Spatial Grid Data Visualization System and Method Based on 3D Model

Abstract

The present invention discloses a spatial grid data visualization system and method based on a three-dimensional model, which relates to the technical fields of computer graphics and real-time visualization. The method first analyzes the original grid data to identify the independent components of the model and their connection relationships; then constructs three independent representations: a topological skeleton layer, a parameterized bounding layer, and a geometric data layer; through quantization calculation and task segmentation, parallelly generates topological skeleton and parameterized bounding volume data, ensuring that the total generation time is less than a preset first-frame display time threshold; then submits for rendering to generate an initial three-dimensional image with complete component semantics; finally, while displaying the initial image, incrementally generates geometric data and performs progressive visual fusion with it; through hierarchical data organization and timing control, the present invention solves the problem of interface lag caused by the serial nature of data generation and display in traditional methods, and achieves a seamless transition from instant structure display to fine geometric rendering.

Description

Technical Field

[0001] This invention relates to the fields of computer graphics and real-time visualization technology, specifically to a spatial grid data visualization system and method based on a three-dimensional model. Background Technology

[0002] In the field of 3D spatial data visualization, such as geographic information systems, large-scale point cloud rendering, computational fluid dynamics result analysis, or medical volume data reconstruction, users often need to load and browse complex 3D spatial mesh datasets composed of massive discrete spatial units (such as vertices, voxels, or particles) in real time. Existing technologies for processing such spatial mesh data follow an inherent linear processing paradigm: the system must first completely read all geometric and attribute data describing the entire spatial domain from storage media or network streams, complete the reconstruction and organization of the data structure in memory, and only then can it call the graphics rendering pipeline to generate the first frame image and output it to the display device.

[0003] This paradigm leads to a pervasive technical problem: the user interface remains unresponsive to any effective visual feedback until the complete spatial grid data has been fully loaded and processed. Users are faced with a blank or static screen, unable to perceive the spatial structure, distribution characteristics, or any continuity information represented by the data, and all interactive intentions are forcibly blocked.

[0004] Take loading a high-precision digital terrain model as an example: This model consists of millions of regular or irregular spatial grid points, each containing three-dimensional coordinates and possible elevation, attribute, and other information. When loading using existing technologies, from the moment the user issues the command until the first terrain map appears, the interface experiences a continuous, unresponsive waiting period. During this time, although storage I / O and CPU computing resources are under high load, the screen cannot display any schematic image of terrain undulations, regional outlines, or overall extent. The user cannot confirm whether the data is loading correctly, nor can they gain a macroscopic understanding of the spatial layout in advance. This abrupt shift from an "informationless" state to a "fully detailed" state severely disrupts the user's cognitive process, reduces analytical efficiency, and is particularly problematic in applications requiring real-time observation and decision-making.

[0005] The system cannot prioritize extracting and presenting lightweight structural information that is easy to parse quickly over massive amounts of geometric detail. Therefore, the time spent on data processing is forcibly tied to the timing of visual feedback, resulting in unavoidable interaction delays.

[0006] To address the above problems, this invention proposes a spatial grid data visualization system and method based on a three-dimensional model. Summary of the Invention

[0007] The purpose of this invention is to provide a spatial grid data visualization system and method based on a three-dimensional model to solve the problems raised in the prior art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: A spatial grid data visualization method based on a 3D model includes the following steps: S1. Analyze the original mesh data of the 3D model, identify and separate the set of independent parts of the model and the connection relationships between the parts; S2. Based on the components and connection relationships, a three-layer data representation is logically constructed. The first layer is the topological skeleton layer describing the connection relationships, the second layer is the bounding volume collection layer that uses parameterized primitives to represent the orientation and size of the components, and the third layer is the geometric data layer containing complete geometric details. S3. Generate the data of the topology skeleton layer and the data of the bounding volume set layer in parallel, and adjust the processing flow so that the sum of their generation times is less than the preset first frame display time threshold. S4. Submit the generated topological skeleton and bounding volume set data for rendering to generate and display an initial 3D image with complete component semantics; S5. While displaying the initial 3D image, continue to generate the third layer of geometric data, and replace and merge it with the corresponding bounding volume on the screen in real time according to the component identification, so as to complete the gradual transition of the visualization details from parametric representation to geometric data.

[0009] S1 further includes the following: Obtain the original mesh data of the three-dimensional model, the original mesh data including a vertex array storing three-dimensional spatial coordinates and patch index information defining surface triangles; The original mesh data is processed according to the preset geometric segmentation rules, and a group of patches that meet the spatial continuity condition are aggregated to form independent components, and a unique identifier is assigned to each component. For each component, traverse all the vertex coordinates contained therein, calculate its coordinate extreme values ​​in the three spatial dimensions, and use the coordinate extreme values ​​as the spatial range boundary of the component. By detecting whether there are shared edges between the facets contained in different components, the geometric contact between components is determined, and an adjacency record is established for component pairs that have contact.

[0010] S2 further includes the following: Based on the unique identifier of the component, each component is mapped to a topology node. Based on the adjacency relationship between the components, connecting edges are established between the topology nodes that have an adjacency relationship to form a topology skeleton layer. For each component, a parameterized bounding body that completely surrounds the component is calculated based on the spatial extent boundary of the component, forming a bounding body collection layer; The center coordinates of the parameterized bounding volume are determined by the average of the coordinate extreme values ​​of the spatial range boundary in three dimensions, and the axial dimension of the parameterized bounding volume is determined by the difference of the coordinate extreme values ​​of the spatial range boundary in three dimensions. The set of facets corresponding to each component is used as the geometric data of that component to form a geometric data layer; The topology skeleton layer, parameterized bounding layer, and geometric data layer are independent of each other in terms of data, and are stored and transmitted separately.

[0011] S3 further includes the following: Obtain the total number of edges E of the topological skeleton layer and the total number of components P of the parameterized bounding layer; Obtain the unit time te for processing a single topological edge and the unit time tp for processing the bounding volume data of a single component from the system processing unit; The estimated total time Te for generating the topology skeleton layer and parameterized bounding layer data is calculated using the following formula: Te = E × te + P × tp; The estimated total time Te is compared with the preset first frame display time threshold Tmax; If the estimated total time Te is not greater than the time threshold Tmax, then the tasks of generating the graphics drawing data of the topology skeleton layer and generating the geometric primitives and rendering parameters of the parameterized bounding layer are executed in parallel. If the estimated total time Te is greater than the time threshold Tmax, the generation tasks of the topology skeleton layer and the parameterized enclosing layer are logically divided, and Te is recalculated based on the divided subtasks until the condition Te≤Tmax is met. Then, parallel generation tasks are executed according to the task plan after the condition is met.

[0012] The logical segmentation of the task of generating the topology skeleton layer and the parameterized bounding layer includes the following: The task of generating the parameterized bounding layer is divided into several consecutive generation subsets based on the total number of components P. Each subset contains bounding volume data processing tasks for K components, where K is a positive integer less than P. The estimated sub-time Tb for generating the first subset of topological skeleton layer data and parameterized bounding layer data is calculated using the following formula: Tb = E × te + K × tp; Adjust the value of K so that Tb ≤ Tmax; Based on the final determined K value, the task of generating the parameterized bounding layer is planned to be executed in parallel with the task of generating the topology skeleton layer in a subset order, with the generation and transmission of each subset performed independently.

[0013] S4 further includes the following: The generated topological skeleton layer corresponding to the graphic drawing data is submitted to the graphics processing unit; the graphics processing unit calculates the projection coordinates of the graphic markers and line segments in the screen space through coordinate transformation based on the spatial position of the graphic markers and line segments, and generates the first set of pixel color and depth data according to the preset vertex shading rules and fragment shading rules to complete the first round of rendering calculation. The generated parameterized bounding layer corresponding to the geometric primitives and rendering parameters is submitted to the graphics processing unit; the graphics processing unit performs the same coordinate transformation based on the vertex coordinate data of the geometric primitives to obtain their screen space projection coordinates, and generates a second set of pixel color and depth data according to the color and transparency defined by the corresponding rendering parameters through the fragment shading rules to complete the second round of rendering calculation; The graphics processing unit synthesizes the first and second sets of pixel color and depth data according to the rules of depth testing and color mixing, and outputs a synthesized image containing all component topology nodes, connecting edges and parameterized bounding volumes, which is then displayed as the initial 3D image with complete component semantics.

[0014] S5 further includes the following: While displaying the initial 3D image, the data generation of the geometric data layer is initiated. The data generation is performed on each component, and based on its component identifier, the corresponding set of facets is converted into geometric mesh data that can be directly drawn by the graphics processing unit. For each component, based on the axial dimension of its parametric bounding volume, an approximate spatial volume of the component is obtained by calculating the product of the components of the axial dimension in three dimensions; all components are then sorted in descending order based on the approximate spatial volume to generate a geometric replacement priority sequence. According to the geometric replacement priority sequence, the geometric mesh data of the generated components is submitted to the graphics processing unit; The graphics processing unit adjusts the display opacity of both the bounding volume and the geometric mesh data of the same component by linear interpolation in consecutive image frames, based on a preset total number of replacement frames. The linear interpolation adjustment gradually decreases the display opacity of the bounding volume from its initial value to complete transparency, while simultaneously increasing the display opacity of the corresponding geometric mesh data from complete transparency back to its initial value. Ultimately, the geometric mesh data visually completely replaces its corresponding bounding volume, achieving a gradual enhancement of visual details.

[0015] A spatial grid data visualization system based on a 3D model includes a data parsing and preprocessing module, a hierarchical data construction module, a temporal control and parallel task generation module, an initial rendering and image synthesis module, and a progressive geometric fusion and display module. The data parsing and preprocessing module is used to acquire raw grid data, divide the model into multiple independent components, calculate the spatial extent boundaries of each component, and determine the adjacency relationship between components. Based on the output of the data parsing and preprocessing module, the hierarchical data construction module logically constructs and generates independent topological skeleton layer data, parameterized bounding layer data, and geometric data layer data. The timing control and parallel task generation module performs logical analysis and calculation based on the data scale and system performance parameters of the topology skeleton layer and parameterized enclosing layer, generates a parallel task scheduling instruction that satisfies the display time constraint of the first frame, and drives the generation task of the topology skeleton layer data and parameterized enclosing layer data. The initial rendering and image synthesis module is used to receive the data output by the timing control and parallel task generation module, schedule the graphics processing unit to execute the rendering calculation of the topological skeleton and parameterized bounding volume in sequence, and fuse the calculation results into an initial three-dimensional image with complete component semantics. The progressive geometry blending and display module is used to schedule the generation of geometric data layer data during the initial 3D image display, and control the graphics processing unit to visually blend and replace the parameterized bounding volume and geometric mesh data according to the priority sequence, so as to complete the progressive enhancement of visualization details.

[0016] The initial rendering and image compositing module includes an image processing unit: In the initial rendering and image compositing module, the graphics processing unit receives the graphics data of the topological skeleton layer and the parameterized bounding layer, performs the transformation projection of vertex coordinates from the three-dimensional model space to the two-dimensional screen space, and generates the corresponding pixel color and depth information according to the preset shading rules. In the progressive geometry blending and display module, the graphics processing unit performs linear interpolation calculation and real-time rendering of the display opacity of the parameterized bounding volume and the subsequently delivered geometric mesh at the same spatial location based on control parameters.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention decouples model data into three logical layers: topological skeleton, parametric bounding volume, and geometric details. It prioritizes and processes the first two layers in parallel, enabling the system to render an initial 3D image representing the overall structure and component relationships of the model before the complete geometric data is generated. This fundamentally eliminates the "white screen" waiting period caused by traditional methods that require waiting for all data processing to complete. 2. After displaying the initial image, the system incrementally generates and integrates precise geometric details according to a predetermined strategy, achieving seamless and progressive enhancement of the visualized content from a macroscopic semantic framework to a microscopic precise shape. Users receive effective visual feedback from the very beginning of the operation and can gradually deepen their understanding of the model through continuous detail refinement, greatly improving the smoothness and continuity of interactive exploration. Attached Figure Description

[0018] Figure 1 This is a flowchart of the spatial grid data visualization method based on a three-dimensional model according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: Figure 1 As shown, the present invention provides a technical solution. A spatial grid data visualization method based on a 3D model includes the following steps: S1. Analyze the original mesh data of the 3D model, identify and separate the set of independent parts of the model and the connection relationships between the parts; S2. Based on the components and connection relationships, a three-layer data representation is logically constructed. The first layer is the topological skeleton layer describing the connection relationships, the second layer is the bounding volume collection layer that uses parameterized primitives to represent the orientation and size of the components, and the third layer is the geometric data layer containing complete geometric details. S3. Generate the data of the topology skeleton layer and the data of the bounding volume set layer in parallel, and adjust the processing flow so that the sum of their generation times is less than the preset first frame display time threshold. S4. Submit the generated topological skeleton and bounding volume set data for rendering to generate and display an initial 3D image with complete component semantics; S5. While displaying the initial 3D image, continue to generate the third layer of geometric data, and replace and merge it with the corresponding bounding volume on the screen in real time according to the component identification, so as to complete the gradual transition of the visualization details from parametric representation to geometric data.

[0021] S1 further includes the following: Obtain the original mesh data of the three-dimensional model, the original mesh data including a vertex array storing three-dimensional spatial coordinates and patch index information defining surface triangles; The original mesh data is processed according to the preset geometric segmentation rules, and a group of patches that meet the spatial continuity condition are aggregated to form independent components, and a unique identifier is assigned to each component. For each component, traverse all the vertex coordinates contained therein, calculate its coordinate extreme values ​​in the three spatial dimensions, and use the coordinate extreme values ​​as the spatial range boundary of the component. By detecting whether there are shared edges between the facets contained in different components, the geometric contact between components is determined, and an adjacency record is established for component pairs that have contact.

[0022] S2 further includes the following: Based on the unique identifier of the component, each component is mapped to a topology node. Based on the adjacency relationship between the components, connecting edges are established between the topology nodes that have an adjacency relationship to form a topology skeleton layer. For each component, a parameterized bounding body that completely surrounds the component is calculated based on the spatial extent boundary of the component, forming a bounding body collection layer; The center coordinates of the parameterized bounding volume are determined by the average of the coordinate extreme values ​​of the spatial range boundary in three dimensions, and the axial dimension of the parameterized bounding volume is determined by the difference of the coordinate extreme values ​​of the spatial range boundary in three dimensions. The set of facets corresponding to each component is used as the geometric data of that component to form a geometric data layer; The topology skeleton layer, parameterized bounding layer, and geometric data layer are independent of each other in terms of data, and are stored and transmitted separately.

[0023] S3 further includes the following: Obtain the total number of edges E of the topological skeleton layer and the total number of components P of the parameterized bounding layer; Obtain the unit time te for processing a single topological edge and the unit time tp for processing the bounding volume data of a single component from the system processing unit; The estimated total time Te for generating the topology skeleton layer and parameterized bounding layer data is calculated using the following formula: Te = E × te + P × tp; The estimated total time Te is compared with the preset first frame display time threshold Tmax; If the estimated total time Te is not greater than the time threshold Tmax, then the tasks of generating the graphics drawing data of the topology skeleton layer and generating the geometric primitives and rendering parameters of the parameterized bounding layer are executed in parallel. If the estimated total time Te is greater than the time threshold Tmax, the generation tasks of the topology skeleton layer and the parameterized enclosing layer are logically divided, and Te is recalculated based on the divided subtasks until the condition Te≤Tmax is met. Then, parallel generation tasks are executed according to the task plan after the condition is met.

[0024] The logical segmentation of the task of generating the topology skeleton layer and the parameterized bounding layer includes the following: The task of generating the parameterized bounding layer is divided into several consecutive generation subsets based on the total number of components P. Each subset contains bounding volume data processing tasks for K components, where K is a positive integer less than P. The estimated sub-time Tb for generating the first subset of topological skeleton layer data and parameterized bounding layer data is calculated using the following formula: Tb = E × te + K × tp; Adjust the value of K so that Tb ≤ Tmax; Based on the final determined K value, the task of generating the parameterized bounding layer is planned to be executed in parallel with the task of generating the topology skeleton layer in a subset order, with the generation and transmission of each subset performed independently.

[0025] Through the above calculations, comparisons, and logical segmentation, the system can proactively control the data generation task to ensure that, under any complex model, the total generation time of the topology skeleton layer and parameterized enclosing layer data does not exceed the first frame display time threshold. This quantitative control mechanism is one of the core innovations of this invention. It enables the system to ensure immediacy from the source of data generation, thereby solving the interface lag problem caused by the serial generation and display of data in the background technology.

[0026] S4 further includes the following: The generated topological skeleton layer corresponding to the graphic drawing data is submitted to the graphics processing unit; the graphics processing unit calculates the projection coordinates of the graphic markers and line segments in the screen space through coordinate transformation based on the spatial position of the graphic markers and line segments, and generates the first set of pixel color and depth data according to the preset vertex shading rules and fragment shading rules to complete the first round of rendering calculation. The generated parameterized bounding layer corresponding to the geometric primitives and rendering parameters is submitted to the graphics processing unit; the graphics processing unit performs the same coordinate transformation based on the vertex coordinate data of the geometric primitives to obtain their screen space projection coordinates, and generates a second set of pixel color and depth data according to the color and transparency defined by the corresponding rendering parameters through the fragment shading rules to complete the second round of rendering calculation; The graphics processing unit synthesizes the first and second sets of pixel color and depth data according to the rules of depth testing and color mixing, and outputs a synthesized image containing all component topology nodes, connecting edges and parameterized bounding volumes, which is then displayed as the initial 3D image with complete component semantics.

[0027] Because the S3 control ensures that the topology skeleton layer and parameterized bounding layer data can be generated in a very short time, this step can start rendering almost immediately after the user operation and quickly synthesize an initial 3D image with complete component semantics. This allows users to gain an intuitive understanding of the overall structure of the model without having to wait for the traditionally lengthy loading process, achieving a fundamental shift from 'waiting for loading' to 'instant interaction'.

[0028] S5 further includes the following: While displaying the initial 3D image, the data generation of the geometric data layer is initiated. The data generation is performed on each component, and based on its component identifier, the corresponding set of facets is converted into geometric mesh data that can be directly drawn by the graphics processing unit. For each component, based on the axial dimension of its parametric bounding volume, an approximate spatial volume of the component is obtained by calculating the product of the components of the axial dimension in three dimensions; all components are then sorted in descending order based on the approximate spatial volume to generate a geometric replacement priority sequence. According to the geometric replacement priority sequence, the geometric mesh data of the generated components is submitted to the graphics processing unit; The graphics processing unit adjusts the display opacity of both the bounding volume and the geometric mesh data of the same component by linear interpolation in consecutive image frames, based on a preset total number of replacement frames. The linear interpolation adjustment gradually decreases the display opacity of the bounding volume from its initial value to complete transparency, while simultaneously increasing the display opacity of the corresponding geometric mesh data from complete transparency back to its initial value. Ultimately, the geometric mesh data visually completely replaces its corresponding bounding volume, achieving a gradual enhancement of visual details.

[0029] A spatial grid data visualization system based on a 3D model includes a data parsing and preprocessing module, a hierarchical data construction module, a temporal control and parallel task generation module, an initial rendering and image synthesis module, and a progressive geometric fusion and display module. The data parsing and preprocessing module is used to acquire raw grid data, divide the model into multiple independent components, calculate the spatial extent boundaries of each component, and determine the adjacency relationship between components. Based on the output of the data parsing and preprocessing module, the hierarchical data construction module logically constructs and generates independent topological skeleton layer data, parameterized bounding layer data, and geometric data layer data. The timing control and parallel task generation module performs logical analysis and calculation based on the data scale and system performance parameters of the topology skeleton layer and parameterized enclosing layer, generates a parallel task scheduling instruction that satisfies the display time constraint of the first frame, and drives the generation task of the topology skeleton layer data and parameterized enclosing layer data. The initial rendering and image synthesis module is used to receive the data output by the timing control and parallel task generation module, schedule the graphics processing unit to execute the rendering calculation of the topological skeleton and parameterized bounding volume in sequence, and fuse the calculation results into an initial three-dimensional image with complete component semantics. The progressive geometry blending and display module is used to schedule the generation of geometric data layer data during the initial 3D image display, and control the graphics processing unit to visually blend and replace the parameterized bounding volume and geometric mesh data according to the priority sequence, so as to complete the progressive enhancement of visualization details.

[0030] The initial rendering and image compositing module includes an image processing unit: In the initial rendering and image compositing module, the graphics processing unit receives the graphics data of the topological skeleton layer and the parameterized bounding layer, performs the transformation projection of vertex coordinates from the three-dimensional model space to the two-dimensional screen space, and generates the corresponding pixel color and depth information according to the preset shading rules. In the progressive geometry blending and display module, the graphics processing unit performs linear interpolation calculation and real-time rendering of the display opacity of the parameterized bounding volume and the subsequently delivered geometric mesh at the same spatial location based on control parameters.

[0031] Consider a specific application scenario: real-time visualization and browsing of a large-scale smart park's 3D spatial mesh data. This spatial mesh data describes a high-tech park covering approximately 5 square kilometers, containing about 120 independent 3D models, such as distinctive office buildings, R&D centers, vehicle models in parking lots, landscape sculptures, and road facilities. Its original mesh data consists of over 2.8 million triangular faces, used to accurately describe the appearance and structural details of each model.

[0032] First, data parsing and model partitioning are performed. The system reads the raw data of the spatial grid from the storage medium. This data consists of an array of vertices recording 3D spatial coordinates and index information defining triangular faces. Based on preset model boundaries and logical affiliation rules, the system automatically aggregates all triangular faces belonging to the same independent object (such as a single building or a single car) together to form an independent model unit. For this purpose, each model unit is assigned a unique identifier, such as "BuildingA", "Car023", or "Sculpture01". Subsequently, the system traverses all vertex coordinates contained in each model unit, calculating the minimum and maximum values ​​of the unit in the X, Y, and Z dimensions. These six coordinate values ​​together define the spatial boundary of the unit. To establish spatial or logical relationships between model units, the system establishes connection records for related model unit pairs based on Geographic Information System (GIS) data or preset rules (such as roads connecting buildings).

[0033] Based on the above analysis results, the system logic constructs a three-level data representation. The first layer is the topological skeleton layer: each model unit is abstracted as a topological node, and each connection between units is abstracted as an edge connecting two nodes. In this way, the 120 model units and their connections in the entire park are mapped into a topological graph consisting of 120 nodes and several connecting edges, with a very small amount of data.

[0034] The second layer is the parametric bounding layer: for each model element, the system calculates a simplified bounding volume for that element using its calculated spatial extent boundaries (i.e., six extreme values). For example, for a building, its bounding volume might be a cuboid with matching position and size; for a car, its bounding volume might be an ellipsoid. These bounding volumes can be described using only a few parameters, such as center coordinates and axial dimensions.

[0035] The third layer is the traditional geometric data layer, which is the original set of triangle faces containing all the details for each model unit. The data in these three layers are completely independent in terms of logic and storage.

[0036] Next comes the crucial data generation timing control. The system obtains the total number of edges in the topology skeleton layer (assuming E = 160) and the total number of model units in the parameterized bounding layer (P = 120). Based on the known performance benchmark of the system's processing unit (CPU), the unit time te for processing a single topology edge is determined to be approximately 0.05 milliseconds, and the unit time tp for processing the bounding volume data of a single model unit is approximately 0.5 milliseconds. The preset first frame display time threshold Tmax is 50 milliseconds. The system performs a preliminary calculation using the formula Te = E × te + P × tp, resulting in Te = 160 × 0.05 + 120 × 0.5 = 8 + 60 = 68 milliseconds. Since the estimated total time of 68 milliseconds is greater than the threshold of 50 milliseconds, the system triggers the logical segmentation mechanism.

[0037] The logical segmentation process is as follows: The system splits the task of generating 120 model units in the parameterized enclosing layer. It attempts to set each subset to contain K units. According to the subtask prediction formula Tb=E×te+K×tp, the goal is to make Tb≤Tmax. After adjustment and calculation, when K=75, Tb=8+75×0.5=45.5 milliseconds, which meets the requirement of less than 50 milliseconds.

[0038] Therefore, the system decides to divide the bounding volume generation task into two subsets: the first subset contains the first 75 units, and the second subset contains the remaining 45 units. The system plans to first execute the generation tasks of all topological skeleton data (160 edges) and the first subset of bounding volume data (75 units) in parallel.

[0039] After data generation, the rendering and display phase begins immediately. The graphical data (node ​​markers and connecting edges) generated from the topology skeleton layer is submitted to the graphics processing unit (GPU). The GPU performs coordinate transformations and coloring based on the spatial positions of these elements, generating the first set of pixel data. Next, the geometric data and rendering parameters of the 75 parametric bounding volumes of the first subset are also submitted to the GPU. The GPU performs rendering calculations, generating the second set of pixel data. Finally, the GPU synthesizes the two sets of pixel data according to depth testing and color blending rules, drawing an image on the screen: a 3D layout overview of the park composed of nodes, connecting lines, and 75 colored basic shapes (representing buildings, vehicles, etc.). Although this image lacks detail, it clearly shows the complete semantic structure of the distribution, type, and connection relationships of the 120 model units within the park. The time from the user issuing the command to the display of this image is controlled within approximately 45.5 milliseconds, achieving near-instantaneous visual feedback.

[0040] While the user begins observing and interacting with the initial 3D layout, the system continues to execute two progressive tasks in the background: First, it continues to generate and transmit the bounding volume data of the second subset, gradually displaying it all. Second, it initiates the most time-consuming geometric data layer processing, progressively calculating the precise 3D model data for each model unit. The system calculates the approximate spatial volume of each model unit based on its bounding volume dimensions, sorts them by volume from largest to smallest, and generates a geometric replacement priority sequence, prioritizing the replacement of visually prominent models (such as tall landmarks or sculptures in the central square).

[0041] The replacement process is performed by the GPU using visual blending. For the highest-priority model unit, the GPU begins rendering the precise geometric mesh of that unit in the same spatial location as its corresponding parametric bounding volume is displayed. Over a preset number of frames, the GPU performs linear interpolation of the opacities of both: the opacity of the bounding volume linearly decreases from 100% to 0%, while the opacity of the precise geometric mesh linearly increases from 0% to 100%. Ultimately, the high-precision appearance of the model smoothly and seamlessly replaces the simple bounding volume shape. All models undergo this progressive replacement sequentially according to their priority.

[0042] Ultimately, users first see a comprehensible overview of the overall layout and relationships of elements within the park in a very short time. Then, without interrupting the interaction, they watch as the details of key models, such as building exteriors and vehicle shapes, gradually become more refined and realistic, until the entire park scene fully displays its original, precise geometric details. The entire system, through hierarchical data organization, quantitative temporal control, and progressive rendering integration, fundamentally changes the predicament of users facing long "white screen" waits when loading complex 3D scenes in the past, achieving a seamless transition from immediate structural understanding to refined geometric observation.

[0043] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A spatial grid data visualization method based on a 3D model, characterized in that: Includes the following steps: S1. Analyze the original mesh data of the 3D model, identify and separate the set of independent parts of the model and the connection relationships between the parts; S2. Based on the components and connection relationships, a three-layer data representation is logically constructed. The first layer is the topological skeleton layer describing the connection relationships, the second layer is the bounding volume collection layer that uses parameterized primitives to represent the orientation and size of the components, and the third layer is the geometric data layer containing complete geometric details. S3. Generate the data of the topology skeleton layer and the data of the bounding volume set layer in parallel, and adjust the processing flow so that the sum of their generation times is less than the preset first frame display time threshold. S4. Submit the generated topological skeleton and bounding volume set data for rendering to generate and display an initial 3D image with complete component semantics; S5. While displaying the initial 3D image, continue to generate the third layer of geometric data, and replace and merge it with the corresponding bounding volume on the screen in real time according to the component identification, so as to complete the gradual transition of the visualization details from parametric representation to geometric data.

2. The spatial grid data visualization method based on a three-dimensional model according to claim 1, characterized in that: S1 further includes the following: Obtain the original mesh data of the three-dimensional model, the original mesh data including a vertex array storing three-dimensional spatial coordinates and patch index information defining surface triangles; The original mesh data is processed according to the preset geometric segmentation rules, and a group of patches that meet the spatial continuity condition are aggregated to form independent components, and a unique identifier is assigned to each component. For each component, traverse all the vertex coordinates contained therein, calculate its coordinate extreme values ​​in the three spatial dimensions, and use the coordinate extreme values ​​as the spatial range boundary of the component. By detecting whether there are shared edges between the facets contained in different components, the geometric contact between components is determined, and an adjacency record is established for component pairs that have contact.

3. The spatial grid data visualization method based on a three-dimensional model according to claim 2, characterized in that: S2 further includes the following: Based on the unique identifier of the component, each component is mapped to a topology node. Based on the adjacency relationship between the components, connecting edges are established between the topology nodes that have an adjacency relationship to form a topology skeleton layer. For each component, a parameterized bounding body that completely surrounds the component is calculated based on the spatial extent boundary of the component, forming a bounding body collection layer; The center coordinates of the parameterized bounding volume are determined by the average of the coordinate extreme values ​​of the spatial range boundary in three dimensions, and the axial dimension of the parameterized bounding volume is determined by the difference of the coordinate extreme values ​​of the spatial range boundary in three dimensions. The set of facets corresponding to each component is used as the geometric data of that component to form a geometric data layer; The topology skeleton layer, parameterized bounding layer, and geometric data layer are independent of each other in terms of data, and are stored and transmitted separately.

4. The spatial grid data visualization method based on a three-dimensional model according to claim 1, characterized in that: S3 further includes the following: Obtain the total number of edges E of the topological skeleton layer and the total number of components P of the parameterized bounding layer; Obtain the unit time te for processing a single topological edge and the unit time tp for processing the bounding volume data of a single component from the system processing unit; The estimated total time Te for generating the topology skeleton layer and parameterized bounding layer data is calculated using the following formula: Te = E × te + P × tp; The estimated total time Te is compared with the preset first frame display time threshold Tmax; If the estimated total time Te is not greater than the time threshold Tmax, then the tasks of generating the graphics drawing data of the topology skeleton layer and generating the geometric primitives and rendering parameters of the parameterized bounding layer are executed in parallel. If the estimated total time Te is greater than the time threshold Tmax, the generation tasks of the topology skeleton layer and the parameterized enclosing layer are logically divided, and Te is recalculated based on the divided subtasks until the condition Te≤Tmax is met. Then, parallel generation tasks are executed according to the task plan after the condition is met.

5. The spatial grid data visualization method based on a three-dimensional model according to claim 4, characterized in that: The logical segmentation of the task of generating the topology skeleton layer and the parameterized bounding layer includes the following: The task of generating the parameterized bounding layer is divided into several consecutive generation subsets based on the total number of components P. Each subset contains bounding volume data processing tasks for K components, where K is a positive integer less than P. The estimated sub-time Tb for generating the first subset of topological skeleton layer data and parameterized bounding layer data is calculated using the following formula: Tb = E × te + K × tp; Adjust the value of K so that Tb ≤ Tmax; Based on the final determined K value, the task of generating the parameterized bounding layer is planned to be executed in parallel with the task of generating the topology skeleton layer in a subset order, with the generation and transmission of each subset performed independently.

6. The spatial grid data visualization method based on a three-dimensional model according to claim 1, characterized in that: S4 further includes the following: The generated topological skeleton layer corresponding to the graphic drawing data is submitted to the graphics processing unit; the graphics processing unit calculates the projection coordinates of the graphic markers and line segments in the screen space through coordinate transformation based on the spatial position of the graphic markers and line segments, and generates the first set of pixel color and depth data according to the preset vertex shading rules and fragment shading rules to complete the first round of rendering calculation. The generated parameterized bounding layer corresponding to the geometric primitives and rendering parameters is submitted to the graphics processing unit; the graphics processing unit performs the same coordinate transformation based on the vertex coordinate data of the geometric primitives to obtain their screen space projection coordinates, and generates a second set of pixel color and depth data according to the color and transparency defined by the corresponding rendering parameters through the fragment shading rules to complete the second round of rendering calculation; The graphics processing unit synthesizes the first and second sets of pixel color and depth data according to the rules of depth testing and color mixing, and outputs a synthesized image containing all component topology nodes, connecting edges and parameterized bounding volumes, which is then displayed as the initial 3D image with complete component semantics.

7. The spatial grid data visualization method based on a three-dimensional model according to claim 1, characterized in that: S5 further includes the following: While displaying the initial 3D image, the data generation of the geometric data layer is initiated. The data generation is performed on each component, and based on its component identifier, the corresponding set of facets is converted into geometric mesh data that can be directly drawn by the graphics processing unit. For each component, based on the axial dimension of its parametric bounding volume, an approximate spatial volume of the component is obtained by calculating the product of the components of the axial dimension in three dimensions; all components are then sorted in descending order based on the approximate spatial volume to generate a geometric replacement priority sequence. According to the geometric replacement priority sequence, the geometric mesh data of the generated components is submitted to the graphics processing unit; The graphics processing unit adjusts the display opacity of both the bounding volume and the geometric mesh data of the same component by linear interpolation in consecutive image frames, based on a preset total number of replacement frames. The linear interpolation adjustment gradually decreases the display opacity of the bounding volume from its initial value to complete transparency, while simultaneously increasing the display opacity of the corresponding geometric mesh data from complete transparency back to its initial value. Ultimately, the geometric mesh data visually completely replaces its corresponding bounding volume, achieving a gradual enhancement of visual details.

8. A spatial grid data visualization system based on a three-dimensional model, applied to the spatial grid data visualization method based on a three-dimensional model as described in any one of claims 1-7, characterized in that: It includes a data parsing and preprocessing module, a hierarchical data construction module, a temporal control and parallel task generation module, an initial rendering and image compositing module, and a progressive geometry blending and display module; The data parsing and preprocessing module is used to acquire raw grid data, divide the model into multiple independent components, calculate the spatial extent boundaries of each component, and determine the adjacency relationship between components. Based on the output of the data parsing and preprocessing module, the hierarchical data construction module logically constructs and generates independent topological skeleton layer data, parameterized bounding layer data, and geometric data layer data. The timing control and parallel task generation module performs logical analysis and calculation based on the data scale and system performance parameters of the topology skeleton layer and parameterized enclosing layer, generates a parallel task scheduling instruction that satisfies the display time constraint of the first frame, and drives the generation task of the topology skeleton layer data and parameterized enclosing layer data. The initial rendering and image synthesis module is used to receive the data output by the timing control and parallel task generation module, schedule the graphics processing unit to execute the rendering calculation of the topological skeleton and parameterized bounding volume in sequence, and fuse the calculation results into an initial three-dimensional image with complete component semantics. The progressive geometry blending and display module is used to schedule the generation of geometric data layer data during the initial 3D image display, and control the graphics processing unit to visually blend and replace the parameterized bounding volume and geometric mesh data according to the priority sequence, so as to complete the progressive enhancement of visualization details.

9. A spatial grid data visualization system based on a three-dimensional model, characterized in that: The initial rendering and image compositing module includes an image processing unit: In the initial rendering and image compositing module, the graphics processing unit receives the graphics data of the topological skeleton layer and the parameterized bounding layer, performs the transformation projection of vertex coordinates from the three-dimensional model space to the two-dimensional screen space, and generates the corresponding pixel color and depth information according to the preset shading rules. In the progressive geometry blending and display module, the graphics processing unit performs linear interpolation calculation and real-time rendering of the display opacity of the parameterized bounding volume and the subsequently delivered geometric mesh at the same spatial location based on control parameters.

Images