Natural reserve data fusion display method based on three-dimensional digital twinborn model

Abstract

The invention relates to the technical field of image data fusion processing, in particular to a natural reserve data fusion display method based on a three-dimensional digital twinborn model. The method comprises the steps of collecting multi-source heterogeneous data of a natural reserve, and performing intelligent preprocessing to obtain preprocessed data; performing preliminary feature extraction and dimension reduction processing on the preprocessed data to obtain feature data after dimension reduction; introducing a multi-dimensional data hierarchical adaptive fusion algorithm, and performing fusion processing on the feature data after dimension reduction to obtain fused data; and processing the fused data through an incremental updating and spatial adaptive rendering algorithm, embedding the data into a three-dimensional digital twinborn model for display, and updating the display data. The technical problems of inaccurate processing of multi-source heterogeneous data in a natural reserve, low display efficiency and poor self-adaption are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of image data fusion processing, and in particular to a method for fusion display of nature reserve data based on a three-dimensional digital twin model. Background Art

[0002] With advances in science and technology, particularly the development of remote sensing technology, sensor networks, and geographic information systems (GIS), more and more nature reserves are beginning to rely on multi-source, heterogeneous data for environmental monitoring and management. This data includes imagery from satellite remote sensing, environmental data collected by ground-based sensors, real-time weather data from meteorological stations, and species distribution data obtained from other monitoring equipment. These data sources vary in format, spatial reference system, time synchronization, and data accuracy, resulting in numerous challenges when directly applying them for analysis and presentation. For example, inconsistent data formats make it difficult to effectively integrate different data sources, while differences in spatial reference systems lead to inconsistent data positions on maps, and different timestamps make data synchronization a challenge. These issues lead to accuracy and real-time issues in digital twin models built based on this multi-source data, making it difficult to accurately reflect dynamic changes within the protected area.

[0003] Therefore, the above technology has technical problems such as inaccurate processing of multi-source heterogeneous data in nature reserves, low display efficiency, and poor adaptability. Summary of the Invention

[0004] The present invention provides a nature reserve data fusion display method based on a three-dimensional digital twin model to solve the technical problems of inaccurate processing of multi-source heterogeneous data, low display efficiency, and poor adaptability in nature reserves.

[0005] The present invention's method for fusion display of nature reserve data based on a three-dimensional digital twin model specifically includes the following technical solutions: The data fusion display method of nature reserve based on 3D digital twin model includes the following steps: S1. Collect multi-source heterogeneous data from nature reserves and perform intelligent preprocessing to obtain preprocessed data; perform preliminary feature extraction and dimensionality reduction on the preprocessed data to obtain reduced-dimensionality feature data; introduce a multidimensional data hierarchical adaptive fusion algorithm to fuse the reduced-dimensionality feature data to obtain fused data; S2. Through incremental updates and spatially adaptive rendering algorithms, the fused data is processed and embedded into the 3D digital twin model for display, and the displayed data is updated.

[0006] Preferably, the S1 specifically includes: The multi-dimensional data hierarchical adaptive fusion algorithm fuses the feature data after dimensionality reduction by introducing multi-dimensional data hierarchical division and adaptive weighting.

[0007] Preferably, the S1 specifically includes: In the process of implementing the hierarchical adaptive fusion algorithm for multidimensional data, a clustering method based on spatiotemporal distance measurement is introduced based on the feature data after dimensionality reduction of multi-source heterogeneous data to calculate the spatiotemporal distance measurement; based on the spatiotemporal distance measurement, the data source is hierarchically divided through the clustering algorithm.

[0008] Preferably, the S1 specifically includes: The spatiotemporal distance metric is calculated by combining time, spatial position and eigenvalues ​​of feature data after dimensionality reduction to quantify the similarity between different data sources.

[0009] Preferably, the S1 specifically includes: In the implementation process of the hierarchical adaptive fusion algorithm for multidimensional data, within each layer, the degree of mutual influence between data sources in each layer is calculated based on the spatiotemporal distance metric, combined with the feature space distance, and the introduction of exponential decay.

[0010] Preferably, the S1 specifically includes: In the implementation process of the hierarchical adaptive fusion algorithm for multidimensional data, the attention mechanism is introduced to define the initial weights for different data sources in each layer. The initial weights are dynamically adjusted based on the degree of mutual influence between the data sources in each layer to obtain the dynamically adjusted weights.

[0011] Preferably, the S1 specifically includes: Based on the dynamically adjusted weights, the feature data after dimensionality reduction of the data source at each level are weighted and fused to obtain the feature data after weighted fusion processing at each level. Finally, the feature data after weighted fusion processing at all levels are merged to obtain the fused data.

[0012] Preferably, the S2 specifically includes: During the implementation of the incremental update and spatial adaptive rendering algorithm, the fused data is compared with the data after the previous fusion to obtain incremental data; the incremental data is combined with the three-dimensional digital twin model to perform local rendering updates.

[0013] Preferably, the S2 specifically includes: In the process of implementing incremental updates and spatially adaptive rendering algorithms, a real-time data update mechanism is introduced to update the displayed data.

[0014] Preferably, the S2 specifically includes: The real-time data update mechanism fuses new multi-source heterogeneous data through a multi-dimensional data hierarchical adaptive fusion algorithm to obtain fused data; then performs predictive analysis on the fused data to obtain predicted fused data; performs weighted fusion processing on the fused data and the predicted fused data to obtain the final fused data; finally, through incremental updating and spatial adaptive rendering algorithms, the final fused data is embedded in the three-dimensional digital twin model for rendering display.

[0015] The beneficial effects of the technical solution of the present invention are: 1. Through the multi-dimensional data hierarchical adaptive fusion algorithm, it can effectively solve the fusion problem of heterogeneous data from different data sources (such as remote sensing images, satellite data, camera data, drone dynamic video data, GIS data, environmental monitoring sensor data, etc.). It can reasonably divide the levels according to the temporal and spatial characteristics of the data sources and through clustering methods (such as K-means or hierarchical clustering), thereby realizing the accurate weighted fusion of multi-source heterogeneous data.

[0016] 2. By combining incremental updates with a spatially adaptive rendering algorithm, data changes in the 3D digital twin model can be quickly displayed and updated locally, rather than re-rendering the entire 3D digital twin model. By updating only the incremental data (i.e., the data that has changed), unnecessary computation and rendering burdens are effectively reduced, significantly improving rendering efficiency.

[0017] 3. Introducing a real-time data update mechanism, combined with a time series prediction algorithm, can ensure the timeliness and accuracy of data by supplementing missing data through prediction when the data source is delayed or unavailable. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of the nature reserve data fusion display method based on the three-dimensional digital twin model described in the present invention. DETAILED DESCRIPTION

[0019] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] The specific scheme of the nature reserve data fusion display method based on the three-dimensional digital twin model provided by the present invention is described in detail below with reference to the accompanying drawings.

[0022] Refer to the attached Figure 1 , which shows a flow chart of a method for fusion display of nature reserve data based on a three-dimensional digital twin model provided by one embodiment of the present invention, the method comprising the following steps: S1. Collect multi-source heterogeneous data from nature reserves and perform intelligent preprocessing to obtain preprocessed data; perform preliminary feature extraction and dimensionality reduction on the preprocessed data to obtain reduced-dimensionality feature data; introduce a multidimensional data hierarchical adaptive fusion algorithm to fuse the reduced-dimensionality feature data to obtain fused data; Collect multi-source heterogeneous data from nature reserves, including remote sensing image data, satellite data, camera data, drone dynamic video data, geographic information system (GIS) data, environmental monitoring sensor data, etc.; in order to avoid multi-source heterogeneous data having different formats, spatial reference systems and timestamps, the multi-source heterogeneous data are intelligently preprocessed to obtain preprocessed data; the steps of the intelligent preprocessing include format conversion, time synchronization, spatial reference unification, standardization and normalization, etc., to ensure the quality and consistency of multi-source heterogeneous data; the time synchronization refers to time synchronization processing through time alignment algorithms (such as Kalman filtering, interpolation algorithms, etc.) to ensure that all data are analyzed and displayed within the same time frame; the spatial reference unification refers to converting multi-source heterogeneous data into a unified spatial reference system, such as the WGS84 coordinate system, through spatial reprojection technology, and optimizing spatial accuracy; the intelligent preprocessing process adopts technical means well known to those skilled in the art and will not be described in detail here; In particular, for the UAV dynamic video data, the UAV dynamic video data is first encoded using existing encoding technology, and then decoded using existing decoders to convert the UAV dynamic video data into a standard image format before the above preprocessing process is performed; Specifically, the intelligent preprocessing process for data from different sources will be different. For example, for remote sensing image data, the intelligent preprocessing process includes denoising, radiation correction, geometric correction, response alignment, standardization and normalization, etc.; for satellite data, the intelligent preprocessing process includes time synchronization, data format conversion, noise filtering, outlier detection, standardization and normalization, etc.; for geographic information system data, the intelligent preprocessing process includes coordinate system conversion, data accuracy correction, combination of vector data and raster data, standardization and normalization, etc.; for environmental monitoring sensor data, the intelligent preprocessing process includes time series processing, data filling, standardization and normalization, etc. The specific process of intelligent processing is selected according to the specific scenario and will not be elaborated here; Further, the pre-processed data is subjected to preliminary feature extraction using existing feature engineering technology to obtain preliminary features; the preliminary features are subjected to dimensionality reduction processing using feature dimensionality reduction technology such as principal component analysis to obtain feature data after dimensionality reduction; the feature engineering technology and feature dimensionality reduction technology are technical means well known to those skilled in the art and will not be described in detail here; A multi-dimensional data hierarchical adaptive fusion algorithm is introduced to fuse the feature data after dimensionality reduction to obtain fused data. The multi-dimensional data hierarchical adaptive fusion algorithm can meet the data fusion requirements from different data sources by introducing multi-dimensional data hierarchical division, adaptive weighting and other mechanisms, and realize data fusion from multiple data sources. The specific implementation process is as follows: A clustering method based on spatiotemporal distance metric is introduced. Based on the feature data of multi-source heterogeneous data after dimensionality reduction, the spatiotemporal distance metric, i.e., the distance between each data source, is calculated. Based on the spatiotemporal distance metric, the data sources are hierarchically divided through the clustering algorithm. The spatiotemporal distance metric is expressed as , combining the influence of time, space and feature data, and calculating based on Euclidean distance to quantify the similarity between different data sources, specifically defined as:

[0023] in, and Respectively represent data sources and The feature data of each data source after dimensionality reduction; and Respectively represent data sources and The first dimension of the feature data after dimensionality reduction of the data source eigenvalues; and Respectively represent data sources and The timestamp of data collection from each data source indicates the collection time of each data point and is obtained through the built-in clock of the data collection device; and Respectively represent data sources and The spatial position corresponding to the feature data after dimensionality reduction of a data source, wherein the spatial position comes from the preprocessed data; is the feature weight coefficient, indicating the The weighted coefficient of each feature value is used to indicate the contribution of each feature to the overall similarity calculation. It is set by expert experience combined with application requirements. The reference value range is ; It is the weighted coefficient of time difference, which reflects the influence of time difference on similarity calculation. It is set according to expert experience method, and the reference value range is ; It is the weighted coefficient of spatial difference, which indicates the influence of spatial position difference on similarity calculation. It is set according to expert experience, and the reference value range is ; is the number of dimensions of the feature data after dimensionality reduction of the data source; to eliminate the influence of dimension, the above timestamps and spatial positions are normalized and will not be described in detail here; Based on the spatiotemporal distance metric, different data sources are divided into levels to ensure that the feature data of the data sources in each level have high similarity after dimensionality reduction. The K-means clustering algorithm and the hierarchical clustering algorithm are well known to those skilled in the art and will not be described in detail here.

[0024] Furthermore, after the hierarchical division is completed, based on the theory of similarity measurement and mutual influence modeling, the degree of mutual influence between different data sources in each level is calculated. The specific calculation formula is as follows:

[0025] in, It is in The first level data sources and The degree of mutual influence among data sources reflects the similarity between the feature data after dimensionality reduction of different data sources. The greater the degree of mutual influence, the greater the contribution to the final fusion. and Respectively represent The first level data sources and The feature data of each data source after dimensionality reduction; It is in The first level data sources and The spatiotemporal distance measurement of the feature data after dimensionality reduction of the data source; It is the weighted coefficient of the time-space distance, which is used to balance the influence of the time-space distance and the similarity of the feature space. It is determined according to the expert experience method, and the reference value range is ; It is the weighted coefficient of feature space similarity, which is used to control the weighted degree of feature space similarity to similarity measurement. It is determined according to expert experience and the reference value range is ; It is in The first level data sources and The feature space distance of the feature data after dimensionality reduction of the data sources is used to measure the feature similarity of different data sources; It is a comprehensive measure of the differences in time, space and feature space, reflecting the combined impact of the differences in time, space and feature space; By combining spatiotemporal distance and feature space distance, we ensure that similarity depends not only on spatial and temporal differences but also on the characteristic information of the data source. Finally, we introduce exponential decay to effectively quantify the mutual influence between the reduced-dimensional feature data of different data sources, so that the similarity decreases rapidly as the difference increases, thus meeting the needs of multi-source heterogeneous data fusion. The attention mechanism is introduced to define initial weights for different data sources in each layer. The initial weights are dynamically adjusted based on the degree of mutual influence between different data sources in each layer to obtain the dynamically adjusted weights. The specific formula is:

[0026] in, Indicates in The first level Dynamically adjusted weights of the feature data after dimensionality reduction for each data source; Indicates in The first level The initial weights of the feature data after dimensionality reduction of each data source are obtained based on the attention mechanism. The attention mechanism is a technical means well known to those skilled in the art and will not be described in detail here; It is The total number of different data sources in the hierarchy; It is a calculation of the mutual influence degree of data sources within the hierarchy, indicating the The weighted influence of a data source within the current level is obtained based on the degree of mutual influence with all other data sources; It is The initial weight of the data source is the same as the The product of the weighted influence of the data sources; It is the sum of the mutual influence between all data sources at all levels, which represents the sum of the mutual influence between all data sources and is used to normalize the weight; Furthermore, based on the dynamically adjusted weights, the feature data of the data source at each level after dimensionality reduction is weighted and fused to obtain the feature data after weighted fusion processing at each level. Finally, the feature data after weighted fusion processing at all levels are merged to obtain the fused data. .

[0027] S2. Through incremental updates and spatially adaptive rendering algorithms, the fused data is processed and embedded into the 3D digital twin model for display, and the displayed data is updated.

[0028] The fused data is processed using an incremental update and spatially adaptive rendering algorithm, and then embedded into a 3D digital twin model for display, reflecting changes in the fused data in real time. The 3D digital twin model is a well-known technical approach for those skilled in the art and will not be described in detail here. The incremental update and spatially adaptive rendering algorithm can render only the changed parts when the fused data changes, avoiding global redrawing and significantly improving rendering efficiency. The specific implementation process is as follows: The fused data With the previous frame data status (i.e. the last fused data) to get the incremental data Input the incremental data into an existing 3D rendering engine (such as Unreal Engine UE, UE5 integrates cutting-edge technologies such as Lumen and Nanite, providing strong support for the development of digital twin bases), map the incremental data through existing spatial indexing and coordinate mapping technologies, and combine the incremental data with the mesh or voxel data of the 3D digital twin model to perform local rendering updates; in the above process, the 3D rendering engine does not need to re-render the entire 3D digital twin model, but only updates the changed parts, thereby reducing the amount of calculation and improving rendering efficiency; At the same time, in order to finely control local rendering, the 3D rendering engine also uses Level of Detail (LOD) technology. During the rendering process, parts farther away from the user will use lower rendering details, while parts closer to the user will use higher rendering details, thereby further reducing the rendering burden while ensuring visual effects. The specific implementation formula is as follows:

[0029] in, It is the incremental rendering result, which represents the output result of incremental rendering, that is, the changed part after being processed by the rendering engine; It is a rendering function based on modern computer graphics technology and algorithms, such as illumination model, projection matrix transformation and 3D modeling rendering technology. The illumination model is such as Phong model, Blinn-Phong model, etc. The 3D modeling rendering technology is such as OpenGL, Vulkan, etc. It describes how to use incremental data Update the rendering result, such as ,in, Represents texture mapping and detail level control. Texture mapping is the process of mapping an image (texture) onto the surface of a 3D digital twin model. Represents a texture mapping operation; This function optimizes the details of incremental data, improving rendering performance by adjusting the level of detail of near and far areas. This is achieved based on LOD (Level of Detail) technology. It is the lighting calculation, which affects the brightness and shadow effects of the surface of the 3D data twin model; It is a geometric transformation used to transform the vertex coordinates of the 3D digital twin model (such as rotation, scaling, translation, etc.) so that the updated geometric shape is consistent with the 3D digital twin model, so as to accurately reflect the position changes in space. The methods used in the rendering function are all technical means well known to those skilled in the art and will not be described in detail here. Furthermore, a real-time data update mechanism is introduced to display data updates in real time to avoid the problem of inaccurate data caused by continuous changes over time in nature reserve scenarios. The real-time data update mechanism combines the synergy of incremental updates and time series prediction synchronization to update the data in the 3D digital twin model in real time. The specific process is as follows: After the new multi-source heterogeneous data is fused through the multi-dimensional data hierarchical adaptive fusion algorithm, the fused data is obtained; then the fused data is predicted and analyzed through the existing time series prediction algorithm (such as the prediction algorithm based on exponential smoothing) to obtain the predicted fused data; the fused data and the predicted fused data are further weightedly fused to obtain the final fused data; finally, the incremental update and spatial adaptive rendering algorithm are used to embed the final fused data into the three-dimensional digital twin model, and it is displayed through GPU rendering technology (such as OpenGL or Vulkan) to realize the data update of the three-dimensional digital twin model; the time series prediction method is a technical means well known to those skilled in the art and will not be elaborated here.

[0030] In summary, the data fusion display method of nature reserve based on three-dimensional digital twin model was completed.

[0031] The order in which the embodiments of the invention are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0032] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0033] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A data fusion display method for nature reserves based on a three-dimensional digital twin model, characterized in that: The following steps are involved: S1. Collect multi-source heterogeneous data from nature reserves and perform intelligent preprocessing to obtain preprocessed data; Perform preliminary feature extraction and dimensionality reduction on the preprocessed data to obtain feature data after dimensionality reduction; introduce a multi-dimensional data hierarchical adaptive fusion algorithm to fuse the feature data after dimensionality reduction to obtain fused data; S2. Through incremental updates and spatially adaptive rendering algorithms, the fused data is processed and embedded into the 3D digital twin model for display, and the displayed data is updated.

2. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 1 is characterized in that: Said S1 specifically includes: The multi-dimensional data hierarchical adaptive fusion algorithm fuses the feature data after dimensionality reduction by introducing multi-dimensional data hierarchical division and adaptive weighting.

3. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 2 is characterized in that: Said S1 specifically includes: In the process of implementing the hierarchical adaptive fusion algorithm for multidimensional data, a clustering method based on spatiotemporal distance measurement is introduced based on the feature data after dimensionality reduction of multi-source heterogeneous data to calculate the spatiotemporal distance measurement; based on the spatiotemporal distance measurement, the data source is hierarchically divided through the clustering algorithm.

4. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 3 is characterized in that: Said S1 specifically includes: The spatiotemporal distance metric is calculated by combining time, spatial position and eigenvalues ​​of feature data after dimensionality reduction to quantify the similarity between different data sources.

5. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 4 is characterized in that: Said S1 specifically includes: In the implementation process of the hierarchical adaptive fusion algorithm for multidimensional data, within each layer, the degree of mutual influence between data sources in each layer is calculated based on the spatiotemporal distance metric, combined with the feature space distance, and the introduction of exponential decay.

6. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 5 is characterized in that: Said S1 specifically includes: In the implementation process of the hierarchical adaptive fusion algorithm for multidimensional data, the attention mechanism is introduced to define the initial weights for different data sources in each layer. The initial weights are dynamically adjusted based on the degree of mutual influence between the data sources in each layer to obtain the dynamically adjusted weights.

7. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 6 is characterized in that: Said S1 specifically includes: Based on the dynamically adjusted weights, the feature data after dimensionality reduction of the data source at each level are weighted and fused to obtain the feature data after weighted fusion processing at each level. Finally, the feature data after weighted fusion processing at all levels are merged to obtain the fused data.

8. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 1 is characterized in that: Said S2 specifically includes: During the implementation of the incremental update and spatial adaptive rendering algorithm, the fused data is compared with the data after the previous fusion to obtain incremental data; the incremental data is combined with the three-dimensional digital twin model to perform local rendering updates.

9. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 8 is characterized in that: Said S2 specifically includes: In the process of implementing incremental updates and spatially adaptive rendering algorithms, a real-time data update mechanism is introduced to update the displayed data.

10. The method for fusion display of nature reserve data based on a three-dimensional digital twin model according to claim 9 is characterized in that: Said S2 specifically includes: The real-time data update mechanism fuses new multi-source heterogeneous data through a multi-dimensional data hierarchical adaptive fusion algorithm to obtain fused data; then performs predictive analysis on the fused data to obtain predicted fused data; performs weighted fusion processing on the fused data and the predicted fused data to obtain the final fused data; finally, through incremental updating and spatial adaptive rendering algorithms, the final fused data is embedded in the three-dimensional digital twin model for rendering display.

Images