Realistic three-dimensional geographical entity and spatial planning display system

Abstract

The present application belongs to the technical field of image data processing, and discloses a real scene three-dimensional geographic entity and space planning display system; the system comprises: a first processing module, which constructs a three-dimensional space tile data set corresponding to N precision levels based on multi-density sampling of real scene point cloud data; a second processing module, which extracts operation behavior characteristics based on zooming operation behavior in a three-dimensional display system to obtain operation behavior characteristic data; a behavior diagnosis module, which inputs the operation behavior characteristic data into an operation behavior diagnosis model to obtain a zooming behavior mode; and an intelligent display module, which intelligently displays the three-dimensional display system based on the operation behavior characteristic data, the zooming behavior mode and the three-dimensional space tile data set corresponding to the N precision levels; the application introduces an operation behavior characteristic extraction and zooming behavior mode recognition mechanism, and intelligently and adaptively schedules and displays, thereby effectively solving the problems of zooming misplacement and visual jump.

Description

Technical Field

[0001] This invention relates to the field of image data processing technology, and more specifically, to a real-scene 3D geographic entity and spatial planning display system. Background Technology

[0002] In existing technologies, when displaying and analyzing real-world 3D geographic entity data and planning model data at multiple scales, the real-world point-to-cloud data is typically downsampled at a uniform granularity to generate a single-level set of 3D spatial tiles. Different resolution data sets are then switched according to a fixed scaling ratio to adapt to changes in the view. Simultaneously, to improve the zooming experience, some systems introduce zoom rate detection, adjusting tile loading priority based on the zoom speed to alleviate rendering latency issues caused by continuous zooming.

[0003] However, existing technologies still have significant shortcomings in handling scenarios with large-scale scaling changes, rapid scaling, or drastic changes in spatial scope. Due to the lack of unified standards for hierarchical division and error control between real-world point-of-view and spatial planning models, issues such as feature misalignment, model overlap, or visual jumps easily occur during scaling, affecting the continuity of the display and the consistency of spatial logic. Furthermore, existing technologies cannot fully extract scaling behavior characteristics and dynamically adapt tile scheduling strategies, resulting in rigid resource scheduling and delayed loading responses during rapid zooming in or out. This can easily lead to bottlenecks in the instantaneous loading of large-scale tiles or delays in detailed loading, making it difficult to balance smooth scaling with the final presentation of visual details.

[0004] Therefore, there is an urgent need to propose a display system that can dynamically identify scaling modes and intelligently optimize loading scheduling based on operational behavior characteristics, so as to improve the overall display continuity, system response performance and spatial expression accuracy of 3D scenes under multi-scale scaling. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and achieve the above objectives, the present invention provides the following technical solution: a real-scene 3D geographic entity and spatial planning display system, comprising:

[0006] The first processing module constructs a set of three-dimensional spatial tile data corresponding to N precision levels based on real-world point cloud data sampled at multiple densities.

[0007] The second processing module extracts operation behavior features based on the zoom operation behavior in the 3D display system to obtain operation behavior feature data.

[0008] The behavior diagnosis module inputs operational behavior feature data into the operational behavior diagnosis model to obtain scaling behavior patterns.

[0009] The intelligent display module, a 3D display system, performs intelligent display based on operation behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N precision levels.

[0010] Furthermore, the scaling behavior modes include slow scaling and fast scaling; the slow scaling includes slow zooming in and slow zooming out; the fast scaling includes fast zooming in and fast zooming out; the method for intelligent display based on operation behavior feature data, scaling behavior modes, and a set of three-dimensional spatial tile data corresponding to N precision levels includes:

[0011] If the zoom behavior mode is slow zoom, the 3D display system will intelligently display the data for slow zoom.

[0012] If the scaling behavior mode is fast scaling, the operation behavior feature data and scaling behavior mode are input into the accuracy level prediction model to obtain the prediction accuracy level; the 3D display system performs intelligent display for fast scaling.

[0013] Furthermore, methods for intelligent display in 3D display systems to handle slow zooming include:

[0014] The second scaling ratio value corresponding to the completion of the scaling operation is obtained from the operation behavior feature data. The second scaling ratio value is matched with a pre-constructed scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the slow scaling target precision level. The scaling ratio-precision level matching table includes the scaling ratio value and the corresponding precision level.

[0015] From the three-dimensional spatial tile data sets corresponding to N precision levels, filter and slowly scale the three-dimensional spatial tile data set corresponding to the target precision level, denoted as the target three-dimensional spatial tile data set;

[0016] The 3D display system performs rendering and display operations based on the target 3D spatial tile data set.

[0017] Furthermore, the methods for intelligent display in 3D display systems for rapid zooming include:

[0018] The first scaling ratio value corresponding to the start of the scaling operation is obtained from the operation behavior feature data. The first scaling ratio value is matched with the pre-built scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the fast scaling target precision level.

[0019] Each 3D spatial tile in the fast-scaling target accuracy level 3D spatial tile dataset is clustered with all 3D spatial tiles in the prediction accuracy level 3D spatial tile dataset using a clustering algorithm to obtain the corresponding tile similarity. For each 3D spatial tile in the fast-scaling target accuracy level 3D spatial tile dataset, the 3D spatial tile with the highest similarity to it is selected from the prediction accuracy level 3D spatial tile dataset to form a one-to-one corresponding 3D spatial tile data pair.

[0020] Preset tile similarity threshold 1 and tile similarity threshold 2; tile similarity threshold 1 is greater than tile similarity threshold 2;

[0021] 3D spatial tile data pairs with tile similarity greater than or equal to tile similarity threshold one are assigned to the preloaded set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold one but greater than or equal to tile similarity threshold two are assigned to the progressive supplement set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold two are assigned to the unloaded set.

[0022] The 3D display system performs loading, delayed supplementation, and unloading rendering operations based on the preloaded set, the progressively supplemented set, and the unloaded set, respectively.

[0023] Furthermore, the method for obtaining the operational behavior feature data includes:

[0024] When a scaling operation is detected, the system timestamp at the start of the scaling operation and the scaling ratio value are obtained and recorded as the first system timestamp and the first scaling ratio value, respectively.

[0025] When the scaling operation is detected to be complete, the system timestamp and scaling ratio value at the time of completion of the scaling operation are obtained and recorded as the second system timestamp and the second scaling ratio value, respectively.

[0026] The scaling change is obtained by subtracting the second scaling value from the first scaling value; if the scaling change is greater than zero, the scaling direction indicator is set to the zoom-in indicator; if the scaling change is less than zero, the scaling direction indicator is set to the zoom-out indicator.

[0027] The scaling time interval is obtained by subtracting the second system timestamp from the first system timestamp.

[0028] The scaling rate is obtained by taking the absolute value of the scaling change magnitude and dividing it by the scaling change time interval.

[0029] The second scaling factor, scaling change magnitude, scaling direction identifier, scaling change time interval, and scaling change rate are used to construct the operation behavior feature data.

[0030] Furthermore, methods for constructing three-dimensional spatial tile datasets corresponding to N precision levels based on real-world point cloud data with multi-density sampling include:

[0031] Acquire real-world location cloud data, which includes spatial three-dimensional coordinates;

[0032] Construct a bounding box of real-site cloud data;

[0033] The multi-scale precision control requirements of the pre-collected 3D display system are input into the precision parameter setting model to obtain the precision setting parameters. The precision setting parameters include N precision levels and corresponding layered sampling granularity. The multi-scale precision control requirements include minimum magnification ratio, maximum magnification ratio, minimum reduction ratio, and maximum reduction ratio.

[0034] Based on N precision levels and corresponding hierarchical sampling granularity, the bounding boxes of real point cloud data are downsampled to obtain the downsampled point cloud data sets corresponding to N precision levels.

[0035] Spatial slicing is performed on hierarchical sampling granularity and downsampled point cloud data sets with N precision levels to obtain three-dimensional spatial tile data sets corresponding to N precision levels.

[0036] Furthermore, methods for constructing a bounding box for real-world location cloud data include:

[0037] Obtain all three-dimensional coordinate points in the real-world point cloud data. Based on all three-dimensional coordinate points, determine the boundary values ​​of the real-world point cloud data in three dimensions in sequence, including the maximum and minimum coordinate values ​​in the X-axis direction, the maximum and minimum coordinate values ​​in the Y-axis direction, and the maximum and minimum coordinate values ​​in the Z-axis direction.

[0038] The length of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the X-axis direction; the width of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Y-axis direction; and the height of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Z-axis direction. The center coordinates of the point cloud data bounding box are then obtained based on its length, width, and height.

[0039] Using the length, width, height, and center coordinates of the real-world point cloud data bounding box as parameters, a geometric boundary volume is constructed that encloses the entire real-world point cloud data space, thus obtaining the real-world point cloud data bounding box.

[0040] Furthermore, the method for downsampling the bounding boxes of real-world point cloud data based on N precision levels and their corresponding hierarchical sampling granularities to obtain downsampled point cloud data sets corresponding to N precision levels includes:

[0041] S100: Let the initial value of n be 1, and the range of n is from 1 to N;

[0042] S101: Using the granularity of the nth precision level as the precision, divide along the X-axis direction. Each point cloud data grid cell is divided along the Y-axis. Each point cloud data grid cell is divided along the Z-axis. Each point cloud data grid cell yields a total of Total. n Each point cloud data grid unit; Indicates rounding up; CD bwh KD is the length of the real-world cloud data bounding box. bwh For the width of the real-world cloud data bounding box, GD bwh For the height of the real-world cloud data bounding box, LD n This represents the stratified sampling granularity for the nth precision level.

[0043] S102: Total n The point cloud data grid cells are used to construct the downsampled point cloud data set of the nth precision level; the downsampled point cloud data set maintains the same spatial boundary range as the real point cloud data bounding box;

[0044] S103: Let n = n + 1. If n is less than or equal to N, continue executing S101 to S102; if n is greater than N, end the current process.

[0045] Furthermore, the method for obtaining three-dimensional spatial tile data sets corresponding to N precision levels by performing spatial block processing based on hierarchical sampling granularity and downsampled point cloud data sets with N precision levels includes:

[0046] S200: Let the initial value of n be 1, and the range of n is from 1 to N;

[0047] S201: Obtain the total number of point cloud data grid cells in the nth downsampled point cloud data set. n ; for Total n Each point cloud data grid cell constructs a corresponding downsampled point cloud data sub-bounding box; the downsampled point cloud data sub-bounding box is a 3D spatial tile, resulting in Total. n A three-dimensional spatial tile;

[0048] S202: For Total respectively n The length, width, height, and center coordinates of each 3D spatial tile are extracted to obtain the Total value. nThe three-dimensional spatial tile corresponds to the three-dimensional spatial tile length, three-dimensional spatial tile width, three-dimensional spatial tile height, and three-dimensional spatial tile center coordinates; the error parameter of the three-dimensional spatial tile is β times the layer sampling granularity of the nth precision level, where β is a constant greater than 0;

[0049] S203: Total n The three-dimensional spatial tile error parameters, three-dimensional spatial tile length, three-dimensional spatial tile width, three-dimensional spatial tile height, and three-dimensional spatial tile center coordinates corresponding to each three-dimensional spatial tile are used to construct the Total. n The three-dimensional spatial tile index information corresponding to each three-dimensional spatial tile; Total n Each 3D spatial tile and its corresponding 3D spatial tile index information constitutes a Total. n Three-dimensional spatial tile data, Total n The data of individual three-dimensional spatial tiles are used to construct a three-dimensional spatial tile data set;

[0050] S204: Let n = n + 1. If n is less than or equal to N, continue executing S201 to S103; if n is greater than N, end the current process.

[0051] Furthermore, the training method for the operational behavior diagnostic model includes:

[0052] A pre-constructed operational behavior diagnostic dataset is provided, which includes XW sets of operational behavior diagnostic data and the scaling behavior patterns corresponding to the XW sets of operational behavior diagnostic data, where XW is a positive integer greater than 0, and the operational behavior diagnostic data includes operational behavior feature data.

[0053] The operational behavior diagnosis dataset is divided into a training set and a validation set. The training set is used for parameter learning, and the validation set is used for evaluating the model's generalization ability.

[0054] Support vector machine (SVM) model is used as the operational behavior diagnosis model. Operational behavior feature data is converted into feature vector input, nonlinear features are extracted through hidden layers, and the scaling behavior pattern with the highest probability is output as the final prediction result in the output layer. The training objective is to minimize the cross-entropy loss function. Training stops when the accuracy of the validation set reaches a preset threshold.

[0055] Compared with existing technologies, the technical effects and advantages of the real-scene 3D geographic entity and spatial planning display system of this invention are as follows:

[0056] The real-scene 3D geographic entity and spatial planning display system provided by this invention is based on a multi-density sampling mechanism. It constructs a 3D spatial tile data set covering different precision levels for real-scene point cloud data. Combining the matching relationship between scaling ratio and precision level, it effectively realizes the structured organization and index optimization of multi-scale 3D data through downsampling and spatial block processing.

[0057] Building upon this foundation, this application introduces an operation behavior feature extraction and scaling behavior pattern recognition mechanism. Based on feature data such as scaling change magnitude, scaling direction, and scaling rate, it can determine in real time whether the user's scaling behavior is in a rapid zoom-in, slow zoom-in, rapid zoom-out, or slow zoom-out mode, and then dynamically adapt the corresponding 3D spatial tile scheduling strategy. Specifically, for slow scaling, the system accurately switches to the 3D spatial tile set at the target accuracy level based on the scaling ratio at the end of the operation, ensuring display continuity and system stability. For rapid scaling, based on the initial scaling ratio and the predicted accuracy level, a clustering algorithm is used for tile similarity matching, dividing the system into a pre-loading set, a progressively supplemented set, and an unloading set. Loading, delayed supplementation, and resource reclamation are then executed in an orderly manner according to priority, effectively mitigating the loading abrupt changes and performance bottlenecks caused by rapid scaling.

[0058] Through the above mechanism, this application not only solves the problems of scaling misalignment, visual jumps and analysis distortion caused by the inconsistency in the hierarchical definition of real-world point cloud and planning model and the lack of unified error control in the prior art, but also significantly improves the rendering continuity, spatial logic consistency and resource scheduling efficiency of the 3D display system under multi-scale scaling. At the same time, it takes into account the comprehensive needs of refined, real-time and large-scale data processing in urban 3D reconstruction, spatial planning display and digital twin applications. It has significant technical effects of clear system structure, superior response performance and strong scalability. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the real-scene three-dimensional geographic entity and spatial planning display system of Embodiment 1 of the present invention;

[0060] Figure 2 This is a flowchart of the real-scene three-dimensional geographic entity and spatial planning display method of Embodiment 2 of the present invention;

[0061] Figure 3 A flowchart illustrating a method for intelligent display in a 3D display system to handle slow zooming.

[0062] Figure 4 A flowchart illustrating a method for intelligent display in a 3D display system for rapid zooming.

[0063] Figure 5 A schematic diagram of the architecture of a real-world 3D geographic entity and spatial planning display system;

[0064] Figure 6 A mind map for displaying real-world 3D geographic entities and spatial planning methods. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be described in detail, clearly, and completely below with reference to the accompanying drawings. It should be particularly noted that the specific embodiments described below are only for better illustrating and explaining the technical solutions of the present invention, and are intended to enable those skilled in the art to better understand and implement the present invention, and should not be construed as limiting the scope of protection of the present invention. Without departing from the spirit and substance of the present invention, those skilled in the art can modify, adjust, or make equivalent substitutions based on the content disclosed in the present invention, and these should all be considered within the scope of protection of the present invention.

[0066] Example 1

[0067] Please see Figure 1 As shown in the figure, this embodiment discloses a real-scene 3D geographic entity and spatial planning display system, including a first processing module, a second processing module, a behavior diagnosis module and an intelligent display module. Each module is connected by wires and / or wirelessly to realize data transmission.

[0068] The first processing module constructs a set of three-dimensional spatial tile data corresponding to N precision levels based on real-world point cloud data sampled at multiple densities.

[0069] Methods for constructing three-dimensional spatial tile datasets corresponding to N precision levels based on real-world point cloud data using multi-density sampling include:

[0070] Obtain real-world point cloud data, which includes spatial three-dimensional coordinates (X,Y,Z); the real-world point cloud data format can be LAS, LAZ, PCD, or BIN, etc.

[0071] Construct a bounding box of real-site cloud data;

[0072] The multi-scale precision control requirements of the pre-collected 3D display system are input into the precision parameter setting model to obtain the precision setting parameters. The precision setting parameters include N precision levels and corresponding layered sampling granularity. The multi-scale precision control requirements include minimum magnification ratio, maximum magnification ratio, minimum reduction ratio, and maximum reduction ratio.

[0073] Based on N precision levels and corresponding hierarchical sampling granularity, the bounding boxes of real point cloud data are downsampled to obtain the downsampled point cloud data sets corresponding to N precision levels.

[0074] Spatial slicing is performed on hierarchical sampling granularity and downsampled point cloud data sets with N precision levels to obtain three-dimensional spatial tile data sets corresponding to N precision levels.

[0075] Methods for constructing a bounding box of real-location cloud data include:

[0076] Obtain all three-dimensional coordinate points in the real-world point cloud data. Based on all three-dimensional coordinate points, determine the boundary values ​​of the real-world point cloud data in three dimensions in sequence, including the maximum and minimum coordinate values ​​in the X-axis direction, the maximum and minimum coordinate values ​​in the Y-axis direction, and the maximum and minimum coordinate values ​​in the Z-axis direction.

[0077] The length of the real-world point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the X-axis direction, the width of the real-world point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Y-axis direction, and the height of the real-world point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Z-axis direction.

[0078] The average of the maximum and minimum coordinate values ​​in the X-axis direction, the average of the maximum and minimum coordinate values ​​in the Y-axis direction, and the average of the maximum and minimum coordinate values ​​in the Z-axis direction are used to construct the center coordinates of the real-world point cloud data bounding box.

[0079] Using the length, width, height, and center coordinates of the aforementioned real-world cloud data bounding box as parameters, a geometric boundary volume enclosing the entire real-world cloud data space is constructed, thus obtaining the real-world cloud data bounding box.

[0080] Methods for obtaining the center coordinates of the real-world point-of-view cloud data bounding box include:

[0081]

[0082] Among them, ZB center X is the center coordinate of the real-world point cloud data bounding box. max The maximum coordinate value in the X-axis direction, X min The minimum coordinate value in the X-axis direction, Y max The maximum coordinate value in the Y-axis direction, Y min The minimum coordinate value in the Y-axis direction, Z max The maximum coordinate value in the Z-axis direction, Z min This is the minimum coordinate value in the Z-axis direction.

[0083] The training method for the accuracy parameter setting model includes:

[0084] A pre-constructed precision parameter setting dataset is provided, comprising precision parameter setting data for the JD group and corresponding precision setting parameters, where JD is a positive integer greater than 0. The precision parameter setting data includes multi-scale precision control requirements, including minimum magnification ratio, maximum magnification ratio, minimum reduction ratio, and maximum reduction ratio. The precision parameter setting dataset is divided into a precision parameter setting data training set and a precision parameter setting data validation set. The precision parameter setting data training set is used for parameter learning of the precision parameter setting model, and the precision parameter setting data validation set is used for real-time evaluation of the generalization ability of the precision parameter setting model.

[0085] During the training process of the precision parameter setting model, the precision parameter setting model is trained using a Bayesian network model. The precision parameter setting data is converted into feature vectors as input, and nonlinear features in the data are extracted through hidden layers. Finally, the probability distribution of the precision setting parameters is generated in the output layer using the softmax activation function, and the precision setting parameter corresponding to the highest probability is output as the final prediction result. The training process aims to minimize the cross-entropy loss function, and an early stopping strategy is introduced to monitor the performance of the precision parameter setting data validation set. When the prediction accuracy on the precision parameter setting data validation set reaches the preset accuracy, the precision parameter setting model is determined to have converged, and the training stops immediately.

[0086] Examples of precision setting parameters are shown in Table 1:

[0087] Table 1 Accuracy Setting Parameters

[0088]

[0089]

[0090] The method for downsampling the bounding boxes of real-world point cloud data based on N precision levels and corresponding hierarchical sampling granularities to obtain downsampled point cloud data sets corresponding to N precision levels includes:

[0091] S100: Let the initial value of n be 1, and the range of n is from 1 to N;

[0092] S101: Using the granularity of the nth precision level as the precision, divide along the X-axis direction. Each point cloud data grid cell is divided along the Y-axis. Each point cloud data grid cell is divided along the Z-axis. Each point cloud data grid unit; Indicates rounding up; CD bwh KD is the length of the real-world cloud data bounding box. bwh For the width of the real-world cloud data bounding box, GDbwh For the height of the real-world cloud data bounding box, LD n This refers to the hierarchical sampling granularity at the nth precision level; that is, dividing the bounding box of the real-world point-cloud data into a voxel mesh to obtain Total. n Each point cloud data grid cell, Total n for and Multiplying them together yields the result.

[0093] It should be noted that,

[0094] S102: Total n The point cloud data grid cells are used to construct the downsampled point cloud data set of the nth precision level; the downsampled point cloud data set maintains the same spatial boundary range as the real point cloud data bounding box;

[0095] S103: Let n = n + 1. If n is less than or equal to N, continue executing S101 to S102; if n is greater than N, end the current process.

[0096] The method for obtaining three-dimensional spatial tile data sets corresponding to N precision levels by spatially segmenting a hierarchical sampling granularity and downsampled point cloud data sets based on N precision levels includes:

[0097] S200: Let the initial value of n be 1, and the range of n is from 1 to N;

[0098] S201: Obtain the total number of point cloud data grid cells in the nth downsampled point cloud data set. n ; for Total n Each point cloud data grid cell constructs a corresponding downsampled point cloud data sub-bounding box; the downsampled point cloud data sub-bounding box is a 3D spatial tile, resulting in Total. n A three-dimensional spatial tile;

[0099] S202: For Total respectively n The length, width, height, and center coordinates of each 3D spatial tile are extracted to obtain the Total value. n The three-dimensional spatial tile corresponds to the three-dimensional spatial tile length, three-dimensional spatial tile width, three-dimensional spatial tile height, and three-dimensional spatial tile center coordinates; the error parameter of the three-dimensional spatial tile is β times the layer sampling granularity of the nth precision level, where β is a constant greater than 0; for example, in this embodiment, the value of β can be set to 2;

[0100] S203: Total nThe three-dimensional spatial tile error parameters, three-dimensional spatial tile length, three-dimensional spatial tile width, three-dimensional spatial tile height, and three-dimensional spatial tile center coordinates corresponding to each three-dimensional spatial tile are used to construct the Total. n The three-dimensional spatial tile index information corresponding to each three-dimensional spatial tile; Total n Each 3D spatial tile and its corresponding 3D spatial tile index information constitutes a Total. n Three-dimensional spatial tile data, Total n The data of individual three-dimensional spatial tiles are used to construct a three-dimensional spatial tile data set;

[0101] S204: Let n = n + 1. If n is less than or equal to N, continue executing S201 to S103; if n is greater than N, end the current process.

[0102] It should be noted that this application generates N-level three-dimensional spatial tiles and their index information based on multi-density sampled real-world point cloud data. This method can structure the original unstructured high-density real-world point cloud data according to the multi-scale precision control requirements of the 3D display system. By setting precision parameters, the model automatically generates precision levels and layered sampling granularity, realizing multi-level downsampling of point cloud data within the spatial range. Based on the downsampling results, it constructs three-dimensional spatial tiles with spatial boundary descriptions and error control parameters, along with corresponding index information, thereby forming a precision-layered point cloud data structure that can be called on demand by the 3D visualization system. This method can not only effectively improve the organization efficiency and spatial schedulability of point cloud data and support multi-level continuous display from a global perspective to local details, but also reduce system resource consumption and improve rendering performance and interactive response speed. It is particularly suitable for application scenarios with high requirements for point cloud precision control and visualization performance, such as urban 3D reconstruction, spatial planning display, and digital twin platforms. It has significant advantages such as clear structure, high processing efficiency, strong system adaptability, and good scalability.

[0103] It should be noted that although existing technologies can perform multi-level simplification and error unification processing on real-world point cloud data and spatial planning models, and generate 3D spatial tiles that meet spatial alignment requirements, the lack of a coordination mechanism in the front-end 3D engine regarding the loading scheduling logic of different data sources and error interpretation standards during actual scaling operations still leads to problems such as geometric jumps, loading lags, display misalignments, and scaling level imbalances between real-world point cloud data and spatial planning models. This affects the continuity and accuracy of the final 3D scene display, resulting in inconsistent spatial analysis results and making it difficult to meet practical engineering needs.

[0104] The second processing module extracts operation behavior features based on the scaling operation behavior in the 3D display system to obtain operation behavior feature data.

[0105] The method for obtaining the operational behavior feature data includes:

[0106] When a scaling operation is detected, the system timestamp at the start of the scaling operation and the scaling ratio value are obtained and recorded as the first system timestamp and the first scaling ratio value, respectively.

[0107] When the scaling operation is detected to be complete, the system timestamp and scaling ratio value at the time of completion of the scaling operation are obtained and recorded as the second system timestamp and the second scaling ratio value, respectively.

[0108] The scaling change is obtained by subtracting the second scaling value from the first scaling value; if the scaling change is greater than zero, the scaling direction indicator is set to the zoom-in indicator; if the scaling change is less than zero, the scaling direction indicator is set to the zoom-out indicator.

[0109] It should be noted that the zoom direction indicator can also be represented by a numerical value. For example, the zoom direction indicator can be set to -1 and 1, where -1 means zoom out and 1 means zoom in.

[0110] The scaling time interval is obtained by subtracting the second system timestamp from the first system timestamp.

[0111] The scaling rate is obtained by taking the absolute value of the scaling change magnitude and dividing it by the scaling change time interval.

[0112] The second scaling factor, scaling change magnitude, scaling direction identifier, scaling change time interval, and scaling change rate are used to construct the operation behavior feature data.

[0113] The method for calculating the scaling change range includes:

[0114] BHFD = DEBL - DYBL;

[0115] Where BHFD is the scaling change range, DEBL is the second scaling ratio value, and DYBL is the first scaling ratio value.

[0116] The method for calculating the scaling time interval includes:

[0117] SJJG = DESJ-DYSJ;

[0118] Where SJJG is the scaling time interval, DESJ is the second system timestamp, and DYSJ is the first system timestamp.

[0119] The method for calculating the scaling rate includes:

[0120]

[0121] Where SFSL is the scaling rate.

[0122] The behavior diagnosis module inputs the operation behavior feature data into the operation behavior diagnosis model to obtain the scaling behavior pattern; the scaling behavior pattern includes slow scaling and fast scaling; the slow scaling includes slow zooming in and slow zooming out; the fast scaling includes fast zooming in and fast zooming out.

[0123] The training method for the operational behavior diagnostic model includes:

[0124] A pre-constructed operational behavior diagnostic dataset is provided, comprising XW groups of operational behavior diagnostic data and corresponding scaling behavior patterns, where XW is a positive integer greater than 0. The operational behavior diagnostic data includes operational behavior feature data. The operational behavior diagnostic dataset is divided into an operational behavior diagnostic data training set and an operational behavior diagnostic data validation set. The operational behavior diagnostic data training set is used for parameter learning of the operational behavior diagnostic model, and the operational behavior diagnostic data validation set is used for real-time evaluation of the generalization ability of the operational behavior diagnostic model.

[0125] During the training of the operational behavior diagnostic model, the model employs a support vector machine (SVM) model. Operational behavior diagnostic data is converted into feature vectors as input. Nonlinear features are extracted from the data through hidden layers, and finally, a softmax activation function is used in the output layer to generate a probability distribution of scaling behavior patterns. The scaling behavior pattern with the highest probability is output as the final prediction result. The training process aims to minimize the cross-entropy loss function and incorporates an early stopping strategy to monitor the performance of the operational behavior diagnostic data validation set. When the prediction accuracy on the validation set reaches a preset accuracy, the operational behavior diagnostic model is considered to have converged, and training is immediately stopped.

[0126] It should be noted that by collecting operational behavior feature data such as scaling amplitude, scaling direction indication, scaling time interval, and scaling rate, the action characteristics and trends of the 3D display system during scaling operations can be effectively characterized. By training a machine learning model based on this operational behavior feature data, the model can learn the feature distribution characteristics corresponding to different scaling behavior patterns. In practical applications, this allows for accurate identification of whether the current scaling behavior is a rapid zoom-in, slow zoom-out, rapid zoom-out, or slow zoom-out pattern based on real-time extracted operational behavior feature data. Introducing a machine learning model for scaling behavior pattern recognition effectively avoids the problems of low accuracy and poor adaptability in complex interactive environments caused by traditional single-threshold rule-based judgments, further improving the accuracy and robustness of scaling behavior recognition. Furthermore, the identified scaling behavior patterns provide an accurate basis for subsequent dynamic adjustment of the 3D spatial tile loading strategy, thereby optimizing resource scheduling and improving the interactive experience during the scaling process.

[0127] The intelligent display module, a 3D display system, performs intelligent display based on operation behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N precision levels.

[0128] Methods for intelligent display based on operational behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N precision levels include:

[0129] If the zoom behavior mode is slow zoom, the 3D display system will intelligently display the data for slow zoom.

[0130] If the scaling behavior mode is fast scaling, the operation behavior feature data and scaling behavior mode are input into the accuracy level prediction model to obtain the prediction accuracy level; the 3D display system performs intelligent display for fast scaling.

[0131] like Figure 3 As shown, the methods for intelligent display in a 3D display system for slow zooming include:

[0132] The second scaling ratio value corresponding to the completion of the scaling operation is obtained from the operation behavior feature data. The second scaling ratio value is matched with a pre-constructed scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the slow scaling target precision level. The scaling ratio-precision level matching table includes the scaling ratio value and the corresponding precision level.

[0133] From the three-dimensional spatial tile data sets corresponding to N precision levels, filter and slowly scale the three-dimensional spatial tile data set corresponding to the target precision level, denoted as the target three-dimensional spatial tile data set;

[0134] The 3D display system performs rendering and display operations based on the target 3D spatial tile data set, realizing dynamic switching of display precision according to the scaling ratio, improving visual continuity and system response efficiency during the scaling interaction process.

[0135] like Figure 4 As shown, the methods for intelligent display in a 3D display system for rapid zooming include:

[0136] The first scaling ratio value corresponding to the start of the scaling operation is obtained from the operation behavior feature data. The first scaling ratio value is matched with the pre-built scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the fast scaling target precision level.

[0137] Each 3D spatial tile in the fast-scaling target accuracy level 3D spatial tile dataset is clustered with all 3D spatial tiles in the prediction accuracy level 3D spatial tile dataset using a clustering algorithm to obtain the corresponding tile similarity. For each 3D spatial tile in the fast-scaling target accuracy level 3D spatial tile dataset, the 3D spatial tile with the highest similarity to it is selected from the prediction accuracy level 3D spatial tile dataset to form a one-to-one corresponding 3D spatial tile data pair.

[0138] It should be noted that clustering algorithms that can be used to obtain similarity include K-Means clustering, K-Medoids clustering, and DBSCAN density clustering.

[0139] Preset tile similarity threshold 1 and tile similarity threshold 2; tile similarity threshold 1 is greater than tile similarity threshold 2;

[0140] 3D spatial tile data pairs with tile similarity greater than or equal to tile similarity threshold one are assigned to the preloaded set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold one but greater than or equal to tile similarity threshold two are assigned to the progressive supplement set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold two are assigned to the unloaded set.

[0141] The 3D display system performs loading, delayed supplementation, and unloading rendering operations based on a preloaded set, a progressively supplemented set, and an unloaded set, respectively, to achieve smooth display and optimized resource scheduling during rapid scaling.

[0142] It should be noted that tile similarity is used to measure the degree of feature similarity between 3D spatial tile data at different precision levels. The value range of tile similarity is defined as [0, 1]. A tile similarity value closer to 1 indicates that the spatial location, size parameters, and error characteristics of the tile data are more similar, while a tile similarity value closer to 0 indicates greater feature differences between the tile data. In practical applications, both the first and second tile similarity thresholds are set within the range of [0, 1]. For example, the first tile similarity threshold can be set to 0.7, and the second tile similarity threshold can be set to 0.4.

[0143] The preloaded set is used to prioritize loading target tile data that is highly similar to the local features of the original viewport when zooming in or out rapidly. By quickly loading tiles from the preloaded set, the new viewport can be filled immediately after the zooming operation, ensuring scene continuity and smooth rendering, and avoiding visible holes or significant stuttering.

[0144] Progressive complement sets are used to gradually load tile data that has some similarity to the original local features but does not meet the requirements for immediate loading after the initial rapid scaling. The existence of progressive complement sets can refine the display details during the scaling stabilization phase, improve visual quality, and balance scaling response speed with final graphic accuracy.

[0145] The unloading set is used to identify tile data that has no effective similarity to the current prediction view and is redundantly occupying resources during rapid scaling. By unloading the tile data in the unloading set in a timely manner, system memory, cache, and bandwidth resources can be effectively released, improving overall system performance, reducing unnecessary resource waste, and ensuring the efficient and stable operation of the 3D display system.

[0146] An example of the scaling ratio-precision level matching table is shown in Table 2:

[0147] Table 2 Scaling Ratio - Precision Level Matching Table

[0148]

[0149]

[0150] It should be noted that the overall architecture diagram of the real-scene 3D geographic entity and spatial planning display system provided in this application is as follows: Figure 5 As shown, it mainly includes three layers: the hardware device layer, the data processing layer, and the intelligent interaction layer.

[0151] The hardware layer includes: 3D data acquisition devices, such as the RIEGL VZ-400 LiDAR, for acquiring real-world point cloud data; computing devices, such as the NVIDIA DGX A100 server, for performing high-performance point cloud processing, cluster analysis, and intelligent scheduling calculations; and display terminals, such as the Planar 8K 3D stereoscopic display, for high-resolution 3D visualization rendering and display.

[0152] The data processing layer includes a first processing module and a second processing module: the first processing module is used to perform spatial tile construction and precision level division, and constructs a three-dimensional spatial tile structure suitable for multi-scale display based on a multi-density sampling mechanism; the second processing module is used to perform tile screening and organization reconstruction based on scaling behavior patterns, and dynamically generate a set of loadable tiles by combining the predicted scaling results and precision requirements.

[0153] The intelligent interaction layer includes a behavior diagnosis module and an intelligent display module: the behavior diagnosis module is used to call the behavior diagnosis model based on user operation behavior feature data to determine whether the scaling behavior is fast zooming, slow zooming, fast zooming or slow zooming; the intelligent display module is used to execute dynamic adjustment and progressive tile loading strategies based on the recognition results to achieve smooth control and optimal resource utilization of the system under different scaling behaviors.

[0154] The mind map of the real-scene 3D geographic entity and spatial planning display method provided in this application is as follows: Figure 6 As shown, the overall process comprises four stages: data preprocessing, interactive feature extraction, intelligent decision-making, and dynamic rendering. The data preprocessing stage extracts the spatial structure and reconstructs the multi-scale structure of the original real-world point cloud data through point cloud bounding box construction, multi-level downsampling, and timestamp recording. The interactive feature extraction stage extracts zoom rate and direction identifiers based on user zoom operations to characterize interactive behavior. The intelligent decision-making stage intelligently identifies zoom behavior and plans resource scheduling through behavior pattern classification and resource prediction models. The dynamic rendering stage generates a preloaded set, a progressively supplemented set, and a tile unloading set based on the decision results, dynamically executing the loading, refinement, and release operations of 3D spatial tiles to achieve continuous display, resource optimization, and precision matching of the 3D scene during the zoom interaction process.

[0155] Example 2

[0156] Please see Figure 2 As shown, this embodiment provides a method for displaying real-world 3D geographic entities and spatial planning, including:

[0157] Construct a three-dimensional spatial tile data set corresponding to N precision levels based on real-world point cloud data with multi-density sampling;

[0158] Based on the zoom operation behavior in the 3D display system, operation behavior feature features are extracted to obtain operation behavior feature data;

[0159] Input the operational behavior feature data into the operational behavior diagnostic model to obtain the scaling behavior pattern;

[0160] The 3D display system uses intelligent display based on operation behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N precision levels.

[0161] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0162] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-scene 3D geographic entity and spatial planning display system, characterized in that, include: The first processing module constructs a set of three-dimensional spatial tile data corresponding to N precision levels based on real-world point cloud data sampled at multiple densities. The second processing module extracts operation behavior features based on the zoom operation behavior in the 3D display system to obtain operation behavior feature data. The behavior diagnosis module inputs the operation behavior feature data into the operation behavior diagnosis model to obtain the scaling behavior pattern; the scaling behavior pattern includes slow scaling and fast scaling. The intelligent display module of the 3D display system performs intelligent display based on operation behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N accuracy levels. When the scaling behavior mode is fast scaling, the operation behavior feature data and scaling behavior mode are input into the accuracy level prediction model to obtain the predicted accuracy level, and fast scaling intelligent display is performed based on the predicted accuracy level. The fast scaling intelligent display includes: The first scaling ratio value corresponding to the start of the scaling operation is obtained from the operation behavior feature data. The first scaling ratio value is matched with the pre-built scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the fast scaling target precision level. For each 3D spatial tile data in the 3D spatial tile data set at the target accuracy level, a clustering algorithm is used to determine the tile similarity between the 3D spatial tile data and each 3D spatial tile data in the 3D spatial tile data set at the prediction accuracy level. Then, the 3D spatial tile data with the highest tile similarity is selected from the 3D spatial tile data set at the prediction accuracy level to form a one-to-one corresponding 3D spatial tile data pair. Preset tile similarity threshold 1 and tile similarity threshold 2, where tile similarity threshold 1 is greater than tile similarity threshold 2; 3D spatial tile data pairs with tile similarity greater than or equal to tile similarity threshold one are assigned to the preloaded set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold one but greater than or equal to tile similarity threshold two are assigned to the progressive supplement set; 3D spatial tile data pairs with tile similarity less than tile similarity threshold two are assigned to the unloaded set. The 3D display system performs loading, delayed supplementation, and unloading rendering operations based on the preloaded set, the progressively supplemented set, and the unloaded set, respectively.

2. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 1, characterized in that, The slow scaling includes slow zooming in and slow zooming out; the fast scaling includes fast zooming in and fast zooming out. Methods for intelligent display based on operational behavior feature data, scaling behavior patterns, and a set of 3D spatial tile data corresponding to N precision levels include: If the scaling behavior mode is slow scaling, the 3D display system will intelligently display the data for slow scaling.

3. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 2, characterized in that, The methods used by 3D display systems to intelligently display images during slow zooming include: The second scaling ratio value corresponding to the completion of the scaling operation is obtained from the operation behavior feature data. The second scaling ratio value is matched with a pre-constructed scaling ratio-precision level matching table to obtain the corresponding precision level, which is denoted as the slow scaling target precision level. The scaling ratio-precision level matching table includes the scaling ratio value and the corresponding precision level. From the three-dimensional spatial tile data sets corresponding to N precision levels, filter and slowly scale the three-dimensional spatial tile data set corresponding to the target precision level, denoted as the target three-dimensional spatial tile data set; The 3D display system performs rendering and display operations based on the target 3D spatial tile data set.

4. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 1, characterized in that, The method for obtaining the operational behavior feature data includes: When a scaling operation is detected, the system timestamp at the start of the scaling operation and the scaling ratio value are obtained and recorded as the first system timestamp and the first scaling ratio value, respectively. When the scaling operation is detected to be complete, the system timestamp and scaling ratio value at the time of completion of the scaling operation are obtained and recorded as the second system timestamp and the second scaling ratio value, respectively. The scaling change is obtained by subtracting the second scaling value from the first scaling value; if the scaling change is greater than zero, the scaling direction indicator is set to the zoom-in indicator; if the scaling change is less than zero, the scaling direction indicator is set to the zoom-out indicator. The scaling time interval is obtained by subtracting the second system timestamp from the first system timestamp. The scaling rate is obtained by taking the absolute value of the scaling change magnitude and dividing it by the scaling change time interval. The second scaling factor, scaling change magnitude, scaling direction identifier, scaling change time interval, and scaling change rate are used to construct the operation behavior feature data.

5. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 1, characterized in that, Methods for constructing three-dimensional spatial tile datasets corresponding to N precision levels based on real-world point cloud data using multi-density sampling include: Acquire real-world location cloud data, which includes spatial three-dimensional coordinates; Construct a bounding box of real-site cloud data; The multi-scale precision control requirements of the pre-collected 3D display system are input into the precision parameter setting model to obtain the precision setting parameters. The precision setting parameters include N precision levels and corresponding layered sampling granularity. The multi-scale precision control requirements include minimum magnification ratio, maximum magnification ratio, minimum reduction ratio, and maximum reduction ratio. Based on N precision levels and corresponding hierarchical sampling granularity, the bounding boxes of real point cloud data are downsampled to obtain the downsampled point cloud data sets corresponding to N precision levels. Spatial slicing is performed on hierarchical sampling granularity and downsampled point cloud data sets with N precision levels to obtain three-dimensional spatial tile data sets corresponding to N precision levels.

6. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 5, characterized in that, Methods for constructing a bounding box of real-location cloud data include: Obtain all three-dimensional coordinate points in the real-world point cloud data. Based on all three-dimensional coordinate points, determine the boundary values ​​of the real-world point cloud data in three dimensions in sequence, including the maximum and minimum coordinate values ​​in the X-axis direction, the maximum and minimum coordinate values ​​in the Y-axis direction, and the maximum and minimum coordinate values ​​in the Z-axis direction. The length of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the X-axis direction; the width of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Y-axis direction; and the height of the point cloud data bounding box is obtained by subtracting the maximum and minimum coordinate values ​​in the Z-axis direction. The center coordinates of the point cloud data bounding box are then obtained based on its length, width, and height. Using the length, width, height, and center coordinates of the real-world point cloud data bounding box as parameters, a geometric boundary volume is constructed that encloses the entire real-world point cloud data space, thus obtaining the real-world point cloud data bounding box.

7. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 5, characterized in that, The method for downsampling the bounding boxes of real-world point cloud data based on N precision levels and corresponding hierarchical sampling granularities to obtain downsampled point cloud data sets corresponding to N precision levels includes: S100: Let the initial value of n be 1, and the range of n is from 1 to N; S101: Using the granularity of the nth precision level as the precision, divide along the X-axis direction. Each point cloud data grid cell is divided along the Y-axis. Each point cloud data grid cell is divided along the Z-axis. A total of point cloud data grid cells were obtained, that is, a total of Each point cloud data grid unit; Indicates rounding up; The length of the real-world cloud data bounding box. The width of the virtual cloud data bounding box for the real-world location. The height of the real-world cloud data bounding box. This represents the stratified sampling granularity for the nth precision level. S102: Will The point cloud data grid cells are used to construct the downsampled point cloud data set of the nth precision level; the downsampled point cloud data set maintains the same spatial boundary range as the real point cloud data bounding box; S103: Let n = n + 1. If n is less than or equal to N, continue executing S101 to S102; if n is greater than N, end the current process.

8. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 5, characterized in that, The method for obtaining three-dimensional spatial tile data sets corresponding to N precision levels by spatially segmenting a hierarchical sampling granularity and downsampled point cloud data sets based on N precision levels includes: S200: Let the initial value of n be 1, and the range of n is from 1 to N; S201: Obtain the number of point cloud data grid cells in the nth downsampled point cloud data set. ;for Each point cloud data grid cell constructs a corresponding downsampled point cloud data sub-bounding box; the downsampled point cloud data sub-bounding box is a 3D spatial tile, resulting in... A three-dimensional spatial tile; S202: respectively for The length, width, height, and center coordinates of each three-dimensional spatial tile are extracted to obtain... The three-dimensional spatial tile corresponds to the tile's length, width, height, and center coordinates; with the nth precision level of layered sampling granularity. The factor is used as a three-dimensional spatial tile error parameter. A constant greater than 0; S203: Will The three-dimensional spatial tile error parameters, three-dimensional spatial tile length, three-dimensional spatial tile width, three-dimensional spatial tile height, and three-dimensional spatial tile center coordinates corresponding to each three-dimensional spatial tile are constructed into a... The three-dimensional spatial tile index information corresponding to each three-dimensional spatial tile; Each three-dimensional space tile and its corresponding three-dimensional space tile index information constitute a... Three-dimensional spatial tile data, The data of individual three-dimensional spatial tiles are used to construct a three-dimensional spatial tile data set; S204: Let n = n + 1. If n is less than or equal to N, continue executing S201 to S103; if n is greater than N, end the current process.

9. The real-scene three-dimensional geographic entity and spatial planning display system according to claim 1, characterized in that, The training method for the operational behavior diagnostic model includes: A pre-constructed operational behavior diagnostic dataset is provided, which includes XW sets of operational behavior diagnostic data and the scaling behavior patterns corresponding to the XW sets of operational behavior diagnostic data, where XW is a positive integer greater than 0, and the operational behavior diagnostic data includes operational behavior feature data. The operational behavior diagnosis dataset is divided into a training set and a validation set. The training set is used for parameter learning, and the validation set is used for evaluating the model's generalization ability. Support vector machine (SVM) model is used as the operational behavior diagnosis model. Operational behavior feature data is converted into feature vector input, nonlinear features are extracted through hidden layers, and the scaling behavior pattern with the highest probability is output as the final prediction result in the output layer. The training objective is to minimize the cross-entropy loss function. Training stops when the accuracy of the validation set reaches a preset threshold.

Images