Methods for Editing and Customizing Construction 3D Reports Based on BIM+GIS

Abstract

This invention relates to the field of information technology and provides a method for editing and customizing templates for 3D construction reports based on BIM+GIS. The method includes: collecting user perspective change information to extract location coordinates and direction vectors to obtain the current visible area; performing resource scheduling and priority ranking on the 3D scene based on this area and its changes to determine the loading sequence; obtaining the loading sequence to perform memory allocation on resource blocks, and unloading low-priority area resource blocks when the limit is exceeded to obtain an optimized resource configuration; extracting state parameters including lighting effects and location data before scene switching based on the optimized resource configuration, and generating a smooth transition path for scene switching through smooth interpolation calculation. This invention achieves dynamic and fine-grained scheduling of 3D scene resources and smooth and continuous transition of scene states, improving the fluency and coherence of 3D visualization.

Description

Technical Field

[0001] This application relates to the field of information technology, and in particular to a method for editing and customizing 3D construction reports based on BIM+GIS. Background Technology

[0002] Currently, in the fields of urban planning and engineering management, 3D visualization technology has become an important means of showcasing project progress and design intent. However, with the continuous expansion of data scale and the increasing complexity of display needs, existing 3D reporting technologies still have many shortcomings in terms of smoothness and response speed during dynamic display processes.

[0003] Especially in business scenarios such as construction reports that require frequent switching between scenes and perspectives, existing technologies often struggle to balance the rapid loading of large-scale scene resources with system stability. On one hand, due to a lack of refined management of scene resources, when the user's perspective changes rapidly, the system cannot accurately determine when to load and unload resources, leading to resource waste or loading delays, and consequently, screen stuttering. On the other hand, when switching between different construction stages or displayed content, existing technologies often lack effective maintenance of scene state continuity, causing abrupt changes in parameters such as viewpoint position and lighting effects, disrupting the continuity of the visual experience. Therefore, how to achieve intelligent scheduling and smooth transition of scene resources while ensuring rendering performance has become a pressing technical problem to be solved. Summary of the Invention

[0004] This application provides a method for editing and customizing 3D construction reports based on BIM+GIS, which solves the technical problems of inaccurate resource scheduling and unsmooth transition of scene states in existing technologies.

[0005] To achieve the above objectives, this application adopts the following technical solution: Firstly, a method for editing and customizing 3D construction reports based on BIM+GIS is provided, including: Collect information on changes in the user's field of view, extract the current position coordinates and direction vector from the information, and obtain the current visible area range; Based on the current visible area and its changes, resource scheduling and priority sorting are performed on the 3D scene to determine the set of resource blocks to be loaded and the sorted loading sequence. Obtain the sorted loading sequence, perform memory allocation operations on the resource blocks, and if the system memory limit is exceeded, unload the resource blocks in the low-priority area to obtain the optimized resource configuration; Based on the optimized resource configuration, the state parameters before scene switching are extracted, and a smooth transition path for scene switching is generated by smooth interpolation calculation; the state parameters include lighting effects and position data.

[0006] Based on the above technical solutions, the construction 3D report editing and template customization method based on BIM+GIS provided in this application accurately determines the current visible area range by collecting changes in user perspective in real time and extracting position coordinates and direction vectors. Then, dynamic resource scheduling and priority sorting are performed based on this range and its changes, ensuring the accuracy and efficiency of resource loading. At the same time, resource configuration is optimized through memory allocation and low-priority resource unloading mechanisms, and state parameters are extracted based on the optimized configuration for smooth interpolation calculation, realizing continuous state and smooth visual transition during scene switching, thereby solving the problems of inaccurate resource scheduling and unsmooth state transition during dynamic switching.

[0007] In conjunction with the first aspect above, in one possible implementation, the collection of user perspective change information includes: The user's head deflection or gaze displacement data is collected by the posture sensing module to obtain a viewpoint change dataset. Extract the current position coordinates and the direction vector from the viewpoint change dataset to determine the user's current position point and direction reference in space; If the current position coordinates deviate from the preset coordinate system, the current position coordinates are standardized through calibration to obtain calibrated position data.

[0008] By accurately collecting viewing angle change data through the attitude sensing module and calibrating and standardizing coordinates with deviations, the accuracy of position data and orientation reference is ensured, providing a reliable data foundation for subsequent accurate calculation of the visible area.

[0009] In conjunction with the first aspect above, in one possible implementation, the resource scheduling of the three-dimensional scene includes: using a spatial region partitioning strategy to divide the three-dimensional scene into blocks to obtain multiple scene content units; Analyze the correspondence between each scene content unit and the current visible area range to determine the set of target units located within the current visible area range; The set of resource blocks to be loaded is determined from the set of target units.

[0010] By dividing the 3D scene into blocks using a spatial region partitioning strategy and filtering the target unit set based on the visible area range, spatial granular management and precise matching of scene resources are achieved, avoiding redundant loading of irrelevant resources.

[0011] In conjunction with the first aspect above, in one possible implementation, the step of using a spatial region division strategy to divide the three-dimensional scene into blocks further includes: using spatial perception technology to collect environmental data within the visible area in real time based on the changes in the visible area, thereby obtaining a dynamic information set of the changes in the area. The grid distribution structure is dynamically reconstructed based on environmental data, and resource blocks that do not match regional changes are removed. The set of resource blocks to be loaded is updated through multi-dimensional data synchronization and priority sorting.

[0012] By using spatial perception technology to collect environmental data in real time and dynamically reconstruct the grid distribution, mismatched resource blocks are eliminated, enabling the resource set to dynamically and adaptively update as the visible area changes, thus improving the real-time performance and flexibility of resource scheduling.

[0013] In conjunction with the first aspect above, in one possible implementation, the priority ordering includes: For the unloaded blocks in the set of resource blocks that need to be loaded, a preset priority mechanism is used for preliminary classification to obtain the classified block partitioning results; Priority scoring is performed based on the distance of resource blocks from the viewpoint center and the resource volume, and unloaded blocks are processed in batches according to the scoring results; Batches exceeding the preset time limit are split to obtain the sorted loading sequence.

[0014] Unloaded blocks are classified and scored using a priority mechanism, and then processed in batches and timed out based on the scores. This ensures the rationality and balance of the loading sequence and avoids system lag caused by instantaneous loading peaks.

[0015] In conjunction with the first aspect above, in one possible implementation, the step of performing memory allocation operations on resource blocks, and unloading resource blocks in low-priority regions if the system memory limit is exceeded, includes: Obtain the sorted loading sequence and perform memory allocation operations on the resource blocks; The system monitors the memory level in real time. If the memory usage exceeds the system memory limit, the specific resource block that caused the memory overflow is locked. Based on a multi-dimensional priority strategy, resource blocks in the low-priority regions are selected for unloading, thereby releasing the corresponding memory space. If the restrictions are still not met after uninstallation, a secondary uninstallation mechanism is triggered to obtain the optimized resource configuration.

[0016] By monitoring memory levels in real time and locking overflowing resource blocks when limits are exceeded, combined with a multi-dimensional priority strategy to accurately unload low-priority resources and trigger secondary unloading when necessary, the system memory is always kept within a safe threshold, ensuring the stability of the rendering process.

[0017] In conjunction with the first aspect above, in one possible implementation, the step of extracting the state parameters before the scene switch and generating a smooth transition path for the scene switch through smooth interpolation includes: Obtain the state information before scene switching from the optimized resource configuration, classify and store the lighting effects and the position data to obtain the initial state dataset; Based on the position data in the initial state dataset, intermediate transition points are determined using the smooth interpolation calculation, and a transition frame sequence is generated. Based on the transition frame sequence and the gradual change requirements of the lighting and shadow effects, the inter-frame changes are adjusted through real-time processing technology to determine the distribution scheme of the switching timing. Based on the distribution scheme of switching timing, a smooth transition path for scene switching is constructed using a dynamic adaptation method.

[0018] The above solution classifies and stores lighting effects and position data, calculates transition frame sequences based on smooth interpolation, and adjusts the inter-frame changes in real time in combination with the requirements of lighting gradation and dynamically adapts to the switching timing, thus ensuring the continuity and naturalness of visual effects during scene switching.

[0019] In conjunction with the first aspect above, in one possible implementation, after generating the transition frame sequence, the method further includes: Obtain the correspondence between the effect transition sequence and the transition frame sequence; If the timestamp deviation between the two sequences exceeds a preset threshold, a time protocol tool is used to perform time axis calibration on the transition frame sequence to obtain synchronized transition data. Based on the synchronous transition data, the final smooth path is constructed, generating a complete scene switching transition scheme.

[0020] The above scheme ensures a high degree of synchronization between light and shadow changes and positional movement in the time dimension by calibrating the timestamp deviation between the effect transition sequence and the transition frame sequence, eliminating visual tearing and further improving the integrity of the transition scheme.

[0021] In conjunction with the first aspect above, in one possible implementation, after generating the smooth transition path for scene switching, the method further includes: Monitor rendering frame rate; If the rendering frame rate is lower than the preset threshold, the detail level of the resource blocks will be downgraded to obtain stable rendering performance.

[0022] By monitoring the rendering frame rate in real time and degrading the level of resource detail when the frame rate is too low, the system dynamically balances image quality and rendering performance, ensuring that the system can maintain stable interactive smoothness even under high load scenarios.

[0023] In conjunction with the first aspect above, in one possible implementation, after generating the smooth transition path for scene switching, the method further includes: Integrate content switching signals and apply update logic to the current scenario; A consistency analysis is performed on the updated scene content. If the state continuity judgment result does not meet the preset standard, the scene update part is processed to obtain the final dynamic display result.

[0024] By integrating content switching signals and applying update logic, and performing consistency analysis and transition processing on the updated scene, the continuity of the scene state after content update is ensured to meet preset standards, ultimately outputting a high-fidelity dynamic display result.

[0025] This application provides a method for editing and customizing 3D construction reports based on BIM+GIS. By collecting user perspective change information, extracting position coordinates and direction vectors, and dynamically determining the visible area, this invention employs grid partitioning and spatial awareness technologies to flexibly adjust resource block loading strategies. Combined with a priority scheduling mechanism, it optimizes the loading sequence and memory allocation, ensuring efficient resource configuration. Simultaneously, this invention generates smooth transition paths through smooth interpolation calculation of transition frame sequences and utilizes real-time rendering processing and dynamic adaptation technologies to guarantee the continuity and stability of scene transitions. Furthermore, this invention monitors the rendering frame rate and dynamically adjusts the level of resource detail to ensure stable rendering performance, ultimately achieving seamless integration of content switching and state continuity. Its core technological effect lies in improving the dynamic display smoothness and resource management efficiency of virtual reality scenes, providing users with an immersive experience.

[0026] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0027] Figure 1 A system architecture diagram of a processing system provided in this application embodiment; Figure 2 A flowchart illustrating the BIM+GIS-based method for editing and customizing construction 3D reports and templates, as provided in this application embodiment; Figure 3This is a schematic diagram of the process for collecting user perspective change information provided in an embodiment of this application; Figure 4 This is a schematic diagram of the process for obtaining a smooth transition path provided in an embodiment of this application. Detailed Implementation

[0028] In the description of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0029] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0030] The method for editing and customizing construction 3D reports based on BIM+GIS provided in this application embodiment can be applied to, for example... Figure 1 In the processing system shown, such as Figure 1 As shown, the communication system includes: a view acquisition terminal 10, a resource scheduling server 20, a memory management server 30, a rendering processing server 40, and a data storage server 50.

[0031] A perspective acquisition terminal is used to collect information on changes in the user's perspective, extract the current position coordinates and direction vector from the information, and obtain the current visible area range. The resource scheduling server is used to perform resource scheduling and priority sorting on the 3D scene based on the current visible area range and its changes, and to determine the set of resource blocks to be loaded and the sorted loading sequence. A memory management server is used to obtain the sorted loading sequence and perform memory allocation operations on the resource blocks; A rendering server is used to extract state parameters before scene switching based on the optimized resource configuration, and generate a smooth transition path for scene switching through smooth interpolation calculation. It is also used to obtain the correspondence between the effect transition sequence and the transition frame sequence. If the timestamp deviation between the two sequences exceeds a preset threshold, the transition frame sequence is calibrated on the time axis using a time protocol tool to obtain synchronous transition data. Based on the synchronous transition data, the final smooth path is constructed to generate a complete scene switching transition scheme. Additionally, it is used to monitor the rendering frame rate. If the rendering frame rate is lower than a preset threshold, the detail level of the resource block is downgraded to obtain stable rendering performance. It is also used to integrate content switching signals, apply update logic to the current scene, perform consistency analysis on the updated scene content, and if the state continuity judgment result does not meet the preset standard, the scene update part is transitioned to obtain the final dynamic display result.

[0032] The data storage server is used to store resource block data, scene content unit data, and resource management logs of the 3D scene, providing data support for resource scheduling, memory management, and rendering processing.

[0033] like Figure 2 As shown, this embodiment provides a method for editing and customizing templates for 3D construction reports based on BIM+GIS. This method constructs a complete dynamic display loop from perspective perception to visual presentation. Through the coordinated operation of four core steps, it fundamentally solves the pain point of simultaneous stuttering and jumps when dynamically switching large-scale 3D scenes.

[0034] Example 1 like Figure 1 As shown in the embodiments of this application, the method for editing and customizing construction 3D reports based on BIM+GIS includes: Step S100: Collect user viewpoint change information, extract current position coordinates and direction vector from the information, and obtain the current visible area range.

[0035] It should be noted that perspective change information is the source signal driving the dynamic flow of resources throughout the entire 3D scene. This step transforms the user's observational intent in the physical world into the visible boundaries of the digital space by continuously capturing changes in the user's spatial posture during interaction.

[0036] For example, when a user turns their head or moves their gaze in a construction reporting scenario, the system records these movements in real time and extracts precise 3D coordinate data representing the user's location and a direction vector indicating the direction of observation. Based on these coordinates and vectors, the system can construct a view frustum model in virtual space, thereby calculating the spatial area that the user can observe at any given moment, i.e., the current visible area.

[0037] It should be noted that although this embodiment uses the view frustum model as an example to calculate the visible area, in other embodiments, spatial geometric algorithms such as spherical projection or polyhedral bounding boxes can also be used to determine the visible range, as long as the user's visual coverage area can be accurately mapped. This step is a causal prerequisite for subsequent resource scheduling: only by accurately understanding where the user is "looking" and "about to look" can the system effectively decide "what data should be prepared," thereby avoiding the blind loading of massive amounts of irrelevant scene resources.

[0038] Step S200: Based on the current visible area range and its changes, perform resource scheduling and priority sorting on the 3D scene to determine the set of resource blocks to be loaded and the sorted loading sequence.

[0039] It should be noted that this step follows the visible area data output from step S100, transforming the static observation range into a dynamic resource preparation plan.

[0040] 3D scenes typically contain extremely large construction models and geographic information data; loading them all at once would inevitably cause system lag. Therefore, the system filters relevant resource blocks falling within the current visible area from the overall scene, forming a set of resource blocks that need to be loaded. More importantly, this step not only focuses on the current static area but also specifically considers changes in the visible area.

[0041] For example, when the system predicts that the user's viewpoint is rapidly moving towards a specific construction area, it will be taken into consideration for resource scheduling even if the area has not yet fully entered the current field of view. Based on this, the system prioritizes the selected set of resource blocks, placing those closer to the viewpoint center and contributing more to the current visual presentation first, generating a sorted loading sequence.

[0042] It's important to note that the criteria for prioritization are not limited to distance; they can also include visual occlusion relationships between resource blocks, the importance of business logic, and other factors, as long as the data that most significantly impacts the user's current experience is prepared first. This step directly addresses the pain point of "untimely loading," ensuring that every minor adjustment to the viewpoint is quickly supported by corresponding data.

[0043] Step S300: Obtain the sorted loading sequence, perform memory allocation operation on the resource blocks, and if the system memory limit is exceeded, unload the resource blocks in the low-priority area to obtain the optimized resource configuration.

[0044] It should be noted that memory optimization is an indispensable prerequisite for achieving smooth scene transitions. Even with a reasonable loading sequence, if system memory is exhausted due to excessive accumulation, subsequent rendering and interpolation calculations will still experience lag or even crashes due to resource scarcity.

[0045] This step rigorously assesses the system's memory capacity before actually writing resource blocks from the loading sequence into memory. When it is determined that injecting a new resource will exceed the system's memory limit, the system does not simply refuse to load it, but instead adopts a dynamic replacement strategy: actively unloading resource blocks in low-priority regions.

[0046] For example, when the user's viewpoint has completely shifted to the details of the eastern area of ​​the construction site, the overall outline model located in the western area and hundreds of meters away from the current viewpoint is determined to be a low-priority area, and the memory it occupies will be released to make room for the high-precision component model in the eastern area.

[0047] It should be noted that the criteria for determining low-priority areas can be based on spatial distance, historical access frequency over time, or visibility probability of the rendering pipeline, as long as the impact of the unloaded resources on the current visual continuity is minimized.

[0048] Through this dynamic allocation and unloading game, the system always maintains a healthy memory level and obtains optimized resource allocation, providing a solid guarantee of computing power and space for subsequent smooth visual computing.

[0049] Step S400: Based on the optimized resource configuration, extract the state parameters before scene switching, and generate a smooth transition path for scene switching through smooth interpolation calculation.

[0050] The status parameters include lighting effects and position data.

[0051] It should be noted that this step is the final visual presentation stage after resources are in place, and it aims to completely eliminate the abruptness of scene transitions.

[0052] In traditional processing methods, scene transitions often manifest as a sudden jump in viewpoint and abrupt changes in lighting conditions, giving users a strong sense of visual disjointedness. This step extracts the state parameters before the transition and packages and stores the lighting effects (such as ambient light intensity and shadow projection angle) and position data (such as camera spatial coordinates and observation target point).

[0053] Subsequently, during the switch to the new scene configuration, the system does not jump directly to the target state, but uses smooth interpolation calculation to generate a series of continuous intermediate transition states between the initial state and the target state, thereby constructing a smooth transition path.

[0054] For example, when switching from a panoramic view of the construction site to a detailed view of the foundation pit, the camera position moves frame by frame along the interpolation path, and the lighting effect gradually transitions from the uniform sunlight in the panoramic view to local point light source illumination inside the foundation pit.

[0055] It should be noted that the specific algorithm for smooth interpolation calculation can be linear interpolation to achieve a uniform transition, or it can be a Bezier curve or a B-spline curve to achieve a natural transition with acceleration and deceleration physical inertia, as long as discrete jumps in state parameters are avoided. This step relies on the optimized resource configuration provided in step S300—only when all the lighting and geometric resources required for interpolation calculation are properly prepared in memory will the generation of the smooth transition path not be interrupted due to data loss.

[0056] In summary, step S100 provides accurate visual perception input, step S200 transforms perception into an efficient resource preparation sequence, step S300 removes the memory capacity obstacle for resource preparation, and step S400 completes the smooth sublimation of vision on the basis of resource guarantee.

[0057] The S100-S400 architecture forms an inseparable dynamic display loop. Without the driving force of perspective changes, resource scheduling will lose direction and fall into blind loading; without the prior guarantee of memory optimization, smooth interpolation will become empty talk due to computing power exhaustion; without the final smooth transition, even if resource loading is fast, the visual experience will still be ruined by abrupt jumps.

[0058] Through the aforementioned closed-loop mechanism of causal synergy, this invention can simultaneously overcome the two major pain points of stuttering and abrupt transitions, achieving smooth and coherent dynamic display of 3D scenes. The technical solutions described in this embodiment are merely illustrative and not restrictive. Those skilled in the art can make various modifications and substitutions based on the inspiration of this embodiment. As long as the core logic follows the closed-loop synergy from viewpoint-driven to resource scheduling, memory optimization, and smooth transition, all such modifications and substitutions should fall within the protection scope of this invention.

[0059] Example 2 This embodiment, based on the dynamic display closed loop constructed in Embodiment 1, further refines the accuracy requirements for front-end data acquisition. In the dynamic display of 3D scenes, the accuracy of viewpoint perception and the rationality of scene space division are the cornerstones determining the quality of subsequent resource scheduling. This embodiment addresses this by introducing specific hardware binding and algorithm implementation.

[0060] For step S100, user viewpoint change information is collected, and the current position coordinates and direction vector are extracted from the information to obtain the current visible area range. This embodiment elaborates on the details of viewpoint acquisition and calibration.

[0061] Specifically, a posture sensing module collects data on the user's head rotation or gaze displacement to obtain a viewpoint change dataset. The posture sensing module acts as a bridge connecting the physical world's observational intent with digital spatial computation logic. It should be noted that the specific type of posture sensing module can include at least one of an inertial measurement unit (IMU), an eye tracker, or a spatial positioning base station. For example, in construction reporting scenarios based on VR devices, a built-in IMU is typically used to collect high-frequency data on the user's head rotation angle and acceleration; while in high-end desktop interaction scenarios, an eye tracker can be introduced to capture pupil gaze displacement, thereby achieving more refined viewpoint tracking. These raw physical motion signals, after digital conversion and filtering / denoising, are aggregated to form a viewpoint change dataset, providing raw material for subsequent spatial mapping.

[0062] Subsequently, the current position coordinates and direction vector are extracted from the viewpoint change dataset to determine the user's current position and orientation reference in space. The system analyzes the underlying spatial mapping parameters in the dataset, performs spatial calculations on the raw signals transmitted from the sensors, accurately identifies the user's specific horizontal and vertical coordinates in the current three-dimensional coordinate system, and calculates the direction vector representing the user's line of sight by combining the attitude angle data. This step refines the chaotic sensor signals into a rigorous mathematical reference.

[0063] However, sensor hardware inevitably produces measurement errors during long-term operation or when subjected to environmental magnetic field interference. Therefore, if the current position coordinates deviate from the preset coordinate system, a calibration process is used to standardize the current position coordinates, obtaining calibrated position data. The core mathematical logic of the calibration process lies in zero-point drift compensation and coordinate system alignment.

[0064] For example, when the IMU outputs a small displacement value even when stationary due to temperature drift, the system calculates the current zero-point offset using a sliding window mean filtering algorithm and subtracts it from the real-time extracted coordinates. Simultaneously, if the user's initial orientation deviates from the preset true north direction, the system uses a rotation matrix transformation to forcibly align the local coordinates to the global preset coordinate system. This standardized processing ensures high accuracy and consistency of location data over long-term interactions, avoiding distortion in the calculation of the visible area due to viewpoint drift, thus providing an accurate and reliable input source for resource scheduling.

[0065] After establishing the precise visible area range, in step S200, resource scheduling and priority sorting are performed on the 3D scene based on the current visible area range and its changes, and the set of resource blocks to be loaded and the sorted loading sequence are determined. This embodiment elaborates on the static scene block division logic.

[0066] In some implementations, a spatial region partitioning strategy is used to divide the 3D scene into blocks, resulting in multiple scene content units. When dealing with a construction 3D scene containing massive amounts of BIM components and GIS terrain data, it is necessary to discretize it into independently schedulable spatial granularity. Specific algorithms for spatial region partitioning strategies can include at least one of quadtree partitioning, KD-tree partitioning, or uniform grid partitioning.

[0067] For example, for construction sites with flat terrain and relatively uniform component distribution, uniform grid partitioning can quickly divide the scene into regular cubic units with low computational cost. However, for complex scenes with undulating terrain or significant differences in component density, quadtree or KD-tree algorithms are more suitable for adaptive partitioning. This allows the densely populated core construction area to be divided into finer units, while the sparsely populated suburban areas are retained as larger units. This algorithm-driven approach ensures that the scene's block topology closely matches the actual business space characteristics, avoiding the memory waste or precision loss that can result from uniform partitioning.

[0068] After scene segmentation, the system analyzes the correspondence between each scene content unit and the current visible area to determine the set of target units located within the current visible area. The system then uses spatial topology analysis technology to calculate the intersection weights between the bounding boxes of each scene content unit and the view frustum of the current visible area in real time.

[0069] For example, when the bounding box of a cell falls completely inside the view frustum, its intersection weight is 1, and it is directly included in the target cell set. When a cell only intersects the edge of the view frustum, the system assigns a corresponding weight based on the proportion of the intersection volume and determines whether to include it in the set based on a preset threshold. This precise filtering based on spatial topology eliminates redundant cells that are geographically close but actually obscured by terrain or buildings, ensuring the purity of the set.

[0070] Finally, the set of resource blocks to be loaded is determined from the target unit set. The system extracts the metadata and resource index of each unit in the target unit set, maps them to the corresponding geometry, texture, and lighting resource blocks in the underlying database, and generates a list of pending items that are directly loaded into memory.

[0071] By mapping spatial topology to resource indexes, the abstract spatial visibility determination is transformed into a concrete resource loading task, completing the key closed loop from perception to scheduling in the front-end data acquisition defense line. The hardware binding and algorithm sinking described in this embodiment are only illustrative. Those skilled in the art can select other types of sensors or spatial partitioning algorithms based on the specific device form and scene characteristics, as long as their core logic follows the principles of accurate calibration and adaptive block division, they should all fall within the protection scope of this invention.

[0072] Example 3 This embodiment, based on the dynamic display closed loop constructed in Embodiment 1, introduces advanced branch limitations for the resource scheduling and priority ranking step in S200. In 3D scenes, especially in BIM+GIS-based construction reporting scenarios, user perspective switching is often sudden and frequent, and static spatial division and simple loading order are insufficient to meet the needs of such dynamic changes. This embodiment, by introducing a dynamic reconstruction mechanism and a refined priority scoring algorithm, jointly constructs a dual defense against the core pain point of "untimely loading".

[0073] Regarding step S200, based on the current visible area range and its changes, resource scheduling and priority sorting are performed on the 3D scene to determine the set of resource blocks to be loaded and the sorted loading sequence. This embodiment elaborates on the dynamic adjustment mechanism therein.

[0074] In some implementations, spatial perception technology is used to collect environmental data in real time within the visible area based on changes in the visible area, resulting in a dynamic set of information on the changes in the area. The environmental data collected by spatial perception technology not only includes the user's current absolute displacement coordinates and static rotation angle, but more importantly, it includes dynamic physical quantities that characterize the trend of change, such as the linear velocity, angular velocity, and acceleration of the viewpoint movement.

[0075] It should be noted that although this embodiment uses displacement and rotational speed as examples of environmental data dimensions, other embodiments may also introduce pupil tremor frequency or inertial prediction vectors of head posture from eye tracking, as long as these dimensions can reflect the impending shift in the user's viewpoint in advance. Proactive data acquisition is a causal prerequisite for dynamic reconstruction: only by predicting the direction of the viewpoint can the system prepare resources along the way in advance.

[0076] Subsequently, the grid distribution structure is dynamically reconstructed based on environmental data, and resource blocks that do not match regional changes are removed. The triggering conditions for dynamic grid reconstruction are not arbitrary continuous calculations, but rather defined acceleration thresholds and directional deflection angle thresholds.

[0077] For example, when the user's visual movement acceleration is detected to exceed 2 m / s² 2 If the angular velocity of the deflection exceeds 15° / s, it is determined that the user is making a rapid viewpoint switch. Only then will the dynamic reconstruction of the grid be triggered to avoid unnecessary waste of computing power when browsing slowly.

[0078] During the reconstruction process, the system adjusts the originally uniformly distributed grid structure into a non-uniform focused grid based on the predicted viewpoint landing point. This increases the grid density in the predicted landing point area, thereby containing more refined scene content units. Simultaneously, the criterion for removing mismatched resource blocks uses a spatial trajectory intersection test: resource blocks that are far from the predicted landing point along the tangent direction of the current motion trajectory and move in the opposite direction, even if they are still at the edge of the visible area, will be judged as "mismatched" and removed.

[0079] This elimination logic provides an alternative solution to prevent circumvention. Instead of relying solely on the static fact that resource blocks are within the current visible area, it dynamically eliminates them based on predictions of movement trends, thereby effectively preventing the problem of old area resources lingering and occupying bandwidth due to users rapidly shifting their view.

[0080] Finally, through multi-dimensional data synchronization and priority sorting, the set of resource blocks to be loaded is updated. The system performs multi-dimensional data synchronization between the reconstructed grid cells and the cells retained after removal, ensuring the consistency of spatial topology and resource index, thereby outputting an updated set of resource blocks that dynamically changes with the viewpoint.

[0081] After establishing the dynamically updated set of resource blocks, this embodiment further refines the specific algorithm for priority sorting to ensure the rationality and balance of the loading sequence.

[0082] For the unloaded blocks in the set of resource blocks that need to be loaded, a preset priority mechanism is used for preliminary classification to obtain the classified block partitioning results. The preliminary classification is based on the business attributes and visual dependencies of the resource blocks.

[0083] For example, in a construction reporting scenario, BIM components of core construction nodes (such as tower crane operation areas and deep foundation pit areas) are classified as "critical business blocks," while distant GIS terrain maps are classified as "background environment blocks." This classification assigns different weighting coefficients to subsequent scoring calculations, ensuring that core business content has an absolute advantage in loading competition.

[0084] Subsequently, priority scores are assigned based on the distance of resource blocks from the viewpoint center and the resource volume, and unloaded blocks are processed in batches according to the score results.

[0085] The specific calculation formula for the priority score adopts a weighted scoring model based on distance and volume, and its mathematical expression is as follows: Where P is the priority score of the resource block, D is the Euclidean distance from the center of the resource block's bounding box to the center of the current viewpoint, and V is the data volume of the resource block. The system's preset maximum volume threshold for a single block. This is the distance weighting coefficient. This is the volume weighting coefficient.

[0086] It should be noted that although the above linear weighting formula is given as an example in this embodiment, exponential decay functions (such as...) can also be used in other embodiments. ;in, This uses a pre-set priority coefficient to more sharply differentiate the priority of resources based on their distance, as long as the core logic of giving higher scores to resources that are closer and larger in size is achieved. In this formula, the distance term uses the reciprocal form (1 / D), ensuring that resources near the viewpoint center receive extremely high scores, guaranteeing that the details that most affect the current visual experience are loaded first; the volume term uses a ratio form (…). This is to prioritize loading core components with large data volumes and long loading times, given the same distance, thus avoiding waiting caused by large resources only starting to load at the last stage. Based on the scoring results, unloaded blocks are divided into high, medium, and low priority batches, and loaded sequentially according to the batches. This batch processing mechanism transforms the originally disordered concurrent loading requests into a serial batch pipeline, effectively avoiding the impact of instantaneous loading peaks on system I / O bandwidth.

[0087] Finally, batches exceeding the preset time limit are split to obtain the sorted loading sequence. The threshold setting is based on the system's rendering frame time budget and network throughput.

[0088] For example, the system's preset single-batch loading time limit is 50ms, which corresponds to one frame in a 20fps frame budget. If a high-priority batch contains multiple large BIM component models, and its estimated total loading time exceeds the 50ms limit, the system will split the batch along the resource block boundaries, dividing it into multiple sub-batches that meet the time limits, and interleaving them into the rendering intervals of subsequent frames. This timeout splitting mechanism ensures that no loading batch will block the main rendering thread for an extended period, thus proving that this sorting mechanism can fundamentally avoid screen stuttering caused by instantaneous loading peaks, achieving a smooth interleaving of loading and rendering tasks.

[0089] The dynamic reconstruction and scoring algorithm described in this embodiment is only illustrative. Those skilled in the art can select other types of prediction models or scoring functions based on specific hardware bandwidth and scene characteristics, as long as their core logic follows dynamic elimination based on motion trends and balanced batching based on distance and volume.

[0090] Example 4 This embodiment, based on the dynamic display closed loop constructed in Embodiment 1, introduces advanced branch limitations for the memory allocation and unloading process in step S300. In large-scale 3D scenes, especially in BIM+GIS-based construction reporting scenarios, system memory capacity is often the physical bottleneck limiting rendering performance and smooth transitions. This embodiment constructs a fundamental guarantee to ensure stable system operation by introducing refined memory level control and a secondary unloading defense.

[0091] For step S300, the sorted loading sequence is obtained, and memory allocation operation is performed on the resource blocks. If the system memory limit is exceeded, the resource blocks in the low-priority area are unloaded to obtain the optimized resource configuration. This embodiment elaborates on the memory management sub-process.

[0092] In some implementations, the sorted loading sequence is obtained, and memory allocation operations are performed on the resource blocks. After receiving the loading sequence after priority scoring and batch splitting, the system starts a deep memory scheduler. Through the built-in traversal processing mechanism, the system scans each resource block in the sequence one by one, extracts its data volume and storage requirement characteristics, and, combined with preset memory allocation criteria, pre-matches the corresponding physical storage address for each resource block, thereby constructing an initial allocation scheme.

[0093] It should be noted that the specific methods of memory allocation operations can include contiguous memory allocation, paging memory management, or memory pool mapping, as long as the reasonable arrangement of resource blocks in memory space can be ensured.

[0094] Subsequently, the system memory level is monitored in real time. If memory usage exceeds the system memory limit, the specific resource block causing the memory overflow is locked. The specific threshold parameter for system memory level monitoring is set to 80% of the upper limit of the total available system memory.

[0095] For example, in a rendering terminal with 16GB of available memory, when the total allocated and unallocated memory reaches 12.8GB, the system determines that the memory usage exceeds the system memory limit. At this point, the system does not simply randomly refuse to load; instead, it calls the underlying resource monitoring module to perform source analysis, accurately locates the specific resource blocks that caused the memory level to instantly exceed the limit, locks their metadata information and memory addresses, and marks them as overflow resource blocks, providing a clear target for subsequent replacement operations.

[0096] After locking the overflowing resource block, resource blocks in the low-priority region are selected and unloaded based on a multi-dimensional priority strategy, releasing the corresponding memory space. The dimensions for selecting low-priority regions in the multi-dimensional priority strategy not only include the conventional line-of-sight distance, but also deeply integrate occlusion relationships and the importance of business logic.

[0097] For example, in a construction reporting scenario, a component inside a foundation pit that is close to the center of the viewpoint but is completely obscured by a giant tower crane model in front has zero visual contribution. If only a single dimension is used for unloading judgment, it is very easy to cause the core visual area to be unloaded incorrectly.

[0098] To demonstrate the necessity of a multi-dimensional strategy, a comparative example is presented here: If the system unloads resources based solely on distance, when memory overflows, the system will prioritize unloading the resource block furthest from the viewpoint. However, in the instant a user quickly switches from a panoramic construction view to a detailed view of a specific area, the distant GIS terrain outline, though far away, is the core background maintaining the user's spatial orientation; conversely, some nearby, obscured temporary scaffolding components, though close, are of no use to the current visual presentation. If unloading is based solely on distance, distant terrain outlines will be mistakenly unloaded, causing a sudden black hole in the scene background, while useless nearby scaffolding will still occupy memory. This mistaken unloading directly disrupts the visual integrity of the scene.

[0099] The multi-dimensional priority strategy adopted in this embodiment can accurately identify resource blocks that are "near but blocked" or "far but indispensable" by comprehensively evaluating the line-of-sight distance, occlusion relationship and business attributes. This allows for the precise elimination of visual redundancy during unloading, achieving the optimal balance between memory release and visual fidelity.

[0100] If the restrictions are still not met after uninstallation, a secondary uninstallation mechanism is triggered to obtain the optimized resource configuration.

[0101] The trigger condition for the secondary unloading mechanism is set as follows: after the first round of low-priority region unloading based on the multi-dimensional priority strategy is completed, the system checks the memory level again. If it finds that the remaining available memory is still insufficient to accommodate the locked overflow resource blocks, that is, the memory usage is still higher than the upper limit threshold of 80%, then the secondary unloading is triggered immediately.

[0102] The execution logic of the second-round unloading differs from the coarse-grained region-level unloading of the first round. Instead, it employs a more aggressive fine-grained component-level unloading and degradation retention strategy. Specifically, the system performs in-depth analysis of the medium-priority resource blocks currently retained in memory. Instead of unloading an entire region, it strips away non-core auxiliary components within that region (such as temporary fencing or non-critical marking lines in a construction site), retaining only the core framework and low-precision outer shell. This allows for the extraction of sufficient memory space to accommodate overflowing resource blocks with minimal visual loss. Through this progressive dual-layer defense, the system ensures that memory levels can be forcibly pushed back below a safe threshold under any extremely complex scenario switching, ultimately achieving a stable and optimized resource configuration and clearing the final physical obstacle for subsequent smooth interpolation calculations.

[0103] The threshold parameters and unloading logic described in this embodiment are for illustrative purposes only. Those skilled in the art can adjust the upper limit of the water level or introduce more dimensional priority strategies based on the specific hardware configuration and scenario complexity, as long as the core logic follows the principle of multi-dimensional evaluation and progressive unloading of secondary defenses. All of these should fall within the protection scope of this invention.

[0104] Example 5: This embodiment, based on the dynamic display closed loop constructed in Embodiment 1, introduces advanced branch limitations for the smooth transition path generation step in S400. In 3D scenes, especially in BIM+GIS-based construction reporting scenarios, the visual tearing during scene transitions is a core pain point that disrupts the immersive experience. This embodiment constructs a dual defense against visual tearing by introducing refined state parameter extraction, smooth interpolation algorithm sinking, and a strict time axis calibration mechanism.

[0105] Regarding step S400, based on the optimized resource configuration, the state parameters before scene switching are extracted. The state parameters include lighting effects and position data. A smooth transition path for scene switching is generated through smooth interpolation calculation. This embodiment elaborates on the details of state parameter extraction and interpolation calculation.

[0106] In some implementations, the state information before the scene switch is obtained from the optimized resource configuration, and the lighting effects and position data are classified and stored to obtain an initial state dataset. The design of the classified storage structure is fundamental to ensuring that data confusion does not occur in subsequent interpolation calculations.

[0107] The lighting and shadow effect data is stored in the lighting parameter cache, which includes attributes directly related to visual brightness and darkness, such as ambient light intensity, shadow map weight, and specular reflectance coefficient; the position data is stored in the spatial topology cache, which includes camera spatial coordinates, observation target point vectors, and pose matrices of key BIM components.

[0108] It should be noted that although this embodiment stores light and shadow and position in two separate caches, other embodiments may use a structured associative array or hash map for unified storage, as long as it ensures accurate indexing by data type during interpolation. This categorized storage mechanism allows for the application of differentiated interpolation strategies to different rates of change of physical properties, avoiding visual distortion caused by mixing all parameters together for interpolation at the same rate.

[0109] Subsequently, based on the position data in the initial state dataset, intermediate transition points are determined using the smooth interpolation calculation, generating a transition frame sequence. The specific algorithm chosen for smooth interpolation should match the physical motion characteristics of the scene transition.

[0110] For example, when a user slowly zooms in and out in a construction report, the system can use a linear interpolation algorithm to generate intermediate transition points at a constant speed between the starting coordinates and the target coordinates according to the time step, achieving a smooth and uniform transition. However, when the user makes a rapid camera shake or jumps from a panoramic view to a detailed view, linear interpolation will appear stiff and lack physical inertia. In this case, the system will switch to Bezier curve or B-spline curve algorithms for interpolation calculation.

[0111] Bezier curves, by introducing control points to adjust the curvature of the transition path, give the camera movement an acceleration effect at the beginning and a deceleration and buffering effect as it approaches the target, which is more in line with the natural visual habits of human observation. Through the spatial mapping of these intermediate transition points, the system can automatically generate a continuous sequence of transition frames.

[0112] After generating the transition frame sequence, based on the sequence and the requirements for the gradual change of the lighting and shadow effects, the inter-frame changes are adjusted using real-time processing technology to determine the distribution scheme of the switching timing. The logic for determining the distribution scheme of the switching timing is not a simple equal-interval time slice, but a dynamic game based on the requirements for the gradual change of lighting and shadow effects and the spatial displacement rate.

[0113] For example, in the transition from a panoramic view of outdoor construction under sunlight to detailed indoor pit work under dim lighting, the gradual change in lighting effects requires a relatively long time span to avoid the glare caused by sudden exposure changes, while the displacement of positional data may have already been completed in the earlier stages. Therefore, real-time processing technology dynamically adjusts the weighting of inter-frame changes: in the early stages of the transition, inter-frame changes focus on rapid interpolation of positional data, while lighting parameters remain relatively stable; in the later stages of the transition, as positional data tends to converge, inter-frame changes focus on the smooth decay of lighting effects. This non-uniform timing distribution ensures a natural transition of visual elements.

[0114] Finally, based on the distribution scheme of switching timing, a smooth transition path for scene switching is constructed using a dynamic adaptation method. Dynamic adaptation means that the system will fine-tune the interpolation step size and frame sequence density of the transition path in real time according to the current hardware rendering load and frame rate feedback, minimizing unnecessary redundant frame calculations while ensuring visual smoothness, thereby constructing a smooth and efficient transition path.

[0115] However, even with a precise interpolation path, a severe visual tearing effect can still occur if the changes in lighting and shadows and the movement of positions are misaligned in the temporal dimension. Therefore, this embodiment introduces a temporal axis calibration mechanism after generating the transition frame sequence.

[0116] In some implementations, the correspondence between the effect transition sequence and the transition frame sequence is obtained. The effect transition sequence is a frame sequence generated by interpolating lighting and shadow parameters, while the transition frame sequence is a frame sequence generated by interpolating position data. Ideally, each frame of these two sequences should be strictly aligned, meaning that the positional change in the Nth frame should be synchronously presented with the lighting and shadow changes in the Nth frame.

[0117] However, in actual rendering pipelines, lighting calculations typically involve complex shadow mapping and ray tracing, which often take longer than simple position matrix transformations. This means that if the timestamp discrepancy between two sequences exceeds a preset threshold, a time protocol tool is used to calibrate the transition frame sequence on the timeline to obtain synchronized transition data. The preset threshold is the critical tolerance for determining visual tearing; in this embodiment, the threshold is set to 50ms.

[0118] When the system detects that the timestamp of a frame in an effect transition sequence lags behind the timestamp of the corresponding transition frame sequence by more than 50ms, it means that the user will still see the light and shadow projection of the old position even after the viewpoint has moved to a new position. This misalignment is particularly noticeable during high-speed transitions. To eliminate this tearing, the specific application scenario for time protocol tools is clock synchronization scheduling within the rendering pipeline. A simplified version of the NTP protocol tool—an internal clock synchronization mechanism—is used to forcibly stretch the timeline of the lagging component or insert a preceding frame.

[0119] For example, when the light and shadow sequence lags, the system calculates the lag offset using the NTP calibration algorithm and inserts tiny empty frames into the transition frame sequence to wait, or compresses the interpolation step size of the light and shadow sequence to accelerate its catching-up progress, thereby forcing the two sequences to realign on the time axis and obtain synchronized transition data.

[0120] It should be noted that although this embodiment uses 50ms as an example of a preset threshold, in display terminals with different refresh rates, this threshold can also be dynamically adjusted to half of the single frame rendering time, as long as it can be ensured that the time deviation is not sensitively detected by the human eye.

[0121] Finally, based on the synchronized transition data, a final smooth path is constructed, generating a complete scene transition scheme. The synchronized transition data, after being calibrated along the time axis, completely eliminates the temporal misalignment of light and shadow with position, ensuring that every minor adjustment of visual elements during execution of the final smooth path perfectly matches the temporal dimension, thus generating a seamless and abrupt complete scene transition scheme.

[0122] The interpolation algorithm and time calibration logic described in this embodiment are for illustrative purposes only and not restrictive. Those skilled in the art can select other types of spline curves or clock synchronization protocols based on the specific characteristics of the rendering pipeline, as long as their core logic follows the principles of classified storage, differentiated interpolation and forced timestamp alignment, they should all fall within the protection scope of this invention.

[0123] Example 6 Based on the dynamic display closed loop constructed in Example 1, this embodiment further restricts the performance fallback mechanism and the final output closed loop after step S400, and substitutes all the above technical features into the actual business scenario of a construction 3D reporting PPT based on BIM+GIS, so as to demonstrate the complete dynamic display result generation process and improve the visibility and commercial value of infringement evidence collection.

[0124] After generating a smooth transition path for scene switching, the rendering of 3D scenes often faces computational bottlenecks, especially in construction reporting scenarios that include massive amounts of BIM components and GIS terrain data. Therefore, this embodiment introduces a performance fallback mechanism.

[0125] In some implementations, the rendering frame rate is monitored. The system calls the underlying performance analysis module to sample the output frequency of the rendering pipeline in real time with extremely short time steps, obtaining the fluctuation data of the current frame rate.

[0126] It should be noted that specific methods for monitoring the rendering frame rate can include operating system-level frame timestamp recording, performance query interfaces of graphics APIs (such as OpenGL or DirectX), or custom rendering loop timers, as long as the rendering time of each frame can be accurately captured.

[0127] If the rendering frame rate falls below a preset threshold, the detail level of resource blocks is downgraded to achieve stable rendering performance. The preset threshold is the critical baseline for determining the smoothness of interaction.

[0128] In this embodiment, the preset threshold is set to two levels: for VR reporting scenarios that require a very high level of immersion, the threshold is set to 60fps, and any drop below this threshold will trigger a degradation; for ordinary desktop screen reporting scenarios, the threshold is set to 30fps.

[0129] When the system detects that the frame rate has fallen below the aforementioned threshold, it means that the computational load on the graphics processor has exceeded its instantaneous processing capacity. At this point, it is necessary to reduce the load by degrading the level of detail. Specific methods for degrading the level of detail include, but are not limited to: LOD (Level of Detail) switching, which dynamically replaces the high-precision BIM component model near the center of the current viewpoint with a simplified model with a low polygon count; shadow quality degradation, which switches from high-resolution soft shadow maps to low-resolution hard shadows, or directly disables real-time shadow calculation for non-core components; and texture resolution scaling, which uses a low-level mipmap layer to sample distant GIS terrain maps.

[0130] The underlying mechanism of the aforementioned downgrade process is that the rendering computational power consumption is in a non-linear proportional relationship with the number of model faces and the complexity of light sources. By actively sacrificing some image quality in dimensions with lower visual sensitivity (such as distant details and secondary shadows), a large amount of GPU computing power can be released instantly, allowing the frame rendering time to fall back to within the safe budget. This ensures that the system can maintain stable interactive smoothness under any extremely complex scene switching and avoids reporting interruptions caused by stuttering.

[0131] After ensuring stable rendering performance, the system enters the final content output closed loop.

[0132] In some implementations, content switching signals are integrated to apply update logic to the current scenario. Content switching signals are external instructions that drive the presentation process. These signals can take the form of page-turning signals generated by users clicking PPT page-turning buttons, specific instruction words parsed by the speech recognition module, or preset timeline automatic playback trigger signals.

[0133] After capturing the signal, it will be deeply integrated with the currently executing smooth transition path. Based on the signal's intent, the corresponding update logic will be applied to the current scene's geometric model, material maps, and lighting parameters, such as switching from the logical state of "earthwork excavation stage" to the logical state of "main structure construction stage".

[0134] Subsequently, a consistency analysis is performed on the updated scene content. If the state continuity determination result does not meet the preset standard, the scene update part is transitioned to obtain the final dynamic display result. The consistency analysis aims to ensure that the transition between the old and new scene states does not produce logical gaps.

[0135] The specific dimensions of the state continuity judgment criteria include: spatial topological continuity, i.e., whether the coordinate difference between the old and new viewpoints is within the allowable range of physical motion laws; light and shadow logic continuity, i.e. whether the ambient light intensity of the new scene and the attenuation transition of the old scene are smoothly connected without exposure jumps; and business attribute continuity, i.e. whether the timestamp of the construction progress model increases sequentially without logical fallacies caused by component regression.

[0136] If the judgment result of any of the above dimensions does not meet the preset standard, for example, if the light intensity of the new scene is detected to increase by 300% instantaneously, causing visual exposure tearing, the system will determine that the continuity of the state is disrupted and immediately start transition processing for the scene update part.

[0137] The specific methods of transition processing include: forcibly inserting linear decay buffer frames into the abnormally changing light and shadow parameters, or applying transparency gradient hiding and revealing effects to components with spatial topological misalignment, thereby repairing visual discontinuities at the logical level and ultimately outputting dynamic display results that conform to the business evolution rules and have high visual fidelity.

[0138] To illustrate the invention more clearly, all the above technical features will be described in detail below by substituting them into the actual business scenario of "BIM+GIS-based 3D construction reporting PPT".

[0139] In this scenario, the user is giving an important construction progress report and needs to switch from the "Overall Planning View" to the "Specific Area Detail View". Initially, the user is in the Overall Planning View, which provides an overview of the GIS terrain and BIM tower crane layout of the entire construction site. At this time, the system memory is loaded with low-precision panoramic outline resource blocks.

[0140] When a user clicks on an interactive component in the PPT to switch to the "details of the excavation area" content, the user's gaze quickly shifts and zooms in on the excavation area. At this time, step S100 collects information on the change in the user's perspective, captures the linear and angular velocities of the line of sight displacement, extracts the current position coordinates and direction vector pointing towards the excavation area, and obtains the visible area focused on the excavation area.

[0141] Next, step S200 performs resource scheduling on the 3D scene based on the sharp shrinkage and deflection of the visible area. The system uses spatial perception technology to predict that the viewpoint is about to fall into the pit, dynamically reconstructs the mesh distribution structure, increases the mesh density in the pit area, and removes distant tower crane outline resource blocks that do not match the area changes. Simultaneously, for unloaded high-precision BIM components within the pit, priority scoring is performed based on distance and volume, placing core components such as pit support piles at the front of the loading sequence for batch processing and timeout splitting, generating a sorted loading sequence.

[0142] Subsequently, step S300 obtains the loading sequence and performs memory allocation operations on the high-precision components. Due to the large amount of data of the high-precision components, the memory usage quickly exceeds the system memory limit (e.g., the upper limit of 80%). The system locks the pit detail resource blocks that caused the overflow and selects low-priority resource blocks for unloading based on a multi-dimensional priority strategy. At this time, the GIS terrain map and temporary road model of the distant construction unit are accurately unloaded because they are far away and do not contribute visually to the current pit view, freeing up a large amount of memory space. If the restriction conditions are still not met after unloading, a second unloading is triggered to strip the non-core annotation lines in the medium-priority area, and finally obtain the optimized resource configuration to ensure that the pit detail resources are successfully loaded into memory.

[0143] After resources are secured, step S400 extracts the state parameters before scene switching, including the uniform sunlight and shadow effects from the overhead view and the long-distance position data. Through Bezier curve smooth interpolation calculation, a smooth transition path is generated for the camera to dive into the pit and gradually switch to local point light source illumination.

[0144] During path execution, the rendering frame rate is monitored. If the frame rate drops below the preset threshold of 30fps due to excessive number of facets in the internal components of the foundation pit, the system immediately degrades the level of detail of the resource blocks, switches the non-core scaffolding model around the foundation pit to a low-poly LOD, and reduces its shadow quality, thereby quickly recovering rendering performance and achieving stable rendering performance.

[0145] Ultimately, the system integrates content switching signals, applies the "excavation phase" update logic to the foundation pit scene, and performs consistency analysis on the updated scene. If a slight jump in the light and shadow transition is detected, the scene update portion is processed to eliminate visual tearing, resulting in the final dynamic display result—users experience a seamless, smooth, and coherent 3D immersive transition from macro planning to micro details in their presentation slides, completely avoiding the stuttering, jumps, and loading waits common in traditional presentations.

[0146] The performance degradation threshold, consistency judgment dimensions, and business scenarios described in this embodiment are for illustrative purposes only. Those skilled in the art, guided by this embodiment, can select other types of frame rate baselines or continuity judgment rules based on specific hardware configurations and reporting business characteristics. As long as their core logic adheres to the principles of performance safety net and state continuity closed loop, they should fall within the protection scope of this invention.

[0147] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0148] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and modifications.

Claims

1. A method for editing and customizing 3D construction reports based on BIM+GIS, characterized in that, Collect information on changes in the user's field of view, extract the current position coordinates and direction vector from the information, and obtain the current visible area range; Based on the current visible area and its changes, resource scheduling and priority sorting are performed on the 3D scene to determine the set of resource blocks to be loaded and the sorted loading sequence. Obtain the sorted loading sequence, perform memory allocation operations on the resource blocks, and if the system memory limit is exceeded, unload the resource blocks in the low-priority area to obtain the optimized resource configuration; Based on the optimized resource configuration, the state parameters before scene switching are extracted, and a smooth transition path for scene switching is generated by smooth interpolation calculation; the state parameters include lighting effects and position data.

2. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 1, characterized in that, The collection of user perspective change information includes: Collect user head rotation or gaze displacement data to obtain a dataset of viewpoint changes; Extract the current position coordinates and the direction vector from the viewpoint change dataset to determine the user's current position point and direction reference in space; If the current position coordinates deviate from the preset coordinate system, the current position coordinates are standardized through calibration to obtain calibrated position data.

3. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 1, characterized in that, The resource scheduling for the 3D scene includes: The three-dimensional scene is divided into blocks using a spatial region partitioning strategy to obtain multiple scene content units; Analyze the correspondence between each scene content unit and the current visible area range to determine the set of target units located within the current visible area range; The set of resource blocks to be loaded is determined from the set of target units.

4. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 3, characterized in that, The step of dividing the 3D scene into blocks using a spatial region partitioning strategy includes: Based on the changes in the visible area, spatial perception technology is used to collect environmental data in the area in real time, resulting in a dynamic information set of regional changes. The grid distribution structure is dynamically reconstructed based on environmental data, and resource blocks that do not match regional changes are removed. The set of resource blocks to be loaded is updated through multi-dimensional data synchronization and priority sorting.

5. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 1, characterized in that, The priority sorting includes: For the unloaded blocks in the set of resource blocks that need to be loaded, a preset priority mechanism is used for preliminary classification to obtain the classified block partitioning results; Priority scoring is performed based on the distance of resource blocks from the viewpoint center and the resource volume, and unloaded blocks are processed in batches according to the scoring results; Batches exceeding the preset time limit are split to obtain the sorted loading sequence.

6. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 1, characterized in that, The process of performing memory allocation operations on resource blocks, and unloading resource blocks in low-priority regions if the system memory limit is exceeded, includes: Obtain the sorted loading sequence and perform memory allocation operations on the resource blocks; The system monitors the memory level in real time. If the memory usage exceeds the system memory limit, the specific resource block that caused the memory overflow is locked. Based on a multi-dimensional priority strategy, resource blocks in the low-priority regions are selected for unloading, thereby releasing the corresponding memory space. If the restrictions are still not met after uninstallation, a secondary uninstallation mechanism is triggered to obtain the optimized resource configuration.

7. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 1, characterized in that, The step of extracting the state parameters before scene switching and generating a smooth transition path for scene switching through smooth interpolation includes: Obtain the state information before scene switching from the optimized resource configuration, classify and store the lighting effects and the position data to obtain the initial state dataset; Based on the position data in the initial state dataset, intermediate transition points are determined using the smooth interpolation calculation, and a transition frame sequence is generated. Based on the transition frame sequence and the gradual change requirements of the lighting and shadow effects, the inter-frame changes are adjusted through real-time processing technology to determine the distribution scheme of the switching timing. Based on the distribution scheme of switching timing, a smooth transition path for scene switching is constructed using a dynamic adaptation method.

8. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 7, characterized in that, After generating the transition frame sequence, the method further includes: Obtain the correspondence between the effect transition sequence and the transition frame sequence; If the timestamp deviation between the two sequences exceeds a preset threshold, a time protocol tool is used to perform time axis calibration on the transition frame sequence to obtain synchronized transition data. Based on the synchronous transition data, the final smooth path is constructed, generating a complete scene switching transition scheme.

9. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 7, characterized in that, Following the smooth transition path for generating scene changes, the following is also included: Monitor rendering frame rate; If the rendering frame rate is lower than the preset threshold, the detail level of the resource blocks will be downgraded to obtain stable rendering performance.

10. The method for editing and customizing construction 3D reports based on BIM+GIS according to claim 7, characterized in that, Following the smooth transition path for generating scene changes, the following is also included: Integrate content switching signals and apply update logic to the current scenario; A consistency analysis is performed on the updated scene content. If the state continuity judgment result does not meet the preset standard, the scene update part is processed to obtain the final dynamic display result.

Images