Construction of a three-dimensional simulation model of a construction enclosure

Abstract

Embodiments of the present disclosure disclose a construction enclosure three-dimensional simulation model construction method, device and equipment. A specific embodiment of the method comprises: acquiring multi-source survey data, vectorized data and construction design drawings; performing model construction processing on construction oblique images to obtain a real scene texture three-dimensional model; generating aligned data, an aligned three-dimensional model and aligned line data; performing fusion modeling processing on the aligned data and the construction oblique images to obtain a terrain model; generating a power transmission line model; performing facility and line modeling processing on the construction design drawings to obtain a facility model and a to-be-built line model; performing multi-source integration processing on the terrain model, the power transmission line model, the to-be-built line model and the facility model to obtain a scene integrated model; and performing enclosure configuration processing on the scene integrated model to obtain an enclosure three-dimensional simulation model. The embodiment improves the engineering usability of the three-dimensional simulation model, enabling it to play a substantial role in construction safety checking.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and specifically to a method, apparatus, and equipment for constructing a three-dimensional simulation model of a construction site enclosure. Background Technology

[0002] With the continuous expansion of power grid construction, the construction of new transmission lines inevitably requires crossing complex scenarios such as existing transmission lines, communication lines, important highways, railways, and rivers during the stringing process. To reduce power outages and traffic disruptions caused by these crossings, and to ensure safety during construction, it is typically necessary to erect crossing frames or use protective netting at the crossing points. A 3D simulation model of construction netting is a technology that uses computer 3D modeling to accurately reproduce the construction site and safety protection facilities in a virtual environment. By constructing this comprehensive model, engineers can simulate key operational processes such as scaffolding erection, insulation netting installation, and conductor traction in advance within a virtual environment, accurately calculate safety distances, and thus optimize construction plans, improving construction safety and efficiency. Currently, when constructing a 3D simulation model of a construction site enclosure, the common approach is to rely solely on UAV oblique photography technology to acquire image data of the construction site. Based on this single data source, a 3D model is reconstructed. Then, construction technicians manually arrange the enclosure structure and temporary construction facilities in the 3D scene according to 2D design drawings and engineering experience to obtain the 3D model of the construction enclosure.

[0003] However, when constructing a three-dimensional simulation model of the construction site enclosure using the above method, the following technical problems often arise: Relying solely on UAV oblique photography to acquire construction site image data, and then using this single data source for 3D model reconstruction, followed by manual placement of the safety net structure and temporary facilities in the 3D scene by construction technicians based on 2D design drawings and engineering experience, yields a 3D model of the construction safety net. While this method can reproduce surface texture, the small linear objects such as conductors and insulator strings are prone to distortion, breakage, or geometric discrepancies during reconstruction. Furthermore, it cannot effectively penetrate vegetation to obtain accurate ground elevation (DEM), resulting in insufficient model accuracy. Simultaneously, the lack of parameter-driven mechanisms means that safety net design parameters such as cable tension, sag curves, and anchor point coordinates are not incorporated into the 3D scene. When crossing spans or changes in construction conditions occur, the safety net geometry cannot be automatically reconstructed, requiring manual adjustments to the model's position and orientation. This significantly increases the time required for multi-scheme comparison and makes it difficult to cover extreme working conditions. Consequently, the simulation model degenerates into a static visual composite, unable to perform reproducible, verifiable, and recalculated engineering simulations of safety net placement and cable laying conditions while maintaining spatial accuracy. This results in low engineering usability of the constructed 3D simulation model.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not form prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide a method, apparatus, electronic device, and computer-readable medium for constructing a three-dimensional simulation model of a construction site enclosure to solve one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a method for constructing a three-dimensional simulation model of a construction site enclosure. The method includes: acquiring multi-source survey data of the construction site, vectorized data of the original route, and construction design drawings, wherein the multi-source survey data includes point cloud data and oblique images of the construction site; performing three-dimensional real-scene model construction processing on the oblique images of the construction site to obtain a real-scene textured three-dimensional model; generating aligned point cloud data, aligned real-scene textured three-dimensional model, and aligned route data based on the point cloud data, the real-scene textured three-dimensional model, and the vectorized data of the original route; and performing terrain and texture fusion modeling processing on the aligned point cloud data and the oblique images of the construction site to obtain a terrain three-dimensional model. A three-dimensional model is generated based on the aforementioned terrain 3D model, aligned real-scene texture 3D model, aligned point cloud data, and aligned line data. The construction design drawings are then used to model temporary construction facilities and the transmission line to be built, resulting in a 3D model of the temporary construction facilities and the transmission line to be built. Multi-source model scene integration processing is performed on the aforementioned terrain 3D model, the aforementioned transmission line and crossing 3D model, the aforementioned transmission line to be built model, and the aforementioned temporary construction facility 3D model to obtain a scene integration model. Based on preset construction netting design parameters and preset construction scene configuration information, the scene integration model is parametrically configured to obtain a 3D simulation model of the netting.

[0008] Secondly, some embodiments of this disclosure provide a three-dimensional simulation model construction device for construction site fencing. The device includes: an acquisition unit configured to acquire multi-source survey data of the construction site, vectorized data of the original route, and construction design drawings, wherein the multi-source survey data of the construction site includes point cloud data of the construction site and oblique imagery of the construction site; a first processing unit configured to perform three-dimensional real-scene model construction processing on the oblique imagery of the construction site to obtain a real-scene texture three-dimensional model; a first generation unit configured to generate aligned point cloud data, aligned real-scene texture three-dimensional model, and aligned route data based on the point cloud data of the construction site, the real-scene texture three-dimensional model, and the vectorized data of the original route; and a second processing unit configured to perform terrain and texture fusion modeling processing on the aligned point cloud data and the oblique imagery of the construction site to obtain a terrain three-dimensional model. The system comprises the following components: a first generation unit, a second generation unit, configured to generate a 3D model of the transmission line and its crossings based on the aforementioned 3D terrain model, the aforementioned aligned real-world texture 3D model, the aforementioned aligned point cloud data, and the aforementioned aligned line data; a third processing unit, configured to model the aforementioned construction design drawings for temporary construction facilities and the transmission line to be built, obtaining a 3D model of the temporary construction facilities and a model of the transmission line to be built; a scene integration unit, configured to perform multi-source model scene integration processing on the aforementioned 3D terrain model, the aforementioned 3D model of the transmission line and its crossings, the aforementioned model of the transmission line to be built, and the aforementioned 3D model of the temporary construction facilities, obtaining a scene integration model; and a parameterized configuration unit, configured to perform parameterized configuration processing on the aforementioned scene integration model based on preset construction netting design parameters and preset construction scene configuration information, obtaining a 3D simulation model of the netting.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0011] The above-described embodiments of this disclosure have the following beneficial effects: the three-dimensional simulation model construction method for construction safety netting according to some embodiments of this disclosure improves the engineering usability of the constructed three-dimensional simulation model and ensures that the three-dimensional simulation model plays a substantial role in construction safety verification. Specifically, the reason why the constructed three-dimensional simulation model has low engineering usability and cannot play a substantial role in construction safety verification is that it relies solely on UAV oblique photography technology to generate a textured three-dimensional model by aerial reconstruction of the site. Although this method can restore the surface texture, since power facilities such as conductors and insulator strings are small linear objects, they are prone to tearing, breakage, or geometric distortion during reconstruction, and cannot effectively penetrate vegetation to obtain the true ground elevation (DEM), resulting in insufficient model accuracy. At the same time, since the design parameters of the netting, such as the tension of the catenary cable, the sag curve, and the coordinates of the anchor points, are not embedded into the three-dimensional scene, there is a lack of parameter-driven mechanism. When crossing spans or changes in construction conditions, the geometry of the netting cannot be automatically reconstructed, and manual adjustment of the model's position and attitude is required, which leads to a significant increase in the time required for multi-scheme comparison and makes it difficult to cover extreme working conditions. This results in the simulation model degenerating into a static visual composite, unable to perform reproducible, verifiable, and recalculated engineering design-level deductions of the netting layout and line laying conditions while ensuring spatial accuracy. Consequently, the constructed 3D simulation model has low engineering usability and cannot play a substantial role in construction safety verification. Based on this, the method for constructing a 3D simulation model for construction netting in some embodiments of this disclosure first acquires multi-source survey data of the construction site, vectorized data of the original line, and construction design drawings, wherein the multi-source survey data of the construction site includes point cloud data and oblique images of the construction site. Next, the oblique images of the construction site are processed to construct a 3D real-scene model, resulting in a real-scene textured 3D model. Thus, a real-scene 3D model with surface texture can be generated based on the oblique images of the construction site. Then, based on the point cloud data of the construction site, the real-scene textured 3D model, and the original line vectorized data, aligned point cloud data, aligned real-scene textured 3D model, and aligned line data are generated. Therefore, the point cloud data of the construction site from LiDAR, the 3D model of the real-scene texture, and the original vectorized data of the route can be unified under the same engineering coordinate system and accurately aligned spatially. This ensures that the multi-source data are consistent in spatial position, providing a data foundation for subsequent high-precision modeling with a unified coordinate reference and accurate matching of geometry and texture. It solves the problem of inconsistent coordinate systems and spatial offsets between multi-source heterogeneous data that prevent fusion modeling. Subsequently, the aligned point cloud data and the oblique image of the construction site are subjected to terrain and texture fusion modeling processing to obtain a 3D terrain model.Therefore, accurate ground elevation data can be generated from the aligned point cloud data to construct a raster terrain grid. A digital orthophoto map generated from oblique imagery of the construction site is then mapped onto the terrain grid surface as texture, generating a 3D terrain model that contains both accurate ground elevation information and realistic surface texture. This solves the problem of insufficient terrain model accuracy caused by relying solely on oblique photography, which cannot effectively penetrate vegetation to obtain the true ground elevation (DEM). This provides an accurate terrain reference surface for the spatial positioning of subsequent power line facilities and fencing structures. Furthermore, based on the aforementioned 3D terrain model, the aligned real-scene texture 3D model, the aligned point cloud data, and the aligned power line data, a 3D model of the transmission line and its crossings is generated. Therefore, based on the tower locations, conductor suspension point coordinates, and conductor sag parameters in the aligned line data, tower point clouds and conductor point clouds can be extracted from the aligned point cloud data. Tower models with hardware details are generated through parametric tower modeling, and smooth conductor and ground wire models are generated through catenary fitting. Simultaneously, road surface point clouds are extracted from the point cloud data to generate a road surface model, water surface points are extracted to generate a river model, and other overhead line points are extracted to generate an overhead line model. This forms a complete set of 3D models including existing transmission lines and all crossings, solving the problem of easily distorted, broken, or geometrically distorted small linear objects such as conductors and insulator strings during oblique photogrammetry reconstruction, thus ensuring the geometric accuracy of transmission line facilities. Next, the construction design drawings are processed to model temporary construction facilities and the transmission line to be built, resulting in 3D models of the temporary construction facilities and the transmission line to be built. Therefore, a model of the transmission line to be built can be generated based on the design drawings, and a model of the protective facilities (including catenary models, protective net models, auxiliary crossarm models, and crossing frame models) can be generated based on the netting scheme design drawings. Furthermore, models of construction machinery such as tensioners, traction machines, and cranes can be generated based on construction machinery parameter information, forming a complete 3D model of the temporary construction facilities and the transmission line to be built. This provides model inputs for the new line and temporary facilities for subsequent scene integration, solving the problem that the temporary construction facilities and the new line lack 3D models and cannot participate in spatial analysis in the simulation scene. Next, multi-source model scene integration processing is performed on the aforementioned terrain 3D model, the aforementioned transmission line and crossing structure 3D model, the aforementioned transmission line to be built model, and the aforementioned temporary construction facility 3D model to obtain a scene integration model. Thus, the various models generated in the above steps can be integrated together through multi-source scene integration to obtain a scene integration model. Finally, based on preset construction netting design parameters and preset construction scene configuration information, the above scene integration model undergoes netting model parameterization configuration processing to obtain a 3D simulation model of the netting.Therefore, based on preset construction netting design parameters (including crossing location coordinates, netting span, netting width, netting height, catenary tension value, catenary sag value, anchor point coordinates, etc.), the catenary cable model, protective netting model, auxiliary crossarm model, and crossing frame model can be automatically generated in the scene integration model, establishing a parametric linkage between the geometry of the netting structure and the design parameters. When the crossing span or construction conditions change, only the corresponding design parameter values ​​need to be modified, and the netting model can be automatically reconstructed. This avoids the problem of significantly increased time consumption for multi-scheme comparison and difficulty in covering extreme working conditions due to the lack of a parameter-driven mechanism and the need to manually adjust the model position and orientation when construction conditions change. It realizes rapid configuration, dynamic adjustment, and reproducible, verifiable, and recalculated engineering design-level simulation of netting schemes. This is because, by acquiring complementary geometric and textural information from multi-source survey data, achieving precise alignment of multi-source data through coordinate unification and spatial registration, solving the problem of obtaining ground elevation in vegetation-covered areas through terrain and texture fusion modeling, resolving geometric distortion issues of small linear objects such as conductors through automated modeling based on line data, achieving efficient organization and smooth rendering of large scenes through multi-source model scene integration, and enabling reconstruction during design changes through parametric configuration of the netting model, the 3D simulation model is upgraded from a static visual composite that can only be viewed to an engineering design-level extrapolation tool that is reproducible, verifiable, and recalculateable. While ensuring spatial accuracy, it supports systematic simulation verification of netting layout and line laying conditions, thereby improving the engineering usability of the constructed 3D simulation model and ensuring that the 3D simulation model plays a substantial role in construction safety verification. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the method for constructing a three-dimensional simulation model of a construction enclosure net according to the present disclosure; Figure 2 These are schematic diagrams of some embodiments of the three-dimensional simulation model construction device for construction netting according to this disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] Figure 1 A flowchart 100 is shown, illustrating some embodiments of a method for constructing a three-dimensional simulation model of a construction site enclosure according to the present disclosure. This method for constructing a three-dimensional simulation model of a construction site enclosure includes the following steps: Step 101: Obtain multi-source survey data of the construction site, vectorized data of the original line, and construction design drawings.

[0021] In some embodiments, the execution entity (e.g., a computing device) of the three-dimensional simulation model construction method for construction site enclosure can acquire multi-source survey data, existing line vectorization data, and construction design drawings at the construction site via wired or wireless connections. The multi-source survey data includes point cloud data and oblique imagery of the construction site. In practice, the execution entity can acquire multi-source survey data collected by drones. The point cloud data can be high-density point cloud (LAS / LAZ) data obtained by scanning the construction site with a drone's LiDAR. The existing line vectorization data can be digitized vector data generated during the design phase of the transmission line project, typically provided by the design unit. This data includes, but is not limited to, line path, tower coordinates, phase / loop information, insulator string and hardware models, and crossing records. The line path can be a projected polygonal line or curve of the conductor on the ground (representing the line direction). The tower coordinates can be the precise plane coordinates (X, Y) and elevation (Z) of the center point or corner point of the tower leg foundation of each tower. The aforementioned phase / loop information can include conductor arrangement (horizontal, vertical, triangular), loop number, and phase sequence. The aforementioned crossing records can be the mileage markers and design clearances for crossing highways, railways, and power lines as marked during the design phase. The aforementioned construction design drawings can be CAD designs of the transmission line to be built and CAD designs of the network enclosure scheme. The oblique images of the construction site are two-dimensional digital images with position and attitude information obtained by taking multi-angle photos of the construction site from vertical and oblique directions (forward, backward, left, and right) using an oblique photography camera mounted on a UAV platform. The aforementioned position and attitude information can be spatial position and attitude data synchronously recorded and calculated by a positioning and attitude measurement system (such as a combination of a global navigation satellite system receiver and an inertial measurement unit) mounted on a UAV or manned aircraft at the moment of oblique image exposure. The aforementioned spatial position data can be the three-dimensional geographic coordinates of the camera projection center (i.e., the camera lens node) in the object space coordinate system. The aforementioned three-dimensional geographic coordinates are typically represented using longitude, latitude, and elevation values ​​(B, L, H) in a geodetic coordinate system (such as WGS-84 or CGCS2000), and can also be converted to X, Y, and Z coordinate values ​​in a projected plane coordinate system. The aforementioned spatial attitude data can be the three rotation angles of the camera's optical axis in the object-space coordinate system, namely the tilt angle, pitch angle, and yaw angle used to describe the camera's shooting orientation. The aforementioned oblique images of the construction site include oblique images from various viewpoints.

[0022] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other currently known or future wireless connection methods.

[0023] Step 102: Perform three-dimensional real-scene model construction processing on the tilted image of the construction site to obtain a real-scene texture three-dimensional model.

[0024] In some embodiments, the aforementioned execution entity may perform three-dimensional real-scene model construction processing on the aforementioned tilted image of the construction site to obtain a real-scene texture three-dimensional model.

[0025] In some optional implementations of certain embodiments, the aforementioned execution entity can perform three-dimensional real-scene model construction processing on the aforementioned tilted image of the construction site through the following steps to obtain a real-scene texture three-dimensional model: The first step involves aerial triangulation of the aforementioned oblique images of the construction site to obtain image spatial attitude information. In practice, the executing entity first extracts image feature points (e.g., scale-invariant feature transform feature points or acceleration-robust feature points) from the oblique images of the construction site and matches corresponding feature points between different oblique images to establish connectivity between them. Then, based on the collinearity equation, a regional network adjustment model is constructed using each ray (i.e., the ray formed by the object point, the camera projection center, and its corresponding image point) as a unit. Next, the regional network adjustment model (e.g., the Gauss-Newton method) is globally optimized using a least-squares adjustment method. By iteratively adjusting the exterior orientation elements (i.e., image spatial position parameters and attitude parameters) of each oblique image, the corresponding ray ray achieves optimal intersection in the object space, yielding the exterior orientation elements of each oblique image. Finally, the obtained exterior orientation elements are determined as image spatial attitude information. The aforementioned collinearity equation represents the collinear relationship between the object point, the camera projection center, and the image point of that object point on the image. The aforementioned image spatial attitude information includes spatial position parameters and spatial attitude parameters for the oblique images. The spatial position parameters are the three-dimensional coordinates (e.g., X, Y, Z) of the camera projection center in the object space coordinate system. The spatial attitude parameters are three angular elements describing the camera's shooting attitude (e.g., yaw angle, pitch angle, roll angle). The aforementioned regional network adjustment model can be a set of nonlinear equations and their constraints, constructed based on collinearity equations, that uniformly solve for the exterior orientation elements of each image and the three-dimensional coordinates of the object point, taking the aforementioned oblique images of the construction site as the overall processing object. Each oblique image from each perspective in the aforementioned construction site corresponds to one exterior orientation element of the oblique image in the aforementioned image spatial attitude information.

[0026] The second step involves constructing a dense point cloud from the aforementioned image spatial pose information to the oblique images of the construction site, resulting in real-world color point cloud data. In practice, the executing entity can first use the oblique images of the construction site as the processing object, and the camera's intrinsic parameters and exterior orientation elements as spatial constraints. A multi-view stereo vision algorithm is employed to perform the following pixel-by-pixel dense matching steps on each viewpoint of the oblique images: First, the set of adjacent images with the largest overlap area is determined based on the exterior orientation elements of the oblique images. Then, a semi-global matching algorithm is used to perform disparity search along the epipolar direction, calculating the disparity value of each pixel in the oblique images relative to its adjacent images. Finally, based on the disparity value and the camera parameters, the physical distance from the ground feature point corresponding to each pixel to the optical center of the oblique image camera is calculated using the forward intersection principle. This distance value is then assigned to the corresponding pixel position in grayscale form to obtain a depth map. Afterward, the obtained depth maps are smoothed using a median filtering algorithm to eliminate discrete isolated noise points, resulting in optimized depth maps. Then, based on the exterior orientation elements of the corresponding images for each optimized depth map, the optimized depth maps are back-projected into a pre-established unified 3D engineering coordinate system. During the back-projection process, for different depth points projected to the same spatial location, a weighted average strategy (the weights can be determined based on the angle between the projection angle and the principal optical axis of the image; the smaller the angle, the larger the weight) is used for fusion to obtain a 3D spatial point matrix. Next, for each 3D spatial point in the above 3D spatial point matrix, based on its 3D coordinates and the exterior orientation elements of the above tilted images, the collinearity equation is used to calculate the projected pixel position of the above 3D spatial point in all visible tilted images. Then, the viewpoint with the projection angle closest to the direction of the principal optical axis of the image is selected as the optimal texture viewpoint, and the RGB color values ​​of the corresponding pixels under the optimal viewpoint are extracted and assigned to the 3D spatial point to obtain real-scene color point cloud data. Each point in the above 3D spatial point matrix contains precise 3D spatial coordinates (X, Y, Z). The aforementioned real-scene color point cloud data can contain several discrete points, each of which contains at least three-dimensional spatial coordinate data (X, Y, Z) and color data (R, G, B). The spatial point density of the aforementioned real-scene color point cloud data reaches the level of hundreds of points per square meter, containing both the geometric shape information of the construction site features and the real texture and color information based on the original image.

[0027] The third step involves constructing a triangular network from the aforementioned real-scene color point cloud data to obtain a 3D geometric white model. In practice, firstly, the execution entity extracts the 3D spatial coordinates (X, Y, Z) and normal vector information of each discrete point in the real-scene color point cloud data. Then, based on the Poisson surface reconstruction algorithm, adjacent discrete points are calculated and connected in 3D space. Using the criterion of minimizing the minimum angle of all triangles or maximizing the empty circle characteristic, the originally disordered set of discrete points is transformed into a 3D curved surface mesh composed of continuous, seamless triangular facets. Finally, by introducing terrain contour lines as constraints, it is ensured that the topology of the triangular network accurately conforms to the real physical boundaries at key terrain boundaries such as rivers, road edges, or tower bases, thereby generating a 3D geometric white model with continuous geometric form. The aforementioned 3D geometric white model can be composed of several triangular facets, each containing the 3D spatial coordinates (X, Y, Z) of three vertices and the normal vector information of each vertex, but not containing texture color information.

[0028] The fourth step involves performing texture mapping on the aforementioned 3D geometric white model based on the tilted images of the construction site, resulting in a realistic textured 3D model. In practice, firstly, the executing entity can combine the 3D spatial coordinates and normal vectors of each triangular facet or vertex in the 3D geometric white model to calculate the projection rays of each tilted image onto the 3D geometric white model. Secondly, for each triangular facet to be mapped on the 3D geometric white model, the tilted images from various perspectives in the aforementioned tilted images of the construction site are traversed. Based on the line-of-sight occlusion relationship, image resolution, and incident angle, the original tilted image with the optimal viewing angle and least occlusion for that facet is automatically selected as the texture source. Then, the RGB color pixel information of the corresponding area in the aforementioned tilted images of the construction site is extracted. Afterward, the aforementioned RGB color pixel information is mapped onto the corresponding triangular facet according to UV coordinates. Finally, a polygon boundary fusion algorithm is used to eliminate texture seams and color differences caused by stitching together multiple tilted images, resulting in a realistic textured 3D model. The aforementioned real-scene textured 3D model can be generated by constructing a 3D geometric white model based on the aforementioned real-scene color point cloud data through triangular network processing, using the aforementioned tilted image of the construction site as the texture source, and then applying real color information to each triangular facet of the 3D geometric white model through texture mapping processing, resulting in a 3D mesh model with realistic texture colors. This real-scene textured 3D model simultaneously contains 3D geometric shape information and realistic texture color information, representing an integrated digital expression of the geometric shape and visual appearance of the surface features at the construction site.

[0029] Step 103: Based on the point cloud data of the construction site, the 3D model of the real scene texture, and the vectorized data of the original line, generate aligned point cloud data, aligned 3D model of the real scene texture, and aligned line data.

[0030] In some embodiments, the aforementioned execution entity may generate aligned point cloud data, aligned real-scene texture 3D model, and aligned line data based on the aforementioned construction site point cloud data, the aforementioned real-scene texture 3D model, and the aforementioned original line vectorization data.

[0031] In some optional implementations of certain embodiments, the aforementioned execution entity may further generate aligned point cloud data, aligned real-scene texture 3D model, and aligned line data based on the aforementioned construction site point cloud data, the aforementioned real-scene texture 3D model, and the aforementioned original line vectorization data through the following steps: The first step is to denoise the aforementioned construction site point cloud data to obtain denoised point cloud data. In practice, the executing entity can use a statistical outlier filtering algorithm to calculate the distribution of points in the neighborhood of each point, removing isolated noise points (such as floating points caused by birds or multipath reflections) to denoise the construction site point cloud data, thus obtaining denoised point cloud data. The denoised point cloud data can be the point cloud data obtained after removing outliers and random error points caused by atmospheric suspended matter reflection, multipath effects, equipment electronic noise, and scanning by moving objects. Outliers can be isolated data points whose spatial distance distribution to their surrounding neighborhood points in three-dimensional space significantly deviates from the overall statistical distribution. Random error points can be data points whose measured values ​​randomly deviate from the true values ​​due to factors such as internal equipment electronic noise, vibration, or environmental interference.

[0032] The second step involves performing ground point separation processing on the denoised point cloud data to obtain ground point cloud data and non-ground point cloud data. In practice, the execution entity can use a Cloth Simulation Filter (CSF) to flip the denoised point cloud data, simulating the fall of a soft cloth. Points in the denoised point cloud data that are in contact with the cloth are identified as ground points, while points that are not in contact with the cloth are identified as non-ground points. The cloth can be a virtual flexible surface model constructed using physical simulation techniques in 3D computer graphics. The cloth consists of a finite number of discrete particles, each a data unit containing three-dimensional spatial coordinates and motion state parameters, interconnected by preset constraints. The cloth is endowed with physical properties such as rigidity, gravitational acceleration, particle mass, and time step. In the CSF algorithm, after flipping the point cloud, the cloth is allowed to fall from above onto the flipped point cloud surface under the influence of gravity. Through collision detection and position correction between cloth particles and the point cloud, the spatial position of each particle is gradually adjusted. The final shape of the cloth is used as a simulated terrain surface. After flipping the final shape of the cloth back to its original orientation, for each point in the original point cloud, its vertical distance to the simulated terrain surface is calculated. Points with a distance less than a preset threshold are classified as ground points, and the rest are classified as non-ground points.

[0033] The third step involves classifying the aforementioned non-ground point cloud data to obtain classified point cloud data. Local geometric features are then extracted from the non-ground point cloud data. In practice, firstly, the executing entity can construct a local point set within a preset radius neighborhood for each point cloud in the non-ground point cloud. Then, based on Principal Component Analysis (PCA) algorithm, the covariance matrix of the local point set is calculated, thereby solving for the geometric feature values ​​characterizing the spatial distribution of the point. Secondly, a rule-based feature threshold segmentation method is used to classify the non-ground point cloud data according to the aforementioned geometric feature values, obtaining the classification result information. Finally, based on the classification results, the non-ground point cloud data is attribute-labeled and partitioned into sets, generating classified point cloud data containing different category labels (such as vegetation, buildings, poles, power lines, etc.). The aforementioned geometric feature values ​​include, but are not limited to, linear features, planar features, scattering features, verticality features, and point cloud density features. The classification results described above can be logical category labels output after evaluating the spatial attributes of each discrete point in the non-ground point cloud using a rule-based feature threshold segmentation method. For example, based on the internationally recognized ASPRS point cloud classification standard or a custom engineering classification standard, category labels "3" to "5" can be mapped to different levels of vegetation, category label "6" to buildings, category label "15" to poles and towers, and category label "14" to power lines. These category labels can be written into the point cloud data structure as independent attribute fields. The rules described above can include, but are not limited to, classifying point clouds with verticality features greater than a preset threshold and significant linear features as pole and tower point clouds, classifying point clouds with significant planar features and vertically upward normal vectors as building points, and classifying point clouds with significant scattering features and disordered normal vectors as vegetation points. The preset threshold can be a pre-defined value used to classify the aforementioned non-ground points.

[0034] The fourth step is to identify the above-mentioned ground point cloud data, the above-mentioned non-ground point cloud data, and the above-mentioned classified point cloud data as the processed point cloud data.

[0035] The fifth step involves performing multi-source data coordinate unification and spatial registration on the processed point cloud data, the real-scene texture 3D model, and the original route vectorized data to obtain aligned point cloud data, aligned real-scene texture 3D model, and aligned route data. In practice, firstly, the executing entity can read the track file of the LiDAR scanning device corresponding to the processed point cloud data. Then, the coordinate system identifier field is extracted from the file header or metadata field of the track file. Next, the instantaneous coordinate system used by the LiDAR scanning device during data acquisition is determined based on the extracted coordinate system identifier field, and this instantaneous coordinate system is designated as the first original coordinate system. Afterward, the aerial triangulation engineering file generated after processing the oblique image of the construction site through aerial triangulation is read. Then, the definition parameters of the temporary 3D coordinate system used in the aerial triangulation are extracted from the engineering settings field of the aerial triangulation engineering file. Then, the temporary 3D coordinate system is determined based on the extracted definition parameters, and this temporary 3D coordinate system is designated as the second original coordinate system. Finally, the metadata file of the original route vectorized data is read. Then, the coordinate system field is extracted from the aforementioned metadata file. Next, based on the extracted coordinate system field, the design coordinate system used in the creation of the original line vectorized data is determined, and this design coordinate system is designated as the third original coordinate system. Then, the coordinate values ​​of at least three ground control points in the first original coordinate system and their coordinate values ​​in the engineering coordinate system are obtained. These two sets of coordinate values ​​are substituted into the seven-parameter transformation model, and the seven unknown parameters are solved using the least squares adjustment method to obtain the first coordinate transformation parameters. Then, the second and third coordinate transformation parameters are generated using the same method. Next, the three translation parameters, three rotation parameters, and one scaling parameter from the first coordinate transformation parameters are substituted into the seven-parameter transformation model to calculate the three-dimensional spatial coordinates of each point cloud in the processed point cloud data in the engineering coordinate system. This is used to perform coordinate transformation on the processed point cloud data, resulting in the coordinate-unified processed point cloud data. Next, following the same coordinate transformation method, the three-dimensional spatial coordinates of all vertices in the aforementioned real-scene texture 3D model are transformed from the aforementioned second original coordinate system to the aforementioned engineering coordinate system using the second coordinate transformation parameters, resulting in a coordinate-unified real-scene texture 3D model. Then, using the same coordinate transformation method and the aforementioned third coordinate transformation parameters, the tower location coordinates and conductor suspension point coordinates in the aforementioned original line vectorized data are transformed from the aforementioned third original coordinate system to the aforementioned engineering coordinate system, resulting in coordinate-unified line data. Finally, at least three pairs of feature points with the same name are extracted from the coordinate-unified point cloud data and the coordinate-unified real-scene texture 3D model.Next, using at least three pairs of identical feature points as initial control conditions, the coordinate-unified 3D model of the real-world texture as the reference target, and the coordinate-unified point cloud data as the source data to be registered, an iterative nearest-point algorithm is executed. After each iteration, the average distance error between the updated point cloud data and the coordinate-unified 3D model of the real-world texture is calculated. Iteration stops when the average distance error is less than a preset first distance error threshold (e.g., set to 0.05 meters) or the number of iterations reaches a preset maximum number of iterations (e.g., set to 50 times), and the currently updated point cloud data is determined as the globally registered point cloud data. Then, the 3D spatial coordinates of each tower location are extracted from the globally registered point cloud data, and the corresponding 3D spatial coordinates of each tower location are extracted from the coordinate-unified line data. The spatial deviation value of each tower in the two datasets is calculated. If the average spatial deviation of all tower locations is greater than a preset second distance error threshold (e.g., set to 0.1 meters), the execution entity uses a least-squares registration algorithm to perform local fine registration on the coordinate-unified line data, and determines the locally finely registered coordinate-unified line data as the registered line data. If the average spatial deviation is less than or equal to the second distance error threshold, the execution entity directly determines the coordinate-unified line data as the registered line data. Finally, the globally registered point cloud data is determined as the aligned point cloud data, the coordinate-unified real-scene texture 3D model is determined as the aligned real-scene texture 3D model, and the registered line data is determined as the aligned line data. The track file records the position (3D spatial coordinates) and attitude information (attitude angle) of the lidar scanning device at each scanning moment during data acquisition. The track file can be in SBET or POS format. The coordinate system identifier character records the name or code of the coordinate system used by the lidar scanning device during data acquisition. The coordinate system identifier field mentioned above can be any of WGS84, CGCS2000, or a local independent coordinate system name. The track file mentioned above can be a data file recording the spatial position and attitude of the lidar scanning equipment at each scan moment during data acquisition. The aerial triangulation engineering file mentioned above records the exterior orientation elements, camera intrinsic parameters, and three-dimensional spatial coordinates of densification points for all oblique images involved in aerial triangulation. The aerial triangulation engineering file can be in XML or JSON format. The defined parameters mentioned above include projection type, reference ellipsoid, central meridian, three-dimensional spatial coordinates of the coordinate origin, and scale factor. The metadata file mentioned above can be a file describing the data format, coordinate system, data source, and production unit of the aforementioned original route vectorized data. The coordinate system field mentioned above records the complete definition of the design coordinate system used in the production of the aforementioned original route vectorized data.The aforementioned lidar scanning equipment can be a UAV equipped with a lidar measurement system. The aforementioned engineering coordinate system can be a unified spatial coordinate reference framework for the aforementioned target construction area, using the CGCS2000 National Geodetic Coordinate System as the plane coordinate reference and the 1985 National Height Datum as the elevation reference. The aforementioned ground control points can be markers pre-set on the ground at the construction site, whose precise three-dimensional coordinates under the engineering coordinate system have been measured. These markers are artificial targets or naturally occurring features with a clearly defined center position; their coordinates under the engineering coordinate system are obtained through GNSS static measurement, with a measurement accuracy better than centimeter level. The aforementioned seven-parameter transformation model is a three-dimensional spatial coordinate transformation mathematical model, which includes three translation parameters. Three rotation parameters A scaling parameter *k* is used to achieve coordinate transformation between two different three-dimensional Cartesian coordinate systems. The three translation parameters represent the offsets of the origins of the two coordinate systems along the X, Y, and Z axes, respectively, in meters. The three rotation parameters represent the rotation angles of the coordinate axes of the two coordinate systems around the X, Y, and Z axes, respectively, in radians or seconds. The scaling parameter represents the proportional difference per unit length between the two coordinate systems and is a dimensionless quantity. This four-parameter transformation model is a two-dimensional planar coordinate transformation mathematical model that includes two translation parameters. One rotation parameter A scaling parameter k is used to achieve coordinate transformation between two two-dimensional plane coordinate systems, applicable to coordinate transformation under the same ellipsoidal reference but different projection zones or different central meridian settings. The aforementioned coordinate transformation parameters can be the mathematical parameters required to transform spatial coordinates in one coordinate system to spatial coordinates in another, including translation parameters, rotation parameters, and scaling parameters. The corresponding physical feature point for the aforementioned pair of identically named feature points is any one of the following: the center point of a tower top, the end point of a tower crossarm, the intersection of road markings, or the corner point of a building.

[0036] Step 104: Perform terrain and texture fusion modeling on the aligned point cloud data and the tilted image of the construction site to obtain a 3D terrain model.

[0037] In some embodiments, the aforementioned execution entity may perform terrain and texture fusion modeling processing on the aligned point cloud data and the aforementioned construction site tilted image to obtain a three-dimensional terrain model.

[0038] In some optional implementations of certain embodiments, the aforementioned execution entity may perform terrain and texture fusion modeling processing on the aligned point cloud data and the aforementioned oblique image of the construction site through the following steps to obtain a three-dimensional terrain model: The first step involves constructing a digital orthophoto map of the construction site from the aforementioned oblique images. In practice, the implementing entity first selects a vertically oriented image from the oblique images as the base image. Then, based on the camera exterior orientation elements and camera interior parameters obtained through aerial triangulation of the oblique images, digital differential correction is performed on the base image, resulting in individual digitally corrected images. These corrected images are then stitched together according to their geographic coordinates to form an orthophoto map covering the entire construction site. Finally, color balancing and feathering are applied to the seams of the orthophoto maps to ensure consistent color tones between adjacent images, resulting in the construction site digital orthophoto map. This construction site digital orthophoto map can be a surface image of the construction site with orthophoto geometric characteristics obtained after the aforementioned digital orthophoto map construction process. Each pixel in the construction site digital orthophoto map carries real-world geographic coordinates.

[0039] The second step involves generating ground elevation data based on the ground point cloud data included in the aligned point cloud data. In practice, firstly, the executing entity can extract the ground point cloud data from the aligned point cloud data. Then, for each ground point in the ground point cloud data, the planar coordinates (X, Y) of each ground point are associated and stored with its corresponding ground elevation value, forming discrete ground elevation data. Each ground point in the ground point cloud data contains three-dimensional spatial coordinates (X, Y, Z), where the Z coordinate value is the ground elevation value at that point's location.

[0040] The third step involves processing the aforementioned ground elevation data into a raster terrain structure to obtain a raster terrain grid. In practice, firstly, the executing entity determines the coverage area of ​​the ground elevation data on the horizontal plane. Then, the raster resolution is set (e.g., 0.5 meters). Next, according to the raster resolution, I grid points are generated along the X-axis and J grid points along the Y-axis within the coverage area, forming an I×J regularly arranged grid point array. Then, for each grid point, based on its planar coordinates, the ground elevation value at that grid point is calculated using the discrete ground elevation data and an inverse distance weighted interpolation algorithm. Finally, the planar coordinates of each grid point and its corresponding ground elevation value are combined to form the raster terrain grid. The raster terrain grid can be a terrain data matrix composed of regularly spaced grid points on the horizontal plane, with each grid point corresponding to a planar location and a ground elevation value. The raster resolution can be the horizontal spacing between adjacent grid points. The aforementioned coverage area can be the region formed by the minimum and maximum values ​​in the X-axis direction and the minimum and maximum values ​​in the Y-axis direction. I is determined as follows: First, the difference between the maximum and minimum values ​​in the X-axis direction within the aforementioned coverage area is determined as a first value. Then, the ratio of the first value to the aforementioned raster resolution is rounded down to obtain a second value. Finally, the sum of the second value and the value 1 is determined as the value of I. The method for determining J is the same as the method for determining I.

[0041] The fourth step involves texture mapping of the aforementioned raster terrain grid and the aforementioned digital orthophoto image of the construction site to obtain a 3D terrain model. In practice, for each grid point in the raster terrain grid, the executing entity can search for pixel locations with the same geographic coordinates as the grid point in the aforementioned digital orthophoto image of the construction site, and extract the RGB color values ​​of these pixel locations. Then, the extracted RGB color values ​​are assigned to the corresponding grid points in the raster terrain grid. Next, for a raster unit consisting of four adjacent grid points, the raster unit is divided into two triangular patches. The color value of each triangle patch vertex is determined by the RGB color value of its corresponding grid point, and the color values ​​inside the triangle patch are obtained through bilinear interpolation. Finally, each raster unit is traversed, and the 3D spatial coordinates and corresponding RGB color values ​​of each grid point are combined to generate a triangular mesh terrain model with realistic texture as the 3D terrain model. As an example, the spatial resolution of the digital orthophoto image of a construction site is 0.05 meters per pixel, covering an area of ​​0 to 500 meters in the X direction and 0 to 300 meters in the Y direction. The grid resolution was set to 0.5 meters, resulting in a 1001-row × 601-column terrain grid with approximately 600,000 grid points. For each grid point, the corresponding geographic coordinate pixel color value was located in the digital orthophoto map of the construction site and assigned to that grid point. Then, four adjacent grid points were combined to form a grid cell, and each grid cell was divided into two triangular patches, generating approximately 1.2 million triangular patches. These approximately 1.2 million triangular patches all carried realistic color values, forming a 3D terrain model covering the entire construction site with realistic texture. This 3D terrain model includes both elevation information and true surface color information.

[0042] Step 105: Based on the terrain 3D model, the aligned real-world texture 3D model, the aligned point cloud data, and the aligned line data mentioned above, generate a 3D model of the transmission line and the crossings.

[0043] In some embodiments, the execution entity can generate a 3D model of the transmission line and its crossings based on the 3D terrain model, the aligned real-world texture 3D model, the aligned point cloud data, and the aligned line data. The aligned point cloud data includes ground point cloud data, non-ground point cloud data, and classified point cloud data.

[0044] In some optional implementations of certain embodiments, the aforementioned execution entity may generate a three-dimensional model of the transmission line and its crossings based on the aforementioned 3D terrain model, the aforementioned aligned real-world texture 3D model, the aforementioned aligned point cloud data, and the aforementioned aligned line data through the following steps: The first step is to construct a tower model based on the aligned line data and the classified point cloud data to obtain the tower model.

[0045] The second step involves extracting conductor and ground wire point clouds from the aligned line data, resulting in a conductor and ground wire point cloud dataset. In practice, the executing entity first extracts all point clouds with conductor or ground wire category labels from the categorized point cloud data, using this as the initial conductor and ground wire point cloud data. Then, based on the coordinates and orientation of the hanging points of each conductor and ground wire in the aligned line data, the initial conductor and ground wire point cloud data is spatially filtered to remove point clouds that exceed a preset distance threshold (e.g., 2 meters) from the conductor or ground wire path in the aligned line data. Finally, the remaining point clouds after filtering are determined as the conductor and ground wire point cloud dataset. The conductor can be a metal cable used to transmit electrical energy in an overhead transmission line. The ground wire can be a metal cable erected above the conductor in an overhead transmission line for lightning protection, also known as a lightning rod.

[0046] The third step involves modeling the conductor and ground wire point cloud dataset to obtain conductor and ground wire models. In practice, firstly, the execution entity can divide the conductor and ground wire point cloud dataset into conductor and ground wire subsets based on the conductor and ground wire identifiers in the aligned line data. Then, for each phase conductor in the conductor point cloud subset, the conductor point cloud is sorted horizontally along the conductor's direction to form an ordered point sequence. Next, the catenary equation is used to fit the ordered point sequence, and the fitted catenary curve is used as the conductor centerline path. Then, using a preset conductor radius (e.g., 0.015 meters to 0.03 meters) as the cross-sectional radius, a tubular 3D mesh model is generated along the conductor centerline path, serving as the conductor model for that phase. Then, the obtained phase conductor models are combined to obtain the conductor model. Finally, the ground wire point cloud subset is processed using the same method, performing catenary fitting and tubular model generation to obtain the ground wire model. The preset conductor radius can be the physical radius of an actual overhead transmission line conductor, i.e., the radius of the conductor's cross-section, expressed in meters (m). This preset conductor radius is an inherent physical parameter of the conductor, determined by its type and specifications. In the 3D modeling process, this preset conductor radius is used as the cross-sectional radius of the tubular model to expand the fitted conductor centerline path into a 3D tubular mesh model with realistic thickness.

[0047] The fourth step involves extracting road surface point clouds from the classified point cloud data based on the aforementioned 3D terrain model, resulting in a road surface point cloud dataset. In practice, firstly, the executing entity extracts surface elevation information from the 3D terrain model, identifying areas with continuous elevation values ​​and gentle slope changes, and designating these areas as potential road surface areas. Then, points located within these potential road surface areas and whose elevation values ​​are more than or equal to a first preset difference (e.g., 0.1 meters) below the surface elevation of the 3D terrain model are extracted from the classified point cloud data as candidate road surface point cloud data. Finally, plane fitting is performed on the candidate road surface point cloud data, calculating the distance from each point to the fitted plane. Outliers with distances greater than a second preset difference (e.g., 0.05 meters) are removed, and the remaining points are designated as the road surface point cloud dataset. The first preset difference can be a pre-defined threshold parameter used to determine the elevation between road surface points and the terrain surface when filtering out point clouds that truly belong to road surface points from potential road surface areas. The first preset difference can be preset according to road design specifications and point cloud accuracy, and is usually set to 0.05 meters to 0.20 meters. The second preset difference can be a preset distance threshold parameter used to determine which points in the candidate road surface point cloud are valid road surface points and which points are outliers. The second preset difference can be preset according to road surface smoothness and point cloud accuracy, and is usually set to 0.02 meters to 0.10 meters.

[0048] The fifth step involves constructing a road surface model based on the aforementioned real-scene textured 3D model and the aforementioned road surface point cloud dataset. In practice, firstly, the execution entity can extract the boundary points of the road surface from the aforementioned road surface point cloud dataset. Then, these boundary points are connected to form a road surface polygon. Next, a regular grid is generated within the aforementioned road surface polygon at a preset road surface grid resolution (e.g., set to 0.5 meters), forming a 3D road surface grid. Then, the texture region corresponding to the aforementioned road surface polygon is extracted from the aforementioned real-scene textured 3D model. Then, the extracted texture region is mapped onto the surface of the aforementioned 3D road surface grid to obtain the road surface model. The elevation values ​​of each grid point in the aforementioned regular grid are obtained by interpolation of the aforementioned road surface point cloud dataset. The aforementioned road surface model can be a 3D road surface grid model with realistic texture. The aforementioned preset road surface grid resolution can be the horizontal spacing between two adjacent grid points in the regular grid generated within the aforementioned road surface polygon, in meters. The aforementioned preset road surface grid resolution can be preset according to the application requirements of the road surface model and the computer rendering performance.

[0049] Step 6: Based on the aforementioned 3D model of real-world texture and 3D model of terrain, the classified point cloud data is processed to construct a river model, resulting in a river model. In practice, firstly, the executing entity can extract points from the classified point cloud data whose reflection intensity values ​​are lower than a preset intensity threshold (e.g., set to 10% of the threshold) and whose elevation values ​​are lower than the surface elevation of the aforementioned 3D terrain model, as candidate water surface points. Then, connectivity analysis is performed on the candidate water surface points to group spatially connected points into the same connected region, thereby selecting the largest connected region as the river water surface region. Next, the boundary points of the river water surface region are extracted. Then, the boundary points are connected to form a water surface polygon. Finally, a water surface mesh model is generated within the aforementioned water surface polygon using a preset fixed elevation value (e.g., the average elevation of all points within the aforementioned river water surface region) as the water surface elevation value. Next, a semi-transparent material is created for the water surface mesh model in a 3D simulation platform (e.g., Cesium). The rendering mode of the semi-transparent material is set to semi-transparent mode, and its transparency value is set to a preset transparency threshold. The semi-transparent material is then applied to all triangular faces of the water surface mesh model. Finally, the water surface mesh model with the semi-transparent material is defined as a river model. This river model can be a 3D model of a river surface with a semi-transparent material. The preset fixed elevation value, in meters (m), is a uniformly set elevation value for the entire river surface area when generating the river surface mesh model. This preset fixed elevation value can be used to determine the vertical position (i.e., Z-coordinate value) of each grid point in the water surface mesh model. The transparency threshold can be preset according to the visual transparency requirements of the river water, with a value range of 0.3 to 0.7 (where 0 is completely transparent and 1 is completely opaque).

[0050] Step 7: Based on the aforementioned 3D model of real-world texture, the classified point cloud data is processed to construct an overhead line model, resulting in the overhead line model. In practice, firstly, the executing entity can extract all points with conductor or ground wire category labels from the classified point cloud data that are not included in the aligned line data, as overhead line point cloud data. Then, the overhead line point cloud data is clustered according to spatial location using the Euclidean distance clustering algorithm, grouping spatially adjacent points into the same cluster, with each cluster corresponding to one overhead line. Next, for each overhead line, a 3D model of the overhead line is generated according to the catenary fitting and tubular model generation method described in Step 3. Finally, all other overhead lines are traversed, and the 3D models of all other overhead lines are combined to obtain the overhead line model. The aforementioned overhead line point cloud data can be point cloud data of other overhead lines (including communication lines, low-voltage distribution lines, etc.) excluding conductors and ground wires. The aforementioned overhead line model can be a 3D model of other overhead lines excluding conductors and ground wires.

[0051] The eighth step is to determine the above-mentioned tower model, conductor model, ground wire model, road surface model, river model, and overhead line model as the three-dimensional model of the transmission line and crossings.

[0052] In the process of adopting technical solutions to address the aforementioned technical problems, the following technical issues often arise during the construction of a 3D simulation model for the construction and sealing of ultra-high voltage transmission lines in mountainous areas of southern China, specifically for the application scenario: Because the towers are located in mountainous areas, the dense shrubs around the tower base and the climbing vines in the middle of the tower obscure some tower materials and fittings, resulting in incomplete point cloud data collection for some tower materials and key fittings (such as the hanging point fittings on the back of the tower), leading to data gaps. Furthermore, since the tower fittings are mostly made of galvanized steel with smooth surfaces and high reflectivity, when scanning with a drone-mounted LiDAR under direct sunlight, specular reflection can easily cause the scanner to fail to receive effective echoes or distort the echo signals, further exacerbating the incompleteness and distortion of the fitting point cloud data. In addition, during long-term scanning in the field, birds (such as swallows and sparrows) often nest or perch on the crossarms, or insects fly across the site. These dynamic objects are recorded by the scanner as isolated point cloud clusters, forming noise. This results in not only missing data in the collected point cloud data but also a large number of irregular noise points. If the point cloud of tower hardware collected under these complex conditions, containing severe occlusion, reflection blind spots, and mixed noise from birds and vegetation trails, is directly used for mesh modeling, the resulting tower model surface will have numerous holes, flying points, and topological breaks. This fails to meet the high-precision geometric restoration and topological integrity requirements of 3D simulation for construction network enclosure, thus affecting the accuracy and reliability of subsequent power grid electromagnetic simulation calculations, live-line work simulations, and refined operation and maintenance management. To address the following requirements for this application scenario: the construction of a 3D simulation model for the construction network enclosure of ultra-high voltage transmission lines in southern mountainous areas requires a high-fidelity, high-precision 3D model construction method that can simultaneously overcome vegetation occlusion, specular reflection, and interference from complex environmental noise to ensure geometric accuracy and topological correctness, supporting engineering-level simulation analysis and applications. Faced with these technical problems, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned execution entity may perform pole model construction processing on the aforementioned classified point cloud data based on the aforementioned aligned line data through the following steps to obtain the pole model: The first step involves extracting the topology and geometric parameters from the aligned line data to obtain the topology and geometric parameter information of the pole and tower hardware. In practice, the executing entity first reads the pole and tower structure description field from the aligned line data to obtain the pole and tower model identifier. Then, using the pole and tower model identifier as an index, it queries the standard configuration information of the hardware corresponding to the pole and tower model identifier from a pre-established standard parameter library for pole and tower hardware. Next, the retrieved standard configuration information is matched and verified with the hanging point coordinates and connection methods recorded in the aligned line data. After confirming their consistency, the topological relationship description (including the connection sequence, suspension relationship, and spatial orientation of each component) in the standard configuration information is determined as the topology. Then, the dimensional parameters (including the length, width, diameter, and installation angle of the hardware) in the standard configuration information are determined as geometric parameters. Finally, the topology and geometric parameters are associated and stored to obtain the topology and geometric parameter information of the pole and tower hardware. The aligned line data is a vectorized data file in DXF or SHP format, containing information such as tower model identification, number of crossarm layers, crossarm length, conductor suspension point coordinates, insulator string type, and connection method. The tower hardware standard parameter library is a pre-stored dataset of hardware types, specifications, and standard topological relationships for various tower models. The topology and geometric parameter information is a structured dataset containing topological descriptions and geometric dimensional parameters of each component of the tower hardware. The tower hardware refers to the metal fittings used on overhead transmission line towers to connect conductors, insulators, crossarms, and other main tower components, including but not limited to suspension clamps, tension clamps, connecting hardware, and protective hardware. The topology refers to the connection relationships and spatial layout of each component of the tower hardware, including adjacency, hierarchical, and relative positional relationships. The geometric parameters refer to the dimensional parameters of each component of the tower hardware, including length, width, height, diameter, and angle, in millimeters (mm).

[0053] The second step involves constructing a semantic constraint framework from the aforementioned topological structure and geometric parameter information. In practice, the executing entity first establishes a hierarchical structure tree for each component of the tower hardware based on the topological relationships described in the topological structure and geometric parameter information. Specifically, first, records for each component are extracted from the topological structure. Then, based on the "suspended" or "connected" relationship direction in the connection relationships, the parent-child hierarchy between components is determined. For any component A and component B, if the topological relationship description records "component B is suspended from component A" or "component B is connected to the lower end of component A," then component A is determined as the parent node and component B as the child node. Subsequently, following the above method, starting from the tower body without a parent node, the parent-child relationships of all nodes are determined layer by layer according to the following hierarchical path: The tower body is the root node; the crossarms directly connected to the tower body are the child nodes of the root node (i.e., the first-level child nodes); the insulator strings suspended at the ends of each crossarm are the child nodes of the corresponding crossarm nodes (i.e., the second-level child nodes); and the fittings connected to the lower ends of each insulator string are the child nodes of the corresponding insulator string nodes (i.e., the third-level child nodes). For multiple child nodes connected to the same parent node (multiple insulator strings on the same level of crossarm), the execution entity numbers and sorts each child node according to the recorded order in the above topological relationship description, and then sequentially attaches each sorted child node to the parent node, obtaining the hierarchical structure tree of each component of the tower fittings. Then, geometric constraints are set for each node in the above hierarchical structure tree according to the above geometric parameters. Next, spatial position constraints are set for each node according to the above topological relationship description. Finally, the hierarchical structure tree with set geometric and spatial constraints is determined as the above semantic constraint framework information. The aforementioned hierarchical structure tree has the tower body as the root node, crossarms as the first-level child nodes, insulator strings as the second-level child nodes, and hardware as the third-level child nodes. The aforementioned geometric constraints can include the standard geometric shape (cylinder, cuboid, or cone) and size range of each component (e.g., hardware length deviation does not exceed ±5% of the standard value). The aforementioned spatial position constraints can include the coordinates of connection points between adjacent components, relative angles, and distance ranges (e.g., the horizontal distance deviation between the lower end hardware of the insulator string and the conductor attachment point does not exceed ±0.02 meters). The aforementioned semantic constraint framework information can be a set of rules constructed based on the aforementioned topology and geometric parameter information, used to constrain the geometric shape and spatial relationships of each component of the tower hardware. This set of rules defines the standard geometric shape, size range, and relative position constraints between each component of the tower hardware. The aforementioned semantic constraint framework information includes the component scope organized in a hierarchical tree form, and the geometric shape and spatial position constraint rules attached to each node of this tree. The aforementioned semantic constraint framework is a set of rules used to constrain the geometric shape and spatial relationships of various components of tower hardware, which can be called upon in subsequent steps.The aforementioned component records include identification information for the tower body, crossarms, insulator strings, and hardware, as well as their interconnections.

[0054] The third step involves extracting pole and tower hardware from the categorized point cloud data to obtain pole and tower hardware point cloud data. In practice, firstly, the executing entity can extract all points with the pole and tower category labels from the categorized point cloud data to form initial pole and tower point cloud data. Then, based on the tower position coordinates of each pole in the aligned line data, spatial clustering is performed on the initial pole and tower point cloud data to group points with spatially adjacent locations into the same cluster, resulting in various pole and tower point cloud clusters. For each pole and tower point cloud cluster, the point cloud within the range of the conductor suspension point coordinates recorded in the aligned line data as the center and a preset search radius (e.g., set to 0.5 meters) is extracted as candidate point clouds for hardware. Next, local geometric feature analysis is performed on the aforementioned candidate point clouds of hardware to calculate the local linearity, flatness, and divergence of each point. Points with divergence less than a preset divergence threshold (e.g., set to 0.3) and spatially distributed in a linear or tubular cluster are retained, and these retained points are identified as the point cloud data of the tower hardware for that base tower. Finally, all tower point cloud clusters are traversed, and all tower hardware point cloud data are merged to obtain the aforementioned tower hardware point cloud data. Each tower point cloud cluster corresponds to one base tower. The aforementioned tower hardware point cloud data can be a set of three-dimensional spatial discrete points extracted from the aforementioned classified point cloud data that belong to the tower hardware category. The aforementioned preset search radius can be the radius value of the three-dimensional spatial search range set when extracting candidate point clouds of hardware centered on the conductor suspension point coordinates recorded in the aforementioned aligned line data, in meters (m). The aforementioned preset divergence threshold can be a pre-set geometric feature determination threshold value used to determine whether each point in the above candidate point cloud of hardware belongs to the surface of the hardware, with a value range of 0 to 1.

[0055] The fourth step involves removing outliers from the aforementioned pole and tower hardware point cloud data to obtain cleaned point cloud data. In practice, for each point cloud in the pole and tower hardware point cloud data, the executing entity first uses a pre-constructed spatial index to search for a preset number (e.g., 20) of nearest neighbors within the point's neighborhood. Then, the average distance from the point cloud to its neighboring points is determined as a first target value. Next, the standard deviation of each obtained first target value is determined as a second target value. Then, in response to any first target value being greater than a preset value, the point cloud is identified as an outlier and removed from the pole and tower hardware point cloud data to update the data. Finally, the updated pole and tower hardware point cloud data is determined as the cleaned point cloud data. The preset value can be the product of the second target value and a preset multiple threshold. The aforementioned preset multiplier threshold can be used as the standard deviation multiplier coefficient for determining outlier noise when removing abnormal noise based on the statistical outlier filtering algorithm.

[0056] The fifth step involves spatially aligning the semantic constraint framework information based on the cleaned point cloud data to obtain aligned semantic constraint framework information. In practice, firstly, the executing entity can perform curvature detection on each point in the cleaned point cloud data to obtain curvature change information. Then, points whose curvature change information exceeds a first preset threshold (e.g., set to 0.1) are marked as key feature points, thereby identifying key feature points of the hardware from the cleaned point cloud data. Next, theoretical key feature point coordinates are extracted from the semantic constraint framework information based on the geometric parameters in the topology and geometric parameter information. Secondly, a theoretical point cloud template is generated in three-dimensional space based on the standard geometric shape and size parameters of each component in the semantic constraint framework information. Then, the endpoints, inflection points, and connection points of each component are extracted from the theoretical point cloud template as theoretical key feature points. Finally, using the key feature points in the cleaned point cloud data as the target point set and the theoretical key feature points as the source point set, the spatial transformation matrix is ​​calculated using the iterative nearest point algorithm. Subsequently, the aforementioned spatial transformation matrix is ​​applied to all theoretical coordinate points in the aforementioned semantic constraint framework information to obtain the transformed semantic constraint framework information, which serves as the aligned semantic constraint framework information. The aforementioned key feature points may include the endpoints, inflection points, and connection points of the hardware. The aforementioned spatial transformation matrix can be a 4×4 homogeneous transformation matrix, including rotation and translation components. The aforementioned curvature variation information can be obtained by statistically analyzing the curvature values ​​of each point in the cleaned point cloud data, serving as a dataset reflecting the changes in the degree of curvature in different regions of the hardware surface. The aforementioned curvature variation information includes the curvature values ​​of each point and the spatial distribution characteristics of the curvature values.

[0057] The sixth step involves performing completion and topology optimization on the cleaned point cloud data and the aligned semantic constraint framework information to obtain the completed point cloud information. In practice, the execution entity can perform spatial distribution density analysis on the cleaned point cloud data to identify sparse and missing regions. Specifically, the three-dimensional space of the cleaned point cloud data is divided into several voxel grids with a preset side length (e.g., 0.02 meters). The number of points in each voxel grid is counted. If the number of points in a voxel grid is less than a preset density threshold (e.g., 3 points), the voxel grid is determined to be a sparse or missing region. Then, based on the aligned semantic constraint framework information, supplementary points are generated in the sparse and missing regions according to standard geometry to obtain the supplemented point cloud data. Specifically, the type and standard geometric parameters of the hardware components belonging to the region are determined. New points are generated within this region using a uniform sampling method, based on the standard geometric shape of the hardware component (e.g., a cylindrical surface, a spherical surface), to achieve a point density that meets a preset target density (e.g., at least 5 points per square centimeter). Then, the supplemented point cloud data is adjusted according to the spatial position constraints in the aligned semantic constraint framework information to ensure that the relative positional relationships between the hardware components satisfy the standard topology. Finally, the adjusted point cloud data is determined as the completed point cloud information. This completed point cloud information can be a point cloud dataset that is geometrically complete and conforms to the standard topology of pole and tower hardware, obtained after the above completion and topology optimization processing. The preset value can be the side length of each voxel grid, in meters, pre-set when dividing the cleaned point cloud data into voxel grids, used to control the granularity of point cloud spatial discretization.

[0058] Step 7: The completed point cloud information is meshed and textured to obtain the tower model. In practice, firstly, the execution entity can calculate the covariance matrix of each point in the completed point cloud information using its neighborhood point set, and take the eigenvector corresponding to the smallest eigenvalue as the normal vector direction of that point. Then, using the Poisson surface reconstruction algorithm, with the completed point cloud information and the normal vectors of each point as input, an implicit indicator function is constructed, and the Poisson equation is solved to extract the isosurface, resulting in an initial triangular mesh. Next, the initial triangular mesh is simplified to a preset target number (e.g., 500,000 to 1,000,000 triangles) using an edge-folding algorithm. Then, the Laplacian smoothing algorithm is used to smooth the mesh surface, resulting in a processed triangular mesh. Finally, the texture image of the corresponding region is extracted from the real-world textured 3D model, and the texture coordinates are assigned to each mesh vertex according to the mapping relationship between each point in the completed point cloud information and the corresponding pixel in the real-world textured 3D model, generating a textured triangular mesh model. Finally, the textured triangular mesh model described above is determined as the tower model described above. The aforementioned preset target number can be a pre-set target value for the total number of triangular faces that the simplified triangular mesh should achieve when the initial triangular mesh is simplified using an edge-folding algorithm, used to control the balance between model accuracy and data volume.

[0059] The above technical solution, combined with steps 106-108 and related content, serves as an inventive point of this disclosure, solving the technical problem of "numerous holes, flying points, and topological breaks appearing on the surface of the tower model, failing to meet the high-precision geometric restoration and topological integrity requirements of the 3D simulation of construction netting." The factors leading to numerous holes, flying points, and topological breaks on the surface of the tower model, failing to meet the high-precision geometric restoration and topological integrity requirements of the 3D simulation of construction netting, are often as follows: Because the tower is located in a mountainous area, the dense shrubs around the base and the climbing plants in the middle of the tower body obscure some tower materials and fittings, resulting in incomplete point cloud data collection for some tower materials and key fittings (such as the hanging point fittings on the back of the tower body), leading to data loss. Simultaneously, because the tower fittings are mostly made of galvanized steel, with smooth surfaces and high reflectivity, when scanning with a drone-mounted lidar under direct sunlight, specular reflection can easily cause the scanner to fail to receive effective echoes or distort the echo signals, further exacerbating the incompleteness and distortion of the fitting point cloud data. Furthermore, during long-term field scanning, birds (such as swallows and sparrows) often nest or perch on crossarms, or insects fly across the site. These dynamic objects are recorded by the scanner as isolated point cloud clusters, forming noise. This results in not only partial data loss in the collected point cloud data but also a large number of irregular noise points. If the point cloud of pole hardware collected under the above complex conditions, which includes severe occlusion, reflection blind spots, and mixed noise from birds and vegetation trails, is directly used for mesh modeling, the resulting pole model surface will have a large number of holes, flying points, and topological breaks. This will fail to meet the high-precision geometric restoration and topological integrity requirements of 3D simulation of construction network enclosure, thus affecting the accuracy and reliability of subsequent power grid electromagnetic simulation calculations, live-line work simulations, and refined operation and maintenance management. If the above factors are resolved, it is possible to avoid a large number of holes, flying points, and topological breaks on the pole model surface, thereby meeting the high-precision geometric restoration and topological integrity requirements of 3D simulation of construction network enclosure. To achieve this effect, firstly, the aligned line data is processed to extract topology and geometric parameters, obtaining the topology and geometric parameters of the tower hardware. This allows us to obtain the connection relationships (such as the connection method between insulator strings and mounting plates), geometric dimensions (such as the diameter and length of the hardware), and spatial relative positions of the various tower hardware components. This provides accurate geometric benchmarks and topological constraints for subsequent completion of missing hardware components caused by shading or reflection, avoiding structural misalignment or dimensional distortion during completion. Then, the aforementioned topology and geometric parameter information is processed to construct a semantic constraint framework, obtaining semantic constraint framework information.Therefore, the aforementioned topological structure and geometric parameter information can be converted into semantic constraint framework information. This guides the generation location, quantity, and geometric shape of supplementary points during subsequent point cloud completion, ensuring that supplementary points are not randomly inserted but generated strictly according to standard topological structure and standard geometric dimensions. This guarantees that the completed tower model is geometrically consistent with the design standard, avoiding secondary distortion caused by blind completion. Next, the classified point cloud data undergoes tower hardware extraction processing to obtain tower hardware point cloud data. This allows for the separation of points belonging to the tower hardware category from the classified point cloud data, while points belonging to other categories such as the tower body, insulator strings, and conductors are removed, resulting in a dedicated point cloud dataset focused on the hardware structure. Then, the tower hardware point cloud data undergoes abnormal noise removal processing to obtain cleaned point cloud data. This allows for the identification and deletion of outlier isolated points and erroneous measurement points caused by specular reflection, bird perches, insects flying by, etc., in the tower hardware point cloud data. Then, based on the cleaned point cloud data, the semantic constraint framework information is spatially aligned to obtain aligned semantic constraint framework information. This allows for a consistency transformation between the spatial coordinate system of the semantic constraint framework information (theoretical standard template) and the spatial coordinate system of the cleaned point cloud data (actual measurement data), ensuring their spatial positions match. Next, the cleaned point cloud data and the aligned semantic constraint framework information are completed and topologically optimized to obtain completed point cloud information. Based on the prior knowledge of the semantic constraint framework information, holes in the hardware point cloud caused by vegetation shading and specular reflection can be intelligently filled, and topological connection errors can be automatically repaired. This restores the model's complete geometry and correct topological relationships, eliminates model breaks caused by data loss, ensures that the connection relationships between hardware conform to engineering reality, and meets the stringent requirements of simulation analysis for model integrity. Finally, the completed point cloud information is meshed and texture-mapped to obtain the tower model. Therefore, the completed point cloud can be converted into a high-precision 3D mesh model and given realistic material textures, generating a digital twin model of the tower with complete geometric details, correct topological structure, and realistic visual effects. This eliminates holes, flying points, and topological breaks, meeting the requirements of 3D simulation of construction network enclosure for geometric accuracy and topological integrity of the model, and providing reliable data support for subsequent power grid electromagnetic simulation calculations, live-line operation simulation, and refined operation and maintenance management.This method also utilizes semantic constraint framework information constructed by extracting standard topology and geometric parameters from aligned line data. This framework guides the precise extraction of hardware point clouds, targeted removal of abnormal noise, alignment and calibration of spatial coordinate systems, and intelligent completion of missing regions. Ultimately, a continuous, complete, and accurate tower model is generated, avoiding numerous holes, flying points, and topological breaks on the tower model surface. Combined with steps 106-108, a 3D simulation model of the power grid enclosure is generated, thus meeting the high-precision geometric restoration and topological integrity requirements for 3D simulation of power grid enclosure construction. This further improves the accuracy and reliability of subsequent power grid electromagnetic simulation calculations, live-line working simulations, and refined operation and maintenance management.

[0060] Step 106: Model the temporary construction facilities and the transmission lines to be built on the construction design drawings to obtain the three-dimensional models of the temporary construction facilities and the transmission lines to be built.

[0061] In some embodiments, the aforementioned implementing entity may perform modeling processing on the aforementioned construction design drawings to obtain three-dimensional models of the temporary construction facilities and the transmission lines to be built. The aforementioned construction design drawings can be a general term for all types of design drawings involved in the construction of a new overhead transmission line crossing. The aforementioned construction design drawings may include design drawings of the transmission lines to be built and design drawings of the network enclosure scheme.

[0062] In some optional implementations of certain embodiments, the aforementioned execution entity may perform modeling processing of the construction design drawings for temporary construction facilities and the transmission line to be built through the following steps to obtain a three-dimensional model of the temporary construction facilities and a model of the transmission line to be built: The first step is to generate a model of the proposed transmission line based on the aforementioned design drawing. This design drawing can be a file containing the tower locations, tower types, conductor types, conductor suspension point coordinates, and conductor sag parameters for the new overhead transmission line; the file format can be DWG or DXF. In practice, the executing entity first reads the design drawing and extracts the tower location coordinates, tower type, conductor suspension point coordinates, and conductor sag parameters from it. Then, based on the tower location coordinates and tower type, a simplified tower model is retrieved from a pre-established standard tower model library. Next, the simplified tower model is placed in a unified 3D scene according to the tower location coordinates. Then, based on the conductor suspension point coordinates and conductor sag parameters, a conductor path curve is generated according to the catenary equation. A tubular model is then stretched along the conductor path curve with a preset conductor radius (e.g., set to 0.02 meters) as a schematic model for the conductor to be deployed. Finally, the simplified tower model and the schematic model of the conductor to be deployed are determined as the model of the transmission line to be built. The simplified tower model can be a low-polygon model that retains the main shape features of the tower, with no more than 5000 triangles. The pre-established standard tower model library can be a dataset pre-stored in non-volatile memory containing simplified tower models of various common tower types and their corresponding model identifiers. This standard tower model library is used during the construction of the new tower model to quickly call the corresponding simplified tower model based on the tower type extracted from the design drawings of the transmission line to be built, avoiding the need for parametric modeling or reverse modeling for each new tower, thus improving the construction efficiency of the new line model.

[0063] The second step involves constructing a protective facility model from the aforementioned netting scheme design drawing. In practice, firstly, the executing entity can read the netting scheme design drawing and extract the layout parameters of the netting structure. Then, based on the anchorage coordinates and sag values ​​of the catenary, a catenary path curve is generated according to the catenary equation. A tubular model is then generated by stretching the catenary path curve along a preset cable diameter (e.g., 0.02 meters), resulting in the catenary model. Next, a planar mesh model is generated above the catenary model according to the width and length of the protective netting, and a mesh texture is superimposed on the planar mesh model to obtain the protective netting model. Afterward, a cuboid assembly model is generated based on the installation position and dimensions of the auxiliary crossarm, resulting in the auxiliary crossarm model. Finally, based on the position and height of the crossing frame, a truss structure model is generated according to the standard dimensions of steel pipe scaffolding (e.g., upright spacing of 1.5 meters and crossbar spacing of 1.8 meters) as the crossing frame model. Next, the aforementioned catenary model, protective net model, auxiliary crossarm model, and crossing frame model are defined as the netting protection facility model. The aforementioned netting scheme design drawing can be a design drawing file recording the catenary arrangement, protective net dimensions, auxiliary crossarm positions, and crossing frame structure of the netting structure used to protect existing lines and objects under crossing during the crossing construction. The format can be DWG or DXF. The aforementioned netting protection facility model can be a three-dimensional digital model of the netting structure used to protect existing lines and objects under crossing during the crossing construction, including the catenary model, protective net model, auxiliary crossarm model, and crossing frame model. The aforementioned catenary can be the steel wire rope in the netting structure that bears the weight of the protective net. The aforementioned protective net can be the protective net body erected above the catenary to catch fallen conductors. The aforementioned auxiliary crossarm can be a temporary support structure installed on the crossarm of the existing line tower to fix the ends of the catenary. The aforementioned crossing frame can be a temporary scaffolding or truss structure erected on both sides of the object crossing to support the netting structure. The aforementioned arrangement parameters may include the anchorage coordinates of the catenary, the span and sag value of the catenary, the width and length of the protective netting, the installation position and dimensions of the auxiliary crossarms, and the position and height of the crossing frame. The catenary model can be a three-dimensional tubular model used in the protective netting facility model to simulate the spatial position and geometry of the catenary. The protective netting model can be a three-dimensional planar mesh model used in the protective netting facility model to simulate the spatial position and coverage area of ​​the protective netting. The auxiliary crossarms can be a three-dimensional composite model used in the protective netting facility model to simulate the spatial position and geometry of the auxiliary crossarms. The crossing frame model can be a three-dimensional truss structure model used in the protective netting facility model to simulate the spatial position and geometry of the crossing frame.

[0064] The third step is to obtain the parameter information of each construction machine. In practice, the aforementioned executing entity can read the parameter information corresponding to each construction machine from a pre-established construction machine parameter database. This database can be a collection of data that pre-stores the external dimensions, operating parameters, and model file paths of various types of construction machines. The construction machines can be various types of machinery and equipment used in the crossing construction, including tension machines, traction machines, and cranes. The tension machine can be a construction machine used to apply a set tension to the conductor during conductor deployment. The traction machine can be a construction machine used to pull the conductor or traction rope. The crane can be a lifting machine used to lift crossing frame components and auxiliary crossarms. The construction machine parameter information can be a set of parameters describing the external dimensions and operating status of the aforementioned construction machines, including machine type, length, width, height, and operating radius.

[0065] The fourth step involves processing the aforementioned construction machinery parameter information to construct construction machinery models, resulting in individual construction machinery models. In practice, firstly, for each construction machinery parameter information, the executing entity can read the pre-established simplified model file corresponding to that model of construction machinery from the pre-established construction machinery parameter database. Then, the read simplified model files are imported into a unified 3D scene to obtain the corresponding construction machinery model. The simplified model file can be a lightweight 3D model data file pre-stored in the aforementioned construction machinery parameter database, used to characterize the 3D geometry of the construction machinery. The simplified model file is a 3D mesh model file in FBX or OBJ format. The simplified model file records the main outline and key structural features of the construction machinery, but omits internal structural details, small parts (such as bolts, pipes, and instrument panels), and high-precision surface details to control the model data volume, enabling it to be quickly loaded and rendered by low-configuration computers. The fifth step involves determining the aforementioned netting protection facility model and the aforementioned construction machinery models as the 3D model of the temporary construction facility. In practice, the aforementioned implementing entities can place the aforementioned netting protection facility model and the aforementioned construction machinery models in the same three-dimensional scene space according to their designed position coordinates, and the combined models can be determined as the three-dimensional model of the temporary construction facility.

[0066] Step 107: Perform multi-source model scene integration processing on the 3D terrain model, the 3D model of the transmission line and crossing, the model of the transmission line to be built, and the 3D model of the temporary construction facilities to obtain the scene integration model.

[0067] In some embodiments, the execution entity may perform multi-source model scene integration processing on the above-mentioned terrain 3D model, the above-mentioned power transmission line and crossing object 3D model, the above-mentioned power transmission line to be built model and the above-mentioned temporary construction facility 3D model to obtain a scene integration model.

[0068] In addressing the technical challenges mentioned above, the digital simulation and visualization of UHV transmission line construction network enclosure schemes in complex terrain environments such as mountainous areas in southern China often presents the following technical problems: In complex terrain environments like mountainous areas in southern China, after constructing high-precision models of towers, lines, and temporary facilities, it is necessary to integrate these massive, heterogeneous 3D model data into a unified virtual simulation platform for dynamic simulation of the entire construction process, safety distance verification, collision detection, and visual disclosure of construction plans. Due to the extremely high polygon count of the original models and large-scale terrain data, direct loading can lead to memory overflow, low rendering frame rates, and even system crashes. Furthermore, the integrated 3D scene contains hundreds of independent models (multiple towers, multi-phase conductors, multiple crossings, multiple construction machines, and multiple temporary facilities), lacking a systematic scene organization structure. The models are stored randomly in the scene in a tiled manner, lacking parent-child hierarchical relationships and logical groupings. This disordered management method results in a spatial query time complexity of [missing information]. As the number of models increases, the computational load grows exponentially, leading to screen stuttering and high operational latency. The following requirements are necessary for this application scenario: digital simulation and visualization of the network enclosure scheme for ultra-high voltage transmission line construction in mountainous areas of southern China necessitate efficient and low-latency integration and hierarchical management of massive heterogeneous 3D models. This is crucial for dynamic simulation of the entire construction process, safety distance verification, collision detection, and visual disclosure of the construction plan based on the integrated 3D scene. To address these technical challenges, we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned execution entity may perform multi-source model scene integration processing on the aforementioned 3D terrain model, the aforementioned 3D model of the transmission line and crossing, the aforementioned model of the transmission line to be built, and the aforementioned 3D model of the temporary construction facilities through the following steps to obtain a scene integration model: The first step involves protective lightweighting of the aforementioned 3D terrain model, the aforementioned 3D model of power transmission line crossings, the aforementioned model of the power transmission line to be built, and the aforementioned 3D model library of temporary construction facilities, resulting in lightweight 3D terrain models, lightweight 3D models of power transmission line crossings, lightweight models of the power transmission line to be built, and lightweight 3D models of temporary construction facilities. In practice, the implementing entity can first simplify the terrain mesh using an edge-folding algorithm based on a quadratic error metric. The upper limit of the number of triangles in the terrain mesh is set to a target value (e.g., 500,000 triangles). During the simplification process, triangles in areas with drastic elevation changes (i.e., areas with slopes greater than a preset slope threshold (e.g., 15 degrees)) are retained first, while triangles in areas with gentle elevation changes (i.e., areas with slopes less than or equal to the preset slope threshold) are simplified first, thus simplifying the aforementioned 3D terrain model. Then, the simplified 3D terrain model is determined as the lightweight 3D terrain model. Next, for the aforementioned 3D models of transmission lines and crossings, the detail level priority of each model within the 3D model can be determined according to a preset priority. Then, protective lightweighting is performed according to the following priority order: tower models have the highest priority, followed by conductor and ground wire models, and road and river models have the lowest priority. For the highest-priority tower models, the original number of triangular faces remains unchanged. For the second-highest-priority conductor and ground wire models, an edge-folding algorithm is used to simplify the number of triangular faces to 50% of the original number. For the lowest-priority road and river models, an edge-folding algorithm is used to simplify the number of triangular faces to 20% of the original number. Then, the simplified 3D model of transmission lines and crossings is determined as the lightweight 3D model of transmission lines and crossings. Subsequently, for the aforementioned transmission line model to be built, the original number of triangular faces remains unchanged, and the aforementioned transmission line model to be built is directly determined as the lightweight model of the transmission line to be built. Next, for the aforementioned 3D model of temporary construction facilities, an edge-folding algorithm is used to simplify the number of triangular faces in each component model to 30% of the original number, resulting in a simplified 3D model of temporary construction facilities. Finally, the simplified 3D model of temporary construction facilities is determined as the aforementioned lightweight 3D model of temporary construction facilities. Specifically, the aforementioned lightweight terrain 3D model can be a terrain 3D model obtained after the aforementioned protective lightweighting process, with reduced data volume but recognizable terrain geometry and texture. The aforementioned lightweight transmission line and crossing 3D model can be a transmission line and crossing 3D model obtained after the aforementioned protective lightweighting process, with reduced data volume. The aforementioned lightweight transmission line model to be built can be a transmission line model to be built, obtained after the aforementioned protective lightweighting process, with reduced data volume.The aforementioned lightweight construction temporary facility 3D model can be the 3D model of the construction temporary facility with reduced data volume obtained after the above-mentioned protective lightweighting process. The aforementioned preset priority can be: "The tower model in the aforementioned transmission line and crossing 3D model has the highest priority; the conductor model and ground wire model in the aforementioned transmission line and crossing 3D model have the second highest priority; the road surface model and river model in the aforementioned transmission line and crossing 3D model have the lowest priority." The second step is to define the aforementioned lightweight terrain 3D model, the aforementioned lightweight transmission line crossing 3D model, the aforementioned lightweight transmission line under construction model, and the aforementioned lightweight construction temporary facility 3D model as standardized 3D models.

[0069] The third step involves scene construction processing of the standardized 3D models to obtain initial 3D scene information. In practice, firstly, the execution entity can create a blank scene file in a 3D simulation platform (e.g., Cesium platform) and set the scene coordinate system to be consistent with the engineering coordinate system to obtain a blank 3D scene. Then, for each model in the standardized 3D model, the spatial coordinate information corresponding to the model is read. Next, the model is placed in its corresponding position in the blank 3D scene according to the spatial coordinate information. Finally, the resulting 3D model is identified as the initial 3D scene information. The blank 3D scene can be an empty scene file that does not contain any models but has a defined scene coordinate system, consistent with the engineering coordinate system. The spatial coordinate information records the 3D spatial position of the standardized 3D model in the engineering coordinate system. The initial 3D scene information can be the 3D scene data with preliminary spatial layout obtained after placing the standardized 3D models in a unified 3D scene space according to the spatial coordinates in the engineering coordinate system. This 3D scene data at least includes the spatial position coordinates and model identifiers of each model.

[0070] The fourth step involves performing hierarchical encoding and linking processing on the initial 3D scene information based on the standardized 3D model, resulting in a layered scene model. In practice, firstly, the executing entity can classify all models according to their feature categories: terrain, existing lines, crossings, new lines, and temporary construction facilities. Then, a category code is assigned to each category. Next, for each independent model within each category, sub-codes are assigned to the standardized model according to their natural order within the category (e.g., for existing lines, each base tower is assigned sub-codes 01, 02, 03, etc., according to its numbering order). Then, the category codes and sub-codes of each model are combined to form a hierarchical code, which is then written into the initial 3D scene information as part of the model identifier. Finally, a parent-child hierarchical relationship is established in the initial 3D scene information, with the scene root node as the parent node, each category as the child node, and each independent model as the leaf node. Finally, the scene information with the established parent-child hierarchical relationship is determined as the layered scene model. The aforementioned category codes can be preset numerical identifiers (e.g., terrain type code 01, existing line type code 02, intersection / crossing type code 03, newly built line type code 04, and temporary construction facility type code 05). The aforementioned layered scene model can be a scene model with a tree-like hierarchical structure obtained by organizing and encoding the various models in the aforementioned initial 3D scene information according to a preset hierarchical structure. The aforementioned tree-like hierarchical structure has the scene root node as the top layer, the terrain type as the middle layer, and each independent model as the bottom leaf node.

[0071] Fifth, based on the standardized model described above, multi-level detail generation and configuration processing is performed on the layered scene model to obtain a hierarchical 3D scene model. In practice, firstly, for each model in the layered scene model, the execution entity can generate three different precision versions based on the original number of triangle faces: the first precision version is the original model with N faces (N being the original number of faces); the second precision version uses an edge-folding algorithm to simplify the number of faces to 40% of N, serving as a medium precision version; the third precision version uses an edge-folding algorithm to simplify the number of faces to 10% of N, serving as a low precision version. Then, view distance switching rules are configured for the three different precision version models. Finally, the layered scene model configured with multi-level detail versions and view distance switching rules is determined as the aforementioned hierarchical 3D scene model. The aforementioned view distance switching rules may include: displaying the first precision version when the viewing distance is less than a first distance threshold (e.g., set to 50 meters); displaying the second precision version when the viewing distance is greater than or equal to the first distance threshold and less than a second distance threshold (e.g., set to 200 meters); and displaying the third precision version when the viewing distance is greater than or equal to the second distance threshold. This multi-level detail technique allows for using different precision versions of the same 3D model at different display distances. Closer distances use higher model precision and more triangles, while farther distances use lower model precision and fewer triangles. The aforementioned hierarchical 3D scene model can be a scene model obtained after the aforementioned multi-level detail generation and configuration processing, where each model is configured with a multi-level detail version and view distance switching rules.

[0072] The sixth step involves enhancing the scene interaction of the hierarchical 3D scene model to obtain an integrated scene model. In practice, firstly, the executing entity can add spatial measurement functionality to the hierarchical 3D scene model. Specifically, a mouse click event listener is registered in the hierarchical 3D scene model. When a user clicks two points in the hierarchical 3D scene model, the current view matrix and projection matrix are obtained through the graphics API. Combining the screen coordinates and window size, the screen's 2D coordinates are converted to 3D world coordinates using inverse matrix transformation. Based on spatial geometry formulas, the 3D spatial distance, height difference, angle, or area of ​​a closed region between the two points is calculated (e.g., calculating the area of ​​a polygon through triangulation). The measurement results are then displayed in real-time at a specified location on the scene interface. Next, a collision detection function is added to the hierarchical 3D scene model. Specifically, the axis-aligned bounding box (AABB) method is used to traverse the vertex coordinates of each model in the hierarchical 3D scene model, constructing a minimum bounding box along the X, Y, and Z axes for each model. Then, by real-time detection of the projection range of each model's bounding box in 3D space, a collision is determined to have occurred if the X, Y, and Z axis ranges of any two models overlap. When a collision is detected, logic prevents further movement of the colliding model and triggers visual or auditory cues to provide feedback on the collision result. Next, view control functionality is added to the hierarchical 3D scene model. Specifically, by listening to user mouse, keyboard, or touch operations, scene camera parameters are updated in real-time to achieve viewpoint changes. Then, rotation operations adjust the scene camera's yaw and pitch angles to change the viewpoint direction; translation operations move the scene camera's position along the camera's right or up vector to achieve scene roaming; zoom operations adjust the scene camera's field of view or zoom the scene camera's distance from the scene along the viewing direction to achieve near / far viewing; and reset operations reset the scene camera parameters to the preset default viewpoint. Next, a model picking function is added to the hierarchical 3D scene model using ray casting. Specifically, when the user clicks on the hierarchical 3D scene model, a virtual ray is first generated based on the current scene camera position and the screen click coordinates. Then, the triangular meshes of all models in the hierarchical 3D scene model are traversed, and the intersections of rays with each triangular face are calculated. Next, the intersection point closest to the scene camera is selected, and the model to which this intersection point belongs is marked as the picking target. Finally, the model is highlighted, and the associated data interface is called to display its material, position, rotation angle, and other attribute information in the sidebar of the interface. Finally, the hierarchical 3D scene model with added spatial measurement, collision detection, view control, and model picking functions is determined as the scene integration model. The aforementioned spatial measurement functions include distance measurement, height measurement, angle measurement, and area measurement. The collision detection function is used to detect whether there are spatial intersections between models in the scene. The view control function includes rotation, translation, scaling, and reset operations.The aforementioned model picking function highlights a model and displays its attribute information when a user clicks on it in the scene. The aforementioned camera right vector can be a spatial unit vector representing the scene camera's horizontal rightward direction within a 3D local coordinate system established with the scene camera as the origin in the aforementioned hierarchical 3D scene model. The aforementioned camera up vector can be a spatial unit vector representing the scene camera's vertical upward direction within a 3D local coordinate system established with the scene camera as the origin in the aforementioned hierarchical 3D scene model. The aforementioned scene camera can be a virtual entity object within the aforementioned hierarchical 3D scene model that defines the user's viewing angle and field of view. The aforementioned scene camera can be a logical unit in the 3D simulation platform that implements scene rendering and viewpoint control, determining what content in the aforementioned hierarchical 3D scene model the user can see on the computer screen and from what angle.

[0073] The above technical solution, combined with step 108 and related content, serves as an inventive point of this disclosure, solving the technical problem of "screen stuttering and high operation response latency." Factors leading to screen stuttering and high operation response latency often include: In complex terrain environments such as mountainous areas in southern China, after constructing high-precision models of towers, lines, and temporary facilities, it is necessary to integrate these massive, heterogeneous 3D model data into a unified virtual simulation platform for dynamic simulation of the entire construction process, safety distance verification, collision detection, and visual disclosure of construction plans. Due to the extremely high polygon count of the original models and large-scale terrain data, direct loading can lead to memory overflow, low rendering frame rates, and even system crashes. Furthermore, because the integrated 3D scene contains hundreds of independent models (multiple towers, multi-phase conductors, multiple crossings, multiple construction machines, and multiple temporary facilities), it lacks a systematic scene organization structure. The models are stored randomly in the scene in a tiled manner, lacking parent-child hierarchical relationships and logical groupings. This disordered management method results in a spatial query time complexity of O(n²). As the number of models increases, the computational load grows exponentially, leading to screen stuttering and high operational response latency. Solving these factors would avoid screen stuttering and reduce operational response latency. To achieve this, the aforementioned 3D terrain model, the aforementioned 3D transmission line crossing model, the aforementioned planned transmission line model, and the aforementioned temporary construction facility 3D model library are first subjected to protective lightweighting processing, resulting in lightweight terrain 3D models, lightweight transmission line crossing models, lightweight planned transmission line models, and lightweight temporary construction facility 3D models. This allows for lightweighting of each model while maintaining the identifiable geometric features and texture quality of each model. By reducing the number of triangle faces in each model through edge folding algorithms and reducing texture image resolution to decrease texture data volume, the resulting models are lightweighted, reducing memory usage and rendering computation. This provides a data foundation for resolving memory overflow and low rendering frame rates caused by the extremely high face count of the original models. Then, the aforementioned lightweight terrain 3D model, lightweight transmission line crossing 3D model, lightweight planned transmission line model, and lightweight temporary construction facility 3D model are identified as standardized 3D models. Next, scene building processing is performed on these standardized 3D models to obtain initial 3D scene information. Thus, the standardized 3D models can be automatically placed in their corresponding positions within a unified 3D scene space according to their respective spatial coordinates in the engineering coordinate system, forming a complete spatial layout and obtaining initial 3D scene information. Following this, based on the standardized 3D models, the initial 3D scene information undergoes hierarchical encoding and linking processing to obtain a layered scene model.Therefore, based on the type of each model, they can be categorized according to land cover types, each category can be assigned a category code, and each independent model can be assigned a sub-code, forming a hierarchical coding system. A tree-like hierarchical structure is then established in the scene, with the scene root node as the parent node, each category as the child node, and each independent model as the leaf node, resulting in a layered scene model. By establishing this tree-like hierarchical structure, when construction workers perform spatial measurements, the computer only needs to retrieve relevant models under the corresponding category node, reducing the computational load required for spatial queries and improving the responsiveness of scene interaction from the perspective of scene organization. Subsequently, based on the above standardized model, the layered scene model undergoes multi-level detail generation and configuration processing to obtain a hierarchical 3D scene model. This allows for the allocation of different precision display levels to models at different distances, ensuring that only the model details of the highest visual importance are rendered within the field of view, thereby reducing GPU load and maintaining a smooth frame rate at different viewing distances, solving the stuttering problem caused by excessive rendering of model details. Finally, the hierarchical 3D scene model undergoes scene interaction enhancement processing to obtain a scene integration model. Therefore, spatial measurement functions (distance measurement, height measurement, angle measurement, area measurement), collision detection functions (real-time collision detection based on axis-aligned bounding box method), view control functions (rotation, translation, scaling, reset), and model picking functions (clicking to highlight and display attribute information) can be added to the above-mentioned hierarchical 3D scene model. While ensuring the smooth operation of the scene integration model at a high frame rate, the scene integration model is endowed with engineering analysis capabilities to support the refined verification of construction schemes. This allows construction personnel to complete spatial measurements, collision detection, and safety distance verification of the netting scheme in a smooth interactive experience, upgrading the 3D simulation scene from a static display model to an interactive engineering tool. This is because, by reducing data volume through protective lightweight processing, eliminating heterogeneity between models through standardization, achieving one-click model integration through scene building, reducing spatial query complexity by establishing an ordered tree-like organizational structure through hierarchical coding, reducing graphics processor load by achieving view distance-based adaptive rendering through multi-detail level generation and configuration, and finally, providing scene engineering analysis capabilities through scene interaction enhancement, the problems of screen lag and operation response delay are solved. Combined with step 108, the above scene integration model is parametrically configured based on preset construction netting design parameters and preset construction scene configuration information to obtain a netting 3D simulation model. Finally, a netting 3D simulation model that can run smoothly on ordinary configuration computers and supports real-time spatial measurement and collision detection is obtained, thus meeting the engineering needs of construction plan review and dynamic simulation.

[0074] Step 108: Based on the preset construction netting design parameters and preset construction scenario configuration information, the scenario integration model is parametrically configured to obtain a three-dimensional simulation model of the netting.

[0075] In some embodiments, the aforementioned execution entity can perform parametric configuration processing on the scenario integration model based on preset construction netting design parameters and preset construction scenario configuration information to obtain a three-dimensional simulation model of the netting. In practice, firstly, the aforementioned execution entity can read the pre-stored preset construction netting design parameters and preset construction scenario configuration information from non-volatile memory. Next, based on the location of the netting work points, the spatial location of the netting structure is located in the scenario integration model. Then, based on the coverage area of ​​the netting, the projection boundary of the netting structure on the horizontal plane is determined. Then, based on the anchor point coordinates, the positions of the two fixed points of the catenary are determined. Then, based on the span, tension, and sag of the catenary, a catenary path curve is generated according to the catenary equation. Secondly, a tubular model is generated by stretching along the catenary path curve with a preset cable diameter (e.g., set to 0.02 meters) to obtain the catenary model. Then, based on the span and width of the netting, the coverage area of ​​the protective netting is determined. Next, a planar mesh model is generated above the aforementioned load-bearing cable model according to the aforementioned protective netting mesh dimensions. Then, a mesh texture is superimposed on the aforementioned planar mesh model to obtain the protective netting model. Next, the installation position of the auxiliary crossarm is determined according to the aforementioned anchor point coordinates. Then, the installation height of the auxiliary crossarm is determined according to the aforementioned netting height. Then, a cuboid assembly model is generated according to the preset standard dimensions of the auxiliary crossarm (e.g., the length is the same as the aforementioned netting width, and the cross-sectional dimensions are 0.3 meters × 0.3 meters) to obtain the auxiliary crossarm model. Then, the erection position and height of the crossing frame are determined according to the aforementioned netting height and the aforementioned netting span. Next, a truss structure model is generated according to the standard dimensions of steel pipe scaffolding (e.g., the upright spacing is 1.5 meters, and the horizontal bar step distance is 1.8 meters) to obtain the crossing frame model. Finally, the aforementioned load-bearing cable model, the aforementioned protective netting model, the aforementioned auxiliary crossarm model, and the aforementioned crossing frame model are combined according to their respective spatial positions to form the netting protection facility model. Next, the aforementioned protective netting model is attached to the scene tree of the aforementioned integrated scene model as a child node of the temporary construction facility category. Finally, the integrated scene model is defined as the aforementioned 3D simulation model of the protective netting. The aforementioned preset construction netting design parameters can be a set of pre-defined geometric layout parameters and physical performance parameters of the netting structure in 3D space, including the coordinates of the crossing position, netting span, netting width, netting height, tension value of the catenary cable, sag value of the catenary cable, mesh size of the protective netting, and coordinates of the anchor points. The aforementioned netting span can be the horizontal projection distance of the netting structure in the crossing direction, in meters. The aforementioned netting width can be the horizontal width of the netting structure perpendicular to the crossing direction, in meters. The aforementioned netting height can be the vertical distance of the lowest point of the netting structure relative to the ground or the top surface of the object being crossed, in meters.The aforementioned tension value of the catenary cable can be the tensile force applied to the catenary cable at the anchorage end, expressed in kilonewtons (kN). This tension value determines the sag shape of the catenary cable. The aforementioned sag value of the catenary cable can be the sag distance at the midpoint of the catenary cable, i.e., the vertical distance between the line connecting the two anchorage points and the lowest point of the catenary cable, expressed in meters. The aforementioned mesh size of the protective net can be the side length of the mesh, expressed in centimeters. The aforementioned anchorage point coordinates can be the three-dimensional spatial coordinates of the fixed points at both ends of the catenary cable, expressed in meters. The aforementioned non-volatile memory can be any of a hard disk, solid-state drive, or database storage device. The aforementioned preset construction scenario configuration information can be the spatial layout constraint information of the pre-set netting scheme in the aforementioned scenario integration model, including the location of the netting work site, the coverage area of ​​the netting, and the parking location of construction machinery. The aforementioned netting work site location can be the specific geographical coordinates of the netting structure at the construction site, expressed in three-dimensional spatial coordinates (X, Y, Z) under the aforementioned engineering coordinate system. The aforementioned coverage area of ​​the enclosure can be the projected area of ​​the enclosure structure on the horizontal plane, defined by the coordinate sequence of the boundary points of this area. The aforementioned 3D simulation model of the enclosure can be a complete 3D simulation scene model obtained after the parametric configuration of the aforementioned enclosure model, integrating a model of the enclosure protection facility that conforms to the aforementioned preset construction enclosure design parameters into the aforementioned scene integration model. The aforementioned preset cable diameter can be the cross-sectional diameter used to generate a tubular 3D mesh model by stretching along the aforementioned load-bearing cable path curve as the axis when generating the load-bearing cable model, in meters. The aforementioned preset standard dimensions of the auxiliary crossarm can be a set of geometric dimension parameters of each component of the auxiliary crossarm pre-set when generating the 3D model of the auxiliary crossarm, used to determine the spatial occupancy and geometric shape of the auxiliary crossarm model in the 3D scene.

[0076] The above-described embodiments of this disclosure have the following beneficial effects: the three-dimensional simulation model construction method for construction safety netting according to some embodiments of this disclosure improves the engineering usability of the constructed three-dimensional simulation model and ensures that the three-dimensional simulation model plays a substantial role in construction safety verification. Specifically, the reason why the constructed three-dimensional simulation model has low engineering usability and cannot play a substantial role in construction safety verification is that it relies solely on UAV oblique photography technology to generate a textured three-dimensional model by aerial reconstruction of the site. Although this method can restore the surface texture, since power facilities such as conductors and insulator strings are small linear objects, they are prone to tearing, breakage, or geometric distortion during reconstruction, and cannot effectively penetrate vegetation to obtain the true ground elevation (DEM), resulting in insufficient model accuracy. At the same time, since the design parameters of the netting, such as the tension of the catenary cable, the sag curve, and the coordinates of the anchor points, are not embedded into the three-dimensional scene, there is a lack of parameter-driven mechanism. When crossing spans or changes in construction conditions, the geometry of the netting cannot be automatically reconstructed, and manual adjustment of the model's position and attitude is required, which leads to a significant increase in the time required for multi-scheme comparison and makes it difficult to cover extreme working conditions. This results in the simulation model degenerating into a static visual composite, unable to perform reproducible, verifiable, and recalculated engineering design-level deductions of the netting layout and line laying conditions while ensuring spatial accuracy. Consequently, the constructed 3D simulation model has low engineering usability and cannot play a substantial role in construction safety verification. Based on this, the method for constructing a 3D simulation model for construction netting in some embodiments of this disclosure first acquires multi-source survey data of the construction site, vectorized data of the original line, and construction design drawings, wherein the multi-source survey data of the construction site includes point cloud data and oblique images of the construction site. Next, the oblique images of the construction site are processed to construct a 3D real-scene model, resulting in a real-scene textured 3D model. Thus, a real-scene 3D model with surface texture can be generated based on the oblique images of the construction site. Then, based on the point cloud data of the construction site, the real-scene textured 3D model, and the original line vectorized data, aligned point cloud data, aligned real-scene textured 3D model, and aligned line data are generated. Therefore, the point cloud data of the construction site from LiDAR, the 3D model of the real-scene texture, and the original vectorized data of the route can be unified under the same engineering coordinate system and accurately aligned spatially. This ensures that the multi-source data are consistent in spatial position, providing a data foundation for subsequent high-precision modeling with a unified coordinate reference and accurate matching of geometry and texture. It solves the problem of inconsistent coordinate systems and spatial offsets between multi-source heterogeneous data that prevent fusion modeling. Subsequently, the aligned point cloud data and the oblique image of the construction site are subjected to terrain and texture fusion modeling processing to obtain a 3D terrain model.Therefore, accurate ground elevation data can be generated from the aligned point cloud data to construct a raster terrain grid. A digital orthophoto map generated from oblique imagery of the construction site is then mapped onto the terrain grid surface as texture, generating a 3D terrain model that contains both accurate ground elevation information and realistic surface texture. This solves the problem of insufficient terrain model accuracy caused by relying solely on oblique photography, which cannot effectively penetrate vegetation to obtain the true ground elevation (DEM). This provides an accurate terrain reference surface for the spatial positioning of subsequent power line facilities and fencing structures. Furthermore, based on the aforementioned 3D terrain model, the aligned real-scene texture 3D model, the aligned point cloud data, and the aligned power line data, a 3D model of the transmission line and its crossings is generated. Therefore, based on the tower locations, conductor suspension point coordinates, and conductor sag parameters in the aligned line data, tower point clouds and conductor point clouds can be extracted from the aligned point cloud data. Tower models with hardware details are generated through parametric tower modeling, and smooth conductor and ground wire models are generated through catenary fitting. Simultaneously, road surface point clouds are extracted from the point cloud data to generate a road surface model, water surface points are extracted to generate a river model, and other overhead line points are extracted to generate an overhead line model. This forms a complete set of 3D models including existing transmission lines and all crossings, solving the problem of easily distorted, broken, or geometrically distorted small linear objects such as conductors and insulator strings during oblique photogrammetry reconstruction, thus ensuring the geometric accuracy of transmission line facilities. Next, the construction design drawings are processed to model temporary construction facilities and the transmission line to be built, resulting in 3D models of the temporary construction facilities and the transmission line to be built. Therefore, a model of the transmission line to be built can be generated based on the design drawings, and a model of the protective facilities (including catenary models, protective net models, auxiliary crossarm models, and crossing frame models) can be generated based on the netting scheme design drawings. Furthermore, models of construction machinery such as tensioners, traction machines, and cranes can be generated based on construction machinery parameter information, forming a complete 3D model of the temporary construction facilities and the transmission line to be built. This provides model inputs for the new line and temporary facilities for subsequent scene integration, solving the problem that the temporary construction facilities and the new line lack 3D models and cannot participate in spatial analysis in the simulation scene. Next, multi-source model scene integration processing is performed on the aforementioned terrain 3D model, the aforementioned transmission line and crossing structure 3D model, the aforementioned transmission line to be built model, and the aforementioned temporary construction facility 3D model to obtain a scene integration model. Thus, the various models generated in the above steps can be integrated together through multi-source scene integration to obtain a scene integration model. Finally, based on preset construction netting design parameters and preset construction scene configuration information, the above scene integration model undergoes netting model parameterization configuration processing to obtain a 3D simulation model of the netting.Therefore, based on preset construction netting design parameters (including crossing location coordinates, netting span, netting width, netting height, catenary tension value, catenary sag value, anchor point coordinates, etc.), the catenary cable model, protective netting model, auxiliary crossarm model, and crossing frame model can be automatically generated in the scene integration model, establishing a parametric linkage between the geometry of the netting structure and the design parameters. When the crossing span or construction conditions change, only the corresponding design parameter values ​​need to be modified, and the netting model can be automatically reconstructed. This avoids the problem of significantly increased time consumption for multi-scheme comparison and difficulty in covering extreme working conditions due to the lack of a parameter-driven mechanism and the need to manually adjust the model position and orientation when construction conditions change. It realizes rapid configuration, dynamic adjustment, and reproducible, verifiable, and recalculated engineering design-level simulation of netting schemes. Furthermore, by acquiring complementary geometric and textural information from multi-source survey data, achieving precise alignment of multi-source data through coordinate unification and spatial registration, resolving geometric distortion issues of small linear objects such as conductors through automated modeling based on line data, achieving efficient organization and smooth rendering of large scenes through multi-source model scene integration, and enabling reconstruction during design changes through parametric configuration of the netting model, the 3D simulation model is upgraded from a static visual composite that can only be viewed to an engineering design-level simulation tool that is reproducible, verifiable, and recalculateable. While ensuring spatial accuracy, it supports systematic simulation verification of netting layout and line laying conditions, thereby improving the engineering usability of the constructed 3D simulation model and ensuring that the 3D simulation model plays a substantial role in construction safety verification.

[0077] Further reference Figure 2 As an implementation of the methods shown in the figures, this disclosure provides some embodiments of a three-dimensional simulation model construction device for construction netting. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0078] like Figure 2As shown, the three-dimensional simulation model construction device 200 for construction netting in some embodiments includes: an acquisition unit 201, a first processing unit 202, a first generation unit 203, a second processing unit 204, a second generation unit 205, a third processing unit 206, a scene integration unit 207, and a parameterization configuration unit 208. The acquisition unit 201 is configured to acquire multi-source survey data of the construction site, vectorized data of the original route, and construction design drawings. The multi-source survey data includes point cloud data and oblique imagery of the construction site. The first processing unit 202 is configured to perform 3D real-scene model construction processing on the oblique imagery of the construction site to obtain a real-scene texture 3D model. The first generation unit 203 is configured to generate aligned point cloud data, aligned real-scene texture 3D model, and aligned route data based on the construction site point cloud data, the real-scene texture 3D model, and the original route vectorized data. The second processing unit 204 is configured to perform terrain and texture fusion modeling processing on the aligned point cloud data and the oblique imagery of the construction site to obtain a terrain 3D model. The second generation unit 205 is configured to perform terrain and texture fusion modeling processing on the aligned point cloud data and the oblique imagery of the construction site to obtain a terrain 3D model. The device generates a 3D model of the transmission line and its crossings using the aforementioned 3D terrain model, aligned real-world texture 3D model, aligned point cloud data, and aligned line data. The third processing unit 206 is configured to model the temporary construction facilities and the transmission line to be built using the aforementioned construction design drawings, obtaining a 3D model of the temporary construction facilities and a model of the transmission line to be built. The scene integration unit 207 is configured to perform multi-source model scene integration processing on the aforementioned 3D terrain model, the aforementioned 3D model of the transmission line and its crossings, the aforementioned model of the transmission line to be built, and the aforementioned 3D model of the temporary construction facilities, obtaining a scene integration model. The parameterization configuration unit 208 is configured to perform parameterization configuration processing on the aforementioned scene integration model based on preset construction netting design parameters and preset construction scene configuration information, obtaining a 3D simulation model of the netting. It is understood that the units described in the device 200 are related to the reference... Figure 1 The steps in the method described above correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the device 200 and the units contained therein, and will not be repeated here.

[0079] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0080] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0081] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0082] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0083] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0084] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0085] A computer-readable medium may be contained within an electronic device or may exist independently, not assembled into the electronic device. The computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire multi-source survey data of the construction site, vectorized data of the existing route, and construction design drawings, wherein the multi-source survey data of the construction site includes point cloud data and oblique imagery of the construction site; perform 3D reality model construction processing on the oblique imagery of the construction site to obtain a real-scene texture 3D model; based on the point cloud data of the construction site, the real-scene texture 3D model, and the vectorized data of the existing route, generate aligned point cloud data, aligned real-scene texture 3D model, and aligned route data; and perform terrain and texture fusion modeling processing on the aligned point cloud data and the oblique imagery of the construction site to obtain… A 3D terrain model is generated. Based on the aforementioned 3D terrain model, the aligned real-world texture 3D model, the aligned point cloud data, and the aligned line data, a 3D model of the transmission line and its crossings is generated. The construction design drawings are then processed to model temporary construction facilities and the transmission line to be built, resulting in a 3D model of the temporary construction facilities and the transmission line to be built. Multi-source model scene integration processing is performed on the aforementioned 3D terrain model, the aforementioned 3D model of the transmission line and its crossings, the aforementioned transmission line to be built, and the aforementioned 3D model of the temporary construction facilities, resulting in a scene integration model. Based on preset construction netting design parameters and preset construction scene configuration information, the scene integration model is parametrically configured to obtain a 3D netting simulation model. Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0086] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0087] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit, a first processing unit, a first generation unit, a second processing unit, a second generation unit, a third processing unit, a scene integration unit, and a parameterization configuration unit. The names of these units do not necessarily limit the specific unit; for example, the first processing unit may also be described as "a unit that performs three-dimensional real-scene model construction processing on the aforementioned tilted images of the construction site to obtain a three-dimensional real-scene texture model."

[0088] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0089] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of technical features, but should also cover other technical solutions formed by arbitrary combinations of technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for constructing a three-dimensional simulation model of construction site fencing, comprising: Acquire multi-source survey data of the construction site, vectorized data of the original line, and construction design drawings. The multi-source survey data of the construction site includes point cloud data and tilted images of the construction site. The tilted image of the construction site is processed to construct a three-dimensional real-scene model, resulting in a three-dimensional real-scene texture model; Based on the point cloud data of the construction site, the real-scene texture 3D model and the original line vectorization data, the aligned point cloud data, the aligned real-scene texture 3D model and the aligned line data are generated. The aligned point cloud data and the tilted image of the construction site are subjected to terrain and texture fusion modeling to obtain a three-dimensional terrain model; Based on the terrain 3D model, the aligned real-scene texture 3D model, the aligned point cloud data, and the aligned line data, a 3D model of the power transmission line and crossings is generated. The construction design drawings are modeled to create temporary construction facilities and transmission lines to be built, resulting in a three-dimensional model of the temporary construction facilities and a model of the transmission lines to be built. Multi-source model scene integration processing is performed on the terrain 3D model, the power transmission line and crossing object 3D model, the power transmission line to be built model and the temporary construction facility 3D model to obtain a scene integration model; Based on the preset construction netting design parameters and preset construction scenario configuration information, the scenario integration model is parametrically configured to obtain a three-dimensional simulation model of the netting.

2. The method of claim 1, wherein, The process of constructing a three-dimensional real-scene model from the tilted image of the construction site to obtain a real-scene texture three-dimensional model includes: Aerial triangulation was performed on the tilted images of the construction site to obtain the spatial attitude information of the images. Based on the spatial pose information of the image, dense point cloud construction processing is performed on the tilted image of the construction site to obtain real-scene color point cloud data. The real-scene color point cloud data is processed by triangulation to obtain a three-dimensional geometric white model; Based on the tilted image of the construction site, texture mapping processing is performed on the three-dimensional geometric white model to obtain a real-scene textured three-dimensional model.

3. The method according to claim 1, wherein, The process of generating aligned point cloud data, aligned real-scene texture 3D model, and aligned line data based on the construction site point cloud data, the real-scene texture 3D model, and the original line vectorization data includes: The point cloud data of the construction site is denoised to obtain denoised point cloud data; The denoised point cloud data is subjected to ground point separation processing to obtain ground point cloud data and non-ground point cloud data; The non-ground point cloud data is classified to obtain classified point cloud data; The ground point cloud data, the non-ground point cloud data, and the classified point cloud data are identified as processed point cloud data. The processed point cloud data, the real-scene texture 3D model, and the original line vectorized data are subjected to multi-source data coordinate unification and spatial registration processing to obtain aligned point cloud data, aligned real-scene texture 3D model, and aligned line data.

4. The method according to claim 1, wherein, The aligned point cloud data includes ground point cloud data, and the terrain model is constructed by processing the aligned point cloud data and the oblique image of the construction site to obtain a three-dimensional terrain model, including: The oblique images of the construction site are processed to construct a digital orthophoto map, thereby obtaining a digital orthophoto map of the construction site. Based on the ground point cloud data included in the aligned point cloud data, ground elevation data is generated; The ground elevation data is processed to construct a raster terrain, resulting in a raster terrain grid. Texture mapping processing is performed on the raster terrain grid and the digital orthophoto of the construction site to obtain a three-dimensional terrain model.

5. The method according to claim 1, wherein, The aligned point cloud data includes ground point cloud data, non-ground point cloud data, and classified point cloud data. Based on the terrain 3D model, the aligned real-world texture 3D model, the aligned point cloud data, and the aligned line data, a 3D model of the transmission line and its crossings is generated, including: Based on the aligned line data, the classified point cloud data is processed to construct a tower model to obtain the tower model. Based on the aligned line data, the classified point cloud data is subjected to conductor and ground wire point cloud extraction processing to obtain a conductor and ground wire point cloud dataset. The conductor and ground wire point cloud dataset is processed to obtain conductor and ground wire models. Based on the aforementioned 3D terrain model, the classified point cloud data is processed to extract road surface point clouds, resulting in a road surface point cloud dataset. Based on the real-scene texture 3D model, the road surface point cloud dataset is processed to construct a road surface model, thus obtaining the road surface model. Based on the real-scene texture 3D model and the terrain 3D model, the classified point cloud data is processed to construct a river model, thus obtaining a river model. Based on the real-scene texture 3D model, the classified point cloud data is processed to construct an overhead line model to obtain the overhead line model. The tower model, conductor model, ground wire model, road surface model, river model, and overhead line model are defined as three-dimensional models of transmission lines and crossings.

6. The method according to claim 1, wherein, The construction design drawings include design drawings of the transmission line to be built and design drawings of the grid enclosure scheme, as well as the modeling processing of the construction design drawings for temporary construction facilities and the transmission line to be built, to obtain a three-dimensional model of the temporary construction facilities and a model of the transmission line to be built, including: Based on the design drawings of the transmission line to be built, a model of the transmission line to be built is generated; The design drawings of the net enclosure scheme are processed to construct a net enclosure protection facility model, thereby obtaining the net enclosure protection facility model; Obtain parameter information for each construction machine; The parameter information of each construction machine is processed to construct a construction machine model, thereby obtaining the model of each construction machine. The protective netting model and the various construction machinery models are defined as three-dimensional models of temporary construction facilities.

7. A three-dimensional simulation model construction device for construction site fencing, comprising: The acquisition unit is configured to acquire multi-source survey data of the construction site, vectorized data of the original line, and construction design drawings, wherein the multi-source survey data of the construction site includes point cloud data of the construction site and oblique images of the construction site. The first processing unit is configured to perform three-dimensional real-scene model construction processing on the tilted image of the construction site to obtain a real-scene texture three-dimensional model. The first generation unit is configured to generate aligned point cloud data, aligned real-scene texture 3D model and aligned line data based on the construction site point cloud data, the real-scene texture 3D model and the original line vectorization data. The second processing unit is configured to perform terrain and texture fusion modeling processing on the aligned point cloud data and the construction site tilt image to obtain a three-dimensional terrain model. The second generation unit is configured to generate a three-dimensional model of power transmission lines and crossings based on the terrain three-dimensional model, the aligned real-scene texture three-dimensional model, the aligned point cloud data, and the aligned line data. The third processing unit is configured to perform modeling processing on the construction design drawings to obtain a three-dimensional model of the temporary construction facilities and the transmission line to be built. The scene integration unit is configured to perform multi-source model scene integration processing on the terrain 3D model, the power transmission line and crossing object 3D model, the power transmission line to be built model and the construction temporary facility 3D model to obtain a scene integration model; The parameterized configuration unit is configured to perform parameterized configuration processing on the scene integration model based on preset construction netting design parameters and preset construction scenario configuration information to obtain a three-dimensional simulation model of the netting.

8. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6.

9. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 6.

Images