heterogeneous data fusion engine receives the raw point cloud data and performs spatiotemporal alignment processing. The spatiotemporal alignment processing is based on the GPS timestamp and local clock synchronization to generate a topological skeleton of all building elements with an absolute coordinate system. The absolute coordinate system adopts the World Geodetic System standard. The 3D topology reconstruction module analyzes the geometric relationships in the topological skeleton, extracts the spatial constraints of building components, and establishes a parameterizable and editable hierarchical building information model. The parameterizable editing supports dynamic adjustment of component size and position. The dynamic visualization rendering platform loads the hierarchical building information model and generates a multi-scale spatial index to optimize rendering efficiency. The multi-scale spatial index dynamically switches the level of detail according to the observation distance, driving real-time lighting effects and interactive section analysis. The real-time lighting effects utilize the graphics processing unit to accelerate calculations.

[0020] Example 1: See Figure 2 The building space data acquisition terminal includes a lidar array and a multispectral sensor group. The lidar array captures the millimeter-precision three-dimensional coordinate sequence of the building facade using a non-contact scanning method. Non-contact scanning measures distance by emitting laser pulses and receiving reflected signals. The lidar array's scanning frequency reaches millions of measurement points per second, each containing three-dimensional coordinates and reflection intensity information. The multispectral sensor group simultaneously acquires the reflectivity data of the building surface material and establishes a pixel-level mapping relationship with the three-dimensional coordinate sequence. This pixel-level mapping is achieved by calibrating the spatial relative positions of the lidar array and the multispectral sensor group. The calibration process uses a high-precision calibration field, which is equipped with targets of known size and coordinates. Both the lidar array and the multispectral sensor group simultaneously acquire calibration field data. The least squares method is used to calculate the transformation matrix between the two. The working modes of the building space data acquisition terminal include fixed monitoring mode and mobile mapping mode. In the fixed monitoring mode, distributed IoT nodes collect the overall deformation monitoring data of the building at a preset period. The deformation monitoring data is calculated by comparing the differences in point clouds at continuous time points. The distributed IoT nodes are installed on fixed bases around the building. The bases have shockproof and temperature control functions to ensure the stability of data acquisition. In the mobile mapping mode, the building space data acquisition terminal is mounted on a mobile platform. The coverage of the indoor space point cloud is incrementally expanded through synchronous positioning and mapping technology. The synchronous positioning and mapping technology uses inertial measurement unit and visual odometry for fusion positioning. The mobile platform can be a ground robot or a drone, equipped with obstacle avoidance sensors and a path planning system.

[0021] The lidar array adopts a rotating or solid-state design to adapt to different scanning scenarios. Rotating lidar achieves wide-angle coverage through mechanical rotation, while solid-state lidar uses phased array technology to improve scanning speed. The scanning range, resolution, and accuracy of the lidar array are configured according to the building size and measurement requirements. For large buildings, multiple lidar arrays work together, and a time synchronization module ensures the consistency of data acquisition. The multispectral sensor group includes visible light and infrared band sensors. The visible light sensor captures color information, and the infrared sensor detects the thermal radiation characteristics of the material. The spectral range of the multispectral sensor group covers 400 nanometers to 2500 nanometers, including blue, green, red, near-infrared, and short-wave infrared bands. The image resolution of each band is matched with the lidar point cloud density.

[0022] In the fixed monitoring mode, the building space data acquisition terminal is installed on a fixed support and periodically generates point cloud time series for long-term deformation analysis. The position of the fixed support is precisely measured and incorporated into the absolute coordinate system. The acquisition cycle of the fixed monitoring mode is set according to the building structure characteristics and environmental factors, ranging from hourly to monthly. The point cloud time series data is stored in a time-series database, supporting fast query and comparative analysis. In the mobile mapping mode, the building space data acquisition terminal is integrated into a robot or drone platform to build a three-dimensional map of the indoor environment in real time. The mobile platform is equipped with a lidar array, a multispectral sensor group, an inertial measurement unit, a global positioning system receiver, and a computer. The computer runs synchronous positioning and mapping algorithms, processes sensor data in real time, and generates point cloud maps.

[0023] The process of establishing pixel-level mapping relationships involves coordinate transformation and image registration algorithms. Coordinate transformation projects the LiDAR point cloud onto the multispectral image plane. Image registration algorithms optimize spatial alignment accuracy. Coordinate transformation uses a rigid body transformation model, including rotation matrices and translation vectors. Image registration algorithms employ feature point matching or mutual information maximization methods. Feature point matching uses a scale-invariant feature transformation algorithm to extract and describe feature points. Mutual information maximization methods utilize information theory criteria to optimize image alignment. The communication module of the building space data acquisition terminal uses IoT protocols to transmit data. IoT protocols include message queue telemetry transmission or restricted application protocols to ensure low-latency data transmission. The communication module supports wired and wireless transmission methods. Wired transmission enables… Using Ethernet or fiber optics, wireless transmission utilizes Wi-Fi or cellular networks. Encryption and verification mechanisms are employed during data transmission to ensure data security and integrity. The data acquisition process of the lidar array is affected by environmental factors such as sunlight, rain, snow, and fog. To reduce environmental impact, the lidar array is equipped with filters and protective covers. The multispectral sensor group uses supplementary lighting devices under low-light conditions. The data acquisition software includes quality control algorithms to monitor data quality and adjust acquisition parameters in real time. The calibration of the multispectral sensor group includes radiometric calibration and geometric calibration. Radiometric calibration converts the sensor output into reflectivity values ​​using a standard reflector as a reference. Geometric calibration corrects lens distortion using a checkerboard calibration plate to calculate the distortion coefficient.

[0024] The power system of the building space data acquisition terminal supports both AC power and battery power. In fixed monitoring mode, AC power is mainly used, and an uninterruptible power supply is provided to prevent power outages. In mobile mapping mode, a high-capacity battery is used, and the battery life meets the needs of continuous operation. The software interface of the data acquisition terminal provides real-time preview and parameter setting functions. Operators can set parameters such as scanning range, resolution, and acquisition cycle. The software interface displays the real-time fusion results of point cloud and image, which facilitates on-site quality control.

[0025] The mechanical structure design of the building space data acquisition terminal considers portability and stability. Terminals in fixed monitoring mode are installed in a protective case that is waterproof and dustproof. Terminals in mobile mapping mode employ a lightweight design to reduce the load on the mobile platform. Maintenance of the data acquisition terminal includes regular cleaning of optical components, inspection of mechanical parts, and software updates. Maintenance records are stored in a log file to track equipment status and historical performance. The data output format of the building space data acquisition terminal supports industry standards. Point cloud data is saved in either the LiDAR data exchange format or the point cloud data format, and multispectral images are saved in either the tag image file format or the bitmap format. Data files contain metadata, recording the acquisition time, equipment parameters, and coordinate system information. The data acquisition terminal interfaces with the multi-source heterogeneous data fusion engine using either an application programming interface (API) or a file transfer protocol. The API is based on a representational state transfer architecture and supports real-time data streaming. The file transfer protocol is used for batch data transmission. Performance testing of the building space data acquisition terminal is conducted in a standard environment. Test metrics include point cloud accuracy, image resolution, acquisition speed, and power consumption. Point cloud accuracy is verified by measuring targets at known distances. Image resolution is evaluated using a resolution test chart. Acquisition speed is calculated by timing the data acquisition time per unit area. Power consumption is measured using a power meter. Test results are used to optimize the configuration and operation process of the data acquisition terminal to ensure its reliability and efficiency in various scenarios.

[0026] Example 2: See Figure 3 The multi-source heterogeneous data fusion engine performs operations including motion distortion compensation, observation parallax elimination, and coordinate system transformation. Motion distortion compensation processes discrete point cloud data acquired by the lidar array, correcting point cloud distortion caused by platform motion according to the scanning sequence. The scanning sequence records the timestamp of each point. Motion distortion compensation is based on inertial measurement unit (IMU) data, which provides the platform's angular velocity and acceleration in three axes. The relative position change of each laser point at the sampling time is calculated through timestamp interpolation. The feature descriptor matching algorithm eliminates the observation parallax between the multispectral sensor group and the lidar array. The sub-matching algorithm extracts key points from the point cloud and the image, and achieves cross-modal alignment through descriptor comparison. The key point extraction uses a scale-invariant feature transformation algorithm, and the descriptor generation adopts an accelerated robust feature algorithm. The matching process calculates the Hamming distance between feature vectors. Ground control points perform coordinate system transformation on the compensated point cloud data to generate a spatial reference frame that conforms to international surveying and mapping standards. The international surveying and mapping standards adopt the universal transverse Mercator projection or the national geodetic coordinate system. The ground control points are set up in the surveying area, and their geodetic coordinates are obtained by measuring through the static global positioning system. The coordinate system transformation adopts the seven-parameter Bursa model.

[0027] The multi-source heterogeneous data fusion engine includes a data quality assessment unit. This unit generates a data reliability index based on point cloud density, spectral resolution, and coordinate system transformation residuals. Point cloud density assessment calculates the number of points per unit volume, using a 3D grid to divide the space and counting the number of points in each grid. Spectral resolution analysis measures the number of bands and bandwidth of the multispectral data, checking the sharpness and noise level of each band image. Coordinate system transformation residuals measure the fitting error of ground control points, calculating the Euclidean distance between the transformed point cloud coordinates and the known ground control point coordinates. The data reliability index quantifies data reliability, ranging from 0 to 1, with higher values ​​indicating better quality. The data reliability index is calculated using a weighted average method, with point cloud density having a weight of 0.4. The resolution weight is 0.3, the coordinate system transformation residual weight is 0.3, and motion distortion compensation uses inertial navigation system data-assisted correction. The inertial navigation system provides platform acceleration and angular velocity information. The inertial navigation system is synchronized with the lidar array hardware, and the time synchronization accuracy reaches the microsecond level. The motion distortion compensation algorithm establishes a platform motion model, which uses the assumption of uniform motion or acceleration motion. The actual position of each laser point is predicted based on the inertial measurement unit data. The feature descriptor matching algorithm uses scale-invariant feature transformation or acceleration robust feature descriptors to achieve robust feature matching. The matching process includes four steps: feature point detection, direction assignment, descriptor generation, and matching filtering. The matching filtering adopts the nearest neighbor distance ratio criterion to eliminate mismatched point pairs.

[0028] The coordinate system transformation process uses the least squares method to optimize the transformation parameters, minimizing the sum of squares of the ground control point residuals. The transformation parameters include three translation parameters, three rotation parameters, and one scale parameter. The output of the data quality assessment unit is fed back to the 3D topology reconstruction module. The 3D topology reconstruction module dynamically adjusts the decision threshold for semantic segmentation of components based on the confidence index. When the confidence index is below 0.8, the confidence threshold for semantic segmentation is relaxed; when the confidence index is above 0.9, the strictness of semantic segmentation is increased. In the observation parallax elimination, the calibration of the multispectral sensor group and the lidar array is completed through a calibration board. The calibration board has a known geometric pattern for accurate calibration. The calibration board pattern adopts a black and white checkerboard or concentric rings, and the pattern size is accurately measured. During the calibration process, multiple sets of calibration board data in different postures are collected. By solving the extrinsic parameters of the camera model and the lidar model, the coordinate transformation relationship between the two sensors is established. The acceptable threshold for coordinate system transformation residuals is set according to the application scenario. High-precision measurement requires residuals to be less than centimeters. Data with residuals exceeding the threshold triggers a re-acquisition mechanism.

[0029] The multi-source heterogeneous data fusion engine adopts a pipeline architecture for its data processing flow. The pipeline includes four stages: data input, preprocessing, fusion calculation, and result output. The data input stage receives LiDAR point clouds and multispectral images. The preprocessing stage performs data decoding and time synchronization. The fusion calculation stage performs motion distortion compensation, feature matching, and coordinate system transformation. The result output stage generates the fused point cloud and a data quality report. The data quality assessment unit integrates machine learning algorithms. The machine learning algorithms use historical data to train a quality prediction model. The quality prediction model automatically assesses the data credibility based on point cloud features. The multi-source heterogeneous data fusion engine supports multimodal data fusion, including LiDAR point clouds, multispectral images, infrared thermal imaging, and synthetic aperture radar data. The fusion algorithm uses a deep learning-based feature extraction network, which includes convolutional layers and pooling layers to learn common feature representations between different modalities. The data quality assessment unit monitors the fusion process in real time. When a decline in data quality is detected, it automatically adjusts the fusion parameters or triggers data re-acquisition.

[0030] The hardware platform of the multi-source heterogeneous data fusion engine is equipped with a graphics processing unit (GPU) accelerator. The GPU accelerator is used for parallel computation of feature descriptor matching and coordinate system transformation. The software architecture adopts a microservice design, with each functional module encapsulated as an independent service. Services communicate with each other through message queues. The interface of the multi-source heterogeneous data fusion engine supports standard data formats. Point cloud data supports the LiDAR data exchange format, image data supports the geotagged image file format, and coordinate system transformation parameters support the geodetic parameter format. The multi-source heterogeneous data fusion engine is tested and verified using a standard dataset, which contains LiDAR point clouds and multispectral images from different scenarios. Test indicators include fusion accuracy, processing speed, and resource consumption. Fusion accuracy is evaluated by comparing the fusion results with the true values. Processing speed measures the amount of data processed per unit time. Resource consumption records the usage of the central processing unit and memory. The test results are used to optimize the algorithm parameters and system configuration of the multi-source heterogeneous data fusion engine to ensure its performance stability under various operating conditions.

[0031] The multi-source heterogeneous data fusion engine supports both local and cloud platform deployment. Local deployment is suitable for scenarios with high real-time requirements, while cloud platform deployment provides elastic computing resources. The maintenance of the multi-source heterogeneous data fusion engine includes regularly updating the fusion algorithm and calibrating sensor parameters. The update process follows version control specifications, recording the content and impact of each change. The user interface of the multi-source heterogeneous data fusion engine provides visualization tools that display the fusion process and data quality indicators, facilitating monitoring and debugging by operators.

[0032] Example 3: The 3D topology reconstruction module includes a component semantic segmentation unit and a spatial relationship reasoning unit. The component semantic segmentation unit identifies the physical boundaries of building exterior walls, windows, and load-bearing columns based on material reflectance data. The material reflectance data comes from multi-band spectral information collected by a multispectral sensor array. The spatial relationship reasoning unit calculates the geometric constraints of perpendicularity, parallelism, and coplanarity between components based on the physical boundaries. The geometric constraints are used to define the spatial relative positional relationships between building components. The generation process of the hierarchical building information model includes parametric template instantiation, parent-child cascade reference establishment, and version management. Parametric template instantiation abstracts building components that meet the geometric constraints. For adjustable-size template instances, the component semantic segmentation unit uses a deep learning-based point cloud segmentation network to process point clouds with material reflectivity data. The point cloud segmentation network adopts a sparse convolutional neural network architecture, and the network structure includes a feature encoding layer and a feature decoding layer. Physical boundary extraction is achieved by calculating the rate of change of the point cloud normal vector and the material reflectivity gradient. The rate of change of the normal vector is calculated through local plane fitting, and the material reflectivity gradient is obtained through the spectral differences between adjacent points. The spatial relationship reasoning unit calculates perpendicularity using the cosine of the angle between the normal vectors, parallelism using the dot product of the direction vectors, and coplanarity using the root mean square value of the distance from the point to the plane. These geometric metrics are uniformly expressed by the following formula:

[0033] Where: C represents the degree of comprehensive constraint, n represents the number of constraints, and W... i f represents the weight coefficient of the i-th constraint. i Let V1 and V2 represent the direction vectors of the component, and P represent the spatial coordinates of the component. When the comprehensive constraint degree C is lower than the set threshold, the spatial relationship reasoning unit determines that there is a topological relationship conflict between the components.

[0034] The parameterized template instantiation process selects standard component templates from the component library, which contains parameterized definitions of building elements such as beams, slabs, columns, and walls. During template instantiation, the template parameters are adjusted based on the actual point cloud data. The parameters include size, position, and orientation parameters. The size parameters are determined by fitting the point cloud bounding box. Parent-child cascading references are established through a directed acyclic graph (DAG) structure. The nodes of the DAG represent component instances, and the edges represent the spatial topological relationships between components. Version management uses snapshot differencing technology, which records the structural differences before and after each modification. The difference data is stored in JSON format.

[0035] The training process for the component semantic segmentation unit uses a pre-labeled building point cloud dataset containing tens of thousands of manually labeled component samples. The training optimizer employs an adaptive moment estimation algorithm, and the loss function combines cross-entropy loss and Dietrich coefficient loss. The parameter optimization for the spatial relationship inference unit uses constrained least squares, iteratively adjusting the geometric constraint weight coefficients W. i This ensures that the overall constraint degree C best reflects the stability requirements of the actual building structure.

[0036] The hierarchical building information model (BIM) uses a tree-like hierarchical storage structure. The root node represents the entire building, intermediate nodes represent floors or functional areas, and leaf nodes represent individual components. Each node stores the component's geometric data, material properties, and topological relationships. Geometric data is stored using boundary representation, and material properties are associated with a material library using material identifiers. Parametric template instantiation supports dynamic updates. When point cloud data changes, template instance parameters are automatically recalculated to maintain consistency with measured data. A loop detection mechanism is implemented during the establishment of parent-child cascaded references to prevent circular reference relationships between components. The version management backtracking function supports model state recovery at any point in time. The backtracking algorithm quickly locates historical versions based on version number timestamps. Post-processing operations for component semantic segmentation units include hole filling and noise filtering. Hole filling uses morphological closing operations, and noise filtering is based on a statistical outlier removal algorithm.

[0037] The parallel computing architecture of the spatial relationship reasoning unit utilizes a multi-core central processing unit to process multiple component pairs in parallel. The parallel computing employs a work-stealing algorithm to dynamically allocate computing tasks. The export function of the hierarchical building information model supports the Industrial Basic Class format and the Green Building Extended Markup Language format. The Industrial Basic Class format ensures compatibility with building information modeling software. The quality assessment of parametric template instantiation is achieved by comparing the fit between the instantiated model and the original point cloud. The fit is calculated using Hausdorff distance to measure the similarity between point clouds.

[0038] The user interface of the 3D topology reconstruction module provides topology visualization functionality. This visualization uses a force-directed graph algorithm to automatically lay out the component relationship diagram. Interactive editing allows users to manually adjust the automatically identified topology relationships, with the adjustment results fed back to the hierarchical building information model in real time. The incremental learning function of the component semantic segmentation unit supports online updates to the segmentation model. Incremental learning uses new labeled data to fine-tune network parameters, avoiding retraining the entire model. The constraint propagation algorithm of the spatial relationship inference unit handles topology changes, automatically updating the topology relationships of related components when the position of a single component is adjusted. The data consistency check mechanism of the hierarchical building information model verifies the consistency between geometric data and topology relationships. The consistency check verifies the integrity of each connection by traversing the component relationship diagram. The data interface between the 3D topology reconstruction module and the dynamic visualization rendering platform uses memory-mapped files for transfer, providing a high-speed data exchange channel.

[0039] The component semantic segmentation unit's real-time processing mode supports streaming point cloud data input. The real-time processing mode uses a sliding window mechanism to process continuous point cloud streams. The spatial relationship reasoning unit's multi-resolution analysis function processes point cloud data of varying precision. Multi-resolution analysis organizes point cloud detail levels through an octree structure. The hierarchical building information model's compressed storage uses a geometric compression algorithm, which reduces storage space requirements based on parallelogram prediction and entropy coding. The 3D topology reconstruction module's performance monitoring records the time consumption and memory usage of each processing stage. Performance data helps optimize system resource allocation. The error handling mechanism detects invalid input data and provides detailed error reports, guiding users to correct the input data format or quality. The 3D topology reconstruction module's calibration function periodically verifies the accuracy of geometric constraint calculations, using a standard geometric model as a benchmark reference.

[0040] See Figure 4 This visualization demonstrates the working mechanism and effect of the 3D topological relationship reconstruction module. By differentiating the point cloud distribution of key components such as building exterior walls, windows, and load-bearing columns through different gray levels, it reflects the ability of the component semantic segmentation unit to identify physical boundaries based on material reflectivity data. The dashed lines in the figure indicate the verticality constraint relationship between components, and the dotted lines show the parallelism constraint relationship. These geometric constraints together constitute the spatial relative position relationship network of building components. This visualization reflects the process of the spatial relationship reasoning unit calculating geometric constraints such as verticality, parallelism, and coplanarity, providing a theoretical basis and visual verification for establishing a parameterizable and editable hierarchical building information model.

[0041] Example 4: The dynamic visualization rendering platform includes a real-time computing unit and an interactive operation unit. The real-time computing unit dynamically loads building component mesh data of different precision according to the viewing angle. The viewing angle is defined by the user's camera position and orientation. The camera position is calculated through a view matrix, which contains the camera's coordinates and orientation in the world coordinate system. When responding to spatial sectioning commands, the interactive operation unit automatically reconstructs the topological connections between components. Spatial sectioning commands are input through the user interface. The sectioning plane is defined by the normal vector and the origin coordinates. The dynamic visualization rendering platform supports operation protocols including the Open Geospatial Consortium (OGC) standard protocol and the Web Graphics Library (WBL) protocol. The OGC standard protocol publishes a tile map service for the Building Information Model (BIM). The tile map service segments the 3D model into image tiles with multiple levels of detail. The WBL protocol implements... The browser-based 3D spatial measurement and analysis functions include point-to-point distance calculation, surface area measurement, and volume estimation. The dynamic loading mechanism of the real-time computing unit is based on hierarchical detail technology, which switches mesh precision according to the distance between the camera and the component. The distance threshold is preset to multiple levels. Mesh data is stored in a distributed file system, which uses a hash table to index the mesh files. The hash key is generated based on the component identifier and the level of detail. The rendering pipeline of the real-time computing unit adopts a multi-threaded architecture, which separates the geometry processing and rasterization stages. The geometry processing thread calculates vertex transformations, and the rasterization thread processes pixel shading. The sectioning algorithm of the interactive operation unit uses a boundary representation model, which stores the topological information of the component's face, edge, and point. The sectioning operation detects component faces that intersect with the sectioning plane.

[0042] The implementation of the Open Geospatial Consortium (OPCAC) standard protocol utilizes web map services and web map tile services. Web map services provide dynamically generated map images, while web map tile services pre-generate static tiles. A tile pyramid structure organizes tiles at multiple scaling levels, with the bottom layer corresponding to full-range low-resolution data and the top layer corresponding to high-resolution details. The web graphics library protocol leverages the WebGL application programming interface (API), which provides low-level graphics operations, including buffer object management and shader programming. Browser-side 3D spatial measurement is achieved through raycasting, which generates rays in world space from screen coordinates and detects the intersections of the rays with the model. The real-time computing unit's memory management employs a Least Recently Used (LRU) strategy, which caches frequently used grid data. The cache size is dynamically adjusted based on available memory. The interactive operation unit's topology reconstruction process uses an incremental update algorithm, which only modifies the connections of components affected by sectioning, reducing computational overhead. The Open Geospatial Alliance standard protocol's data format supports key-value tagged image file format and portable network graphics format. The key-value tagged image file format stores georeferenced information, while the portable network graphics format provides lossless compression. The measurement accuracy of the network graphics library protocol is optimized through a depth buffer, which stores the depth value of each pixel. The depth value is used to accurately calculate spatial coordinates.

[0043] The user interface of the dynamic visualization rendering platform integrates 3D view controls, which support zooming, panning, and rotation operations. Operation responses are handled through event listeners. The real-time computing unit uses geometric deformation technology to smoothly switch between different precision meshes, avoiding visual abrupt changes. The interactive operation unit highlights the section sections in its section visualization, which are rendered using specific colors that can be customized. The coordinate reference system of the Open Geospatial Alliance standard protocol uses the EPO coding, which uniquely identifies the coordinate system. The performance analysis tool of the Web Graphics Library protocol monitors the rendering frame rate, which is displayed on the user interface to assist in debugging. The mesh compression algorithm of the real-time computing unit uses vertex buffer objects, which reduce data transmission between the central processing unit and the graphics processing unit. The undo / redo function of the interactive operation unit records the section operation history, which is stored in a stack data structure. The Open Geospatial Alliance standard protocol uses the Hypertext Transfer Protocol (HTTP) for tile acquisition requests, which includes tile row and column numbers and zoom level parameters. The Web Graphics Library protocol implements component highlighting through point-and-click interaction, which is achieved through color coding or identifier buffers.

[0044] The dynamic visualization rendering platform's multi-view layout supports simultaneous display of multiple perspectives. The multi-view layout uses framebuffer objects to render each view. Occlusion culling in the real-time computing unit optimizes rendering performance. Occlusion culling uses a hierarchical depth buffer to detect invisible components. The interactive operation unit's section plane editing allows dynamic adjustment of position and orientation. Editing operations are achieved by dragging control points. The Open Geospatial Alliance standard protocol's style description uses a style layer descriptor, which defines the visual style of tiles. The network graphics library protocol's virtual reality mode supports head-mounted displays, and the virtual reality mode uses stereoscopic rendering technology.

[0045] The real-time computing unit's batch processing mode pre-computes meshes at all levels of detail and is used for offline scene preparation. The interactive operation unit's sectioning animation generates smooth transitions and interpolates the sectioning plane position over time. The Open Geospatial Consortium (OGC) standard protocol's caching mechanism stores commonly used tiles, reducing the number of server requests. The Web Graphics Library (WDL) protocol's touchscreen adaptation optimizes mobile device interaction, using gesture recognition algorithms. The dynamic visualization rendering platform's error handling mechanism captures rendering anomalies, logs errors, and restores the system to a safe state. The real-time computing unit's mesh simplification algorithm reduces the number of vertices and is based on edge-folding operations. The interactive operation unit's topology verification ensures model consistency after sectioning and checks the integrity of component connections. The OGC standard protocol's metadata query returns layer information using the Web Feature Service (WLS) protocol. The WDL protocol's print output generates high-quality images using a high-resolution off-screen buffer. See Table 1 for the parameter configurations at different levels of detail in the tile map service.

[0046] Table 1: Detailed parameters of tile map service level

[0047] The detail level switching of the real-time computing unit is based on the geometric error threshold in the table. The geometric error threshold compares screen space error with tolerance. Tile resolution determines the image quality of each tile. The data size range affects network transmission time. The sectioning response time of the interactive operation unit is optimized to the millisecond level. The sectioning response time is achieved through pre-computed topology. The tile encoding of the Open Geospatial Alliance standard protocol uses Z-order curves. Z-order curves optimize the tile storage order. The measurement tool calibration of the Network Graphics Library protocol uses a real-world scale. The real-world scale is obtained from the model metadata.

[0048] The network module of the dynamic visualization rendering platform handles protocol communication, using asynchronous requests to avoid interface freezing. The real-time computing unit manages GPU memory, allocating buffer objects and monitoring usage to prevent overflows. The interactive operation unit renders section planes using a stencil buffer, which marks section regions. The Open Geospatial Alliance standard protocol supports layer overlay of multiple services using alpha blending. The Network Graphics Library protocol handles user interaction events using an event loop, distributing mouse and keyboard input. The real-time computing unit calculates shadows using shadow mapping, rendering depth maps from the light source's perspective. The interactive operation unit saves its sectioning history as a project file containing section plane parameters. The Open Geospatial Alliance standard protocol's time dimension supports dynamic data, filtering tiles based on timestamps. The Network Graphics Library protocol's network worker handles complex calculations, running in a background thread.

[0049] The installer for the dynamic visualization rendering platform includes a configuration wizard that sets graphics parameters and protocol endpoints. The debug mode of the real-time computing unit displays rendering statistics, including the number of triangles and drawing calls. The interactive operation unit's section export generates a standard exchange format, which is an industry-standard format. The security mechanism of the Open Geospatial Consortium (OCC) standard protocol uses authentication tokens to control access permissions. Cross-browser testing of the Web Graphics Library (BLL) protocol ensures compatibility, covering major browser versions. The geometric data of the real-time computing unit is organized using an index buffer, which shares vertex data to reduce memory usage. The interactive operation unit's section face interaction allows rotation and scaling, represented using quaternions. The OCC standard protocol's projection transformation uses a reprojection algorithm to transform different coordinate systems. Mobile optimizations of the BLL protocol reduce power consumption and dynamically adjust rendering quality.

[0050] The dynamic visualization rendering platform's documentation system provides help guides explaining operation steps and protocol configurations. Real-time monitoring in the real-time computing unit adjusts rendering parameters based on frame rate feedback. The interactive operation unit's sectioning collaboration supports multi-user editing using operation transformation algorithms. The Open Geospatial Alliance standard protocol's streaming progressive tile loading prioritizes loading visible areas. The Web Graphics Library protocol's plug-in architecture extends functionality, allowing third-party module integration. The real-time computing unit's grid update handles dynamic model changes, using differential encoding to reduce data volume. The interactive operation unit's sectioning templates save commonly used sectioning configurations, enabling rapid application of predefined settings. The Open Geospatial Alliance standard protocol's metadata standard uses the Dublin Core Metadata Initiative, defining descriptive fields. Performance benchmarks for the Web Graphics Library protocol compare different hardware, generating detailed reports.

[0051] The user training for the dynamic visualization rendering platform simulates a typical workflow. The training includes interactive tutorials, a fault recovery mechanism for the real-time computing unit (RTU) that backs up the rendering state, and a restart mechanism from a checkpoint. The interactive operation unit's section analysis generates statistical reports listing the number of affected components. Extended support for the Open Geospatial Alliance standard protocol allows for custom coordinate systems defined via parameter files. Code obfuscation protects intellectual property using the Network Graphics Library protocol, renaming variables and functions. The RTU's optimization techniques utilize instantiated rendering, enabling batch rendering of identical components. The interactive operation unit's section replay reproduces the operation sequence, displaying steps chronologically. Open Geospatial... The quality metrics of the Alliance Standard Protocol evaluate service performance, including response time and availability. It also enhances the integration of open-source libraries into the Network Graphics Library Protocol, such as the Three.js framework. Maintenance tools for the dynamic visualization rendering platform clean up cached data, deleting expired tiles and temporary files. Hardware acceleration for the real-time computing unit leverages the characteristics of the graphics processing unit, using computational shaders. Section verification for the interactive operation unit checks model integrity, ensuring no isolated components remain after sectioning. Version management for the Open Geospatial Alliance Standard Protocol handles protocol updates and supports backward compatibility. Automated testing of the Network Graphics Library Protocol verifies functional correctness, using simulated user interaction.

[0052] Example 5: The IoT-based intelligent analysis and visualization system for surveying and mapping information includes a building code verification module. This module logically matches a hierarchical building information model with current building design code clauses. These clauses are sourced from a national standards database, which is regularly updated to reflect the latest code versions. A 3D topology reconstruction module automatically marks spatial topology conflict points that violate the code based on the matching results. These conflicts include insufficient fire separation distances, inadequate evacuation route widths, and non-compliant structural component dimensions. The logical matching process of the building code verification module employs rule engine technology. The rule engine parses the conditional statements in the code clauses, converting them into computer-executable logic. The expression uses Extensible Markup Language (XML) format for the digital storage of current building design code clauses. XML files structurally store code clauses, chapter relationships, and applicable conditions. The logical matching process traverses each building component in the hierarchical building information model, checking whether the component attributes meet the requirements of the relevant code clauses. Component attributes include geometric dimensions, material properties, and spatial location information. The rule engine employs a forward chain reasoning mechanism, deriving conclusions from known facts. The known facts are the attribute data of the building components, and the conclusions are the compliance determination results. Spatial topological conflict points are marked using 3D visualization technology, which renders conflict points as red warning markers. These warning markers contain information on the conflict type and severity.

[0053] The building code verification module's code clause library contains standard documents such as the Code for Fire Protection Design of Buildings, the Code for Structural Loads of Buildings, and the General Code for Design of Civil Buildings. These standard documents are stored categorized by professional field. The clause parser processes the code clauses written in natural language, using semantic analysis techniques to extract key parameters and constraints. These semantic analysis techniques include dependency parsing and named entity recognition. The logical matching reasoning process adopts formalism, which defines concepts, attributes, and relationships within the building domain, supporting complex code logic expressions. An automatic conflict point correction suggestion generation module provides modification schemes based on parametric adjustment principles. To minimize design changes while meeting specifications, the conflict detection algorithm in the 3D topology reconstruction module calculates the deviation between the actual design value and the specification requirements. When the deviation exceeds the allowable tolerance, it is marked as a conflict. The verification report generator in the building code verification module creates a detailed compliance report. The compliance report is output in portable document format, which includes violation entries, involved components, and rectification suggestions. The performance optimization of the rule engine uses the Rete algorithm, which improves matching efficiency by sharing rule network nodes. The rule network nodes store some matching results. The version management of specification clauses handles standard updates and records the effective date and scope of application of the clauses to ensure the timeliness of the verification results.

[0054] The building code verification module's batch processing mode supports automatic verification of the entire model. In batch processing mode, all relevant code clauses are checked in a preset order. The interactive verification mode allows users to selectively check specific clauses. In interactive verification mode, the verification scope and depth can be selected through a graphical interface. Spatial distribution analysis of conflict points identifies areas with concentrated design problems. Spatial distribution analysis uses a clustering algorithm to group adjacent conflict points. The clustering algorithm is based on Euclidean distance. Cross-reference processing of code clauses enables collaborative verification of related clauses. Cross-reference analysis establishes logical dependencies between clauses.

[0055] The hierarchical building information model's data extraction interface reads component parameters and supports industrial basic class formats and green building extended markup language formats. The building code verification module's verification log records detailed operation processes, including verification time, user ID, relevant clause numbers, and verification results. The rule engine's debugging tool provides rule testing functions, allowing step-by-step execution of the rule matching process and viewing intermediate inference results. The conflict point historical tracking function records the rectification process and compares the conflict point status changes at different time points. The building code verification module's access control manages access to the code library, distinguishing between administrator and ordinary user permissions. Administrators maintain the code library content. The verification result visualization comparison function supports multi-scheme comparison, displaying the conflict point distribution of different design schemes side-by-side. The code clause applicability check filters irrelevant clauses. The applicability check is based on building type, scale, and functional characteristics. The severity of conflict points is graded using a red, yellow, and green three-color system: red indicates serious violations, yellow indicates general violations, and green indicates compliance with the code.

[0056] The building code verification module features an online update mechanism that automatically acquires the latest codes, downloading standard database update packages via a secure connection. It also includes a resume function to save progress during verification interruptions, allowing verification to resume from the most recent checkpoint after a system failure. The module handles inconsistencies in code clauses through conflict detection, analyzing different clauses' requirements for the same design element. Statistical analysis of verification results generates a compliance rate index, calculated as the ratio of the number of satisfied clauses to the total number of clauses. The module's API supports third-party system integration, employing a descriptive state-transfer architecture and providing standardized data exchange formats. A mobile-optimized version supports on-site verification. Optimizes touch operation and offline working mode; the customizable verification rule function allows the addition of enterprise standards, and the custom function can create a private clause library through the rule editor; the automatic sorting function of conflict points arranges them by priority, and the automatic sorting function takes into account safety impact and rectification costs; the case library of the building code verification module stores typical violation cases, and the case library provides solution references for similar problems; the quality control of the verification process checks the data integrity, and the quality control verifies the geometric consistency and information completeness of the input model; the intelligent recommendation function of code clauses prompts relevant entries, and the intelligent recommendation function is based on design features and historical verification records; the visual navigation of conflict points enables quick location, and the visual navigation automatically adjusts the view focus by clicking on the conflict mark.

[0057] The building code verification module supports concurrent verification through multi-user collaboration, with collaborative work coordinated through operation locks and change merging mechanisms. The verification result export function supports multiple formats, including Extensible Markup Language (XML), JavaScript Object Notation (JavaScript Object Notation), and comma-separated value files. The semantic search function for code clauses allows for quick clause location, supporting keyword retrieval and natural language queries. The module tracks and manages the rectification progress of conflict points, recording the responsible person, time nodes, and completion status. Verification templates in the building code verification module save frequently used settings, including clause selection, parameter configuration, and output format settings. Automated verification tasks support scheduled execution, processing batch models during system idle periods. The difference comparison function for code clauses analyzes the impact of standard updates, highlighting content changes between versions. The correlation analysis of conflict points identifies potential problem chains. The system uncovers causal relationships between different conflict points; the performance monitoring module of the building code verification module records verification efficiency and statistically analyzes the verification time and memory consumption of each clause; the user interface features an accessible design to meet special needs, supporting screen readers and keyboard navigation; multilingual support for code clauses adapts to international requirements, storing clause texts in different languages; a trend analysis function for conflict points predicts design quality, building a quality change model based on historical data; a backup and recovery mechanism for the building code verification module protects data security, periodically archiving the code library and verification records; the verification algorithm uses standard test cases for validation, including known compliant and non-compliant design models; the authoritative identifier for code clauses indicates the source standard, displaying the standard number and issuing organization information; intelligent filtering of conflict points hides resolved items, and the intelligent filter dynamically updates the displayed content based on the rectification status.

[0058] The cloud service deployment of the building code verification module provides elastic computing, automatically allocating resources based on the verification scale; real-time preview of the verification process displays the ongoing results, and the real-time preview function dynamically updates the verification status of completed clauses; links to related documents for the code clauses provide detailed evidence, pointing to the original standard text and technical specifications; batch processing of conflict points supports group operations, allowing the batch processing function to apply the same rectification plan to multiple conflict points simultaneously; the audit logs of the building code verification module meet compliance requirements, recording all code library modifications and verification operations; digital signatures of the verification results ensure the authenticity of the report, and the digital signature uses public key infrastructure technology to prevent [further issues]. The system includes: 1) Filtering applicable building types to optimize the verification scope of code clauses; 2) Visualizing conflict points using heatmaps to display concentrated areas, with color depth representing conflict point density; 3) Intelligent reminders from the building code verification module to notify users of standard updates, monitoring updates from the standard issuing agency; 4) Visualized configuration to simplify the verification process, allowing users to set the verification flow through drag-and-drop; 5) Automatic classification of code clauses to optimize retrieval efficiency, using machine learning algorithms to identify clause themes; and 6) A conflict point rectification suggestion library storing expert solutions and standard solutions to common problems.

[0059] The building code verification module features an interactive learning function that optimizes verification rules, adjusting rule parameters based on user feedback. Customizable verification report templates meet institutional needs, allowing for the addition of institutional identifiers and specific formatting requirements. Timeliness checks on code clauses prevent the use of outdated standards, comparing model design dates with standard effective dates. Risk assessments of conflict points calculate potential impacts, combining code requirements with actual usage conditions. The module supports multiple standards for handling local codes and managing special code requirements from different regions. Automated testing during the verification process ensures system reliability, simulating various violation scenarios to verify detection capabilities. A visual relationship graph of code clause associations displays clause references using node connections. The conflict point rectification plan generation function manages project progress, outputting task allocation and schedules.

[0060] The building code verification module uses data encryption to protect sensitive information, employing advanced encryption standards for data transmission and storage. Its comparative analysis function assesses the effectiveness of improvements and overlays historical verification results to display rectification progress. Intelligent interpretation of code clauses provides explanations, analyzing the intent and applicable conditions of each clause. A conflict warning function identifies potential problems in advance, predicting possible violations based on design patterns. The module's system integration framework supports extended development, defining standard interfaces and data structures. Optimization algorithms improve processing speed, using parallel computing and indexing techniques to accelerate data access. Change impact analysis assesses updated standards, identifying affected design elements and conflict points. A conflict point knowledge management solution accumulates information, building a searchable case knowledge base.

[0061] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0062] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A mapping information intelligent analysis and visualization system based on the Internet of Things, characterized in that, This includes a building space data acquisition terminal, a multi-source heterogeneous data fusion engine, a 3D topology reconstruction module, and a dynamic visualization rendering platform; The building space data acquisition terminal acquires the point cloud of geometric features on the surface of the building structure and environmental parameters in real time through distributed Internet of Things nodes. The multi-source heterogeneous data fusion engine performs spatiotemporal alignment processing on geometric feature point clouds to generate a topological skeleton of all building elements with an absolute coordinate system. The 3D topology reconstruction module extracts the spatial constraint relationships of building components based on the topological skeleton and establishes a hierarchical building information model that can be parametrically edited. The dynamic visualization rendering platform generates multi-scale spatial indexes based on hierarchical building information models, driving real-time lighting effects and interactive cross-sectional analysis.

2. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 1, characterized in that, The building space data acquisition terminal includes a lidar array and a multispectral sensor group; The lidar array captures the three-dimensional coordinate sequence of the building facade with millimeter-level precision using a non-contact scanning method; The multispectral sensor array synchronously collects reflectivity data of building surface materials and establishes a pixel-level mapping relationship with the three-dimensional coordinate sequence.

3. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 2, characterized in that, The operations performed by the multi-source heterogeneous data fusion engine include: Motion distortion compensation is performed on the discrete point cloud data acquired by the lidar array according to the scanning time sequence; The feature descriptor matching algorithm is used to eliminate the observation parallax between the multispectral sensor group and the lidar array; By using ground control points to perform coordinate system transformation on the compensated point cloud data, a spatial reference frame conforming to international surveying and mapping standards is generated.

4. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 3, characterized in that, The 3D topology reconstruction module includes a component semantic segmentation unit and a spatial relationship reasoning unit; The component semantic segmentation unit identifies the physical boundaries of building exterior walls, windows, and load-bearing columns based on material reflectivity data; The spatial relationship reasoning unit calculates the geometric constraints of perpendicularity, parallelism, and coplanarity between components based on physical boundaries.

5. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 4, characterized in that, The generation process of a hierarchical building information model includes: Abstract building components that meet geometric constraints into parametric template instances; Establish parent-child cascading references between template instances based on the spatial topological relationships between components; Versioning of cascading reference structures supports tracing historical modifications.

6. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 5, characterized in that, The dynamic visualization rendering platform includes a real-time computing unit and an interactive operation unit; The real-time computing unit dynamically loads building component mesh data of varying precision based on the viewing angle. When the interactive operation unit responds to a spatial sectioning command, it automatically reconstructs the topological connection relationships between components.

7. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 6, characterized in that, The working modes of the building space data acquisition terminal include: The overall building deformation monitoring data is updated periodically under a fixed monitoring mode; In mobile mapping mode, the coverage of indoor spatial point cloud is incrementally expanded through synchronous positioning and mapping technology.

8. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 7, characterized in that, The multi-source heterogeneous data fusion engine also includes a data quality assessment unit; The data quality assessment unit generates data credibility indicators based on point cloud density, spectral resolution, and coordinate system transformation residuals. The 3D topology reconstruction module dynamically adjusts the decision threshold for semantic segmentation of components based on the credibility index.

9. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 8, characterized in that, The dynamic visualization rendering platform supports the following operation protocols: Publish tile map services for building information models using the Open Geospatial Consortium standard protocol; This implementation uses a network graphics library protocol to achieve 3D spatial measurement and analysis functions on the browser side.

10. The Internet of Things-based intelligent analysis and visualization system for surveying and mapping information according to claim 9, characterized in that, The system also includes a building code verification module; The building code verification module logically matches the hierarchical building information model with the current building design code clauses; The 3D topology reconstruction module automatically marks the points of spatial topology conflict that violate the specifications based on the matching results.

Images