Virtual scene construction method driven by live-action three-dimensional data

Abstract

The invention relates to the technical field of virtual scene construction, and discloses a virtual scene construction method driven by live-action three-dimensional data. The method comprises the steps of obtaining multi-source three-dimensional data of a target area, extracting geometric feature parameters of a scene data unit, calculating topological feature indexes, generating a construction input set, loading a pre-configured three-dimensional modeling algorithm to determine dynamic evolution parameters, labeling the parameters to generate a three-dimensional scene map, and constructing a virtual scene after space consistency detection. And if the parameter conflict is detected, fusing the parameters according to the multi-source data priority. And after construction is completed, monitoring a spatial difference value between a scene data unit and an adjacent unit in real time, and correcting parameters to update the scene when the spatial difference value exceeds a threshold value. According to the method, multi-source data is fully utilized, the accuracy and efficiency of virtual scene construction are improved, the problems existing in a traditional method are effectively solved, and the method has wide application prospects in the fields of urban planning, cultural heritage protection and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual scene construction, and particularly to a method for constructing a virtual scene driven by real - scene three - dimensional data. Background Art

[0002] With the rapid development of computer technology, digital image processing technology, and three - dimensional modeling technology, virtual scene construction has shown great application potential in many fields, such as urban planning, cultural heritage protection, game development, film and television production, and virtual reality (VR) and augmented reality (AR) and other fields. However, the current virtual scene construction technology still faces many challenges.

[0003] Most traditional virtual scene construction methods rely on manual modeling, which requires a large amount of human, material, and time costs. Modelers need to create various objects in the scene one by one. From the shape design of the object to the texture painting and then to the planning of the scene layout, all need to be carefully crafted manually. When constructing large - scale and complex scenes, the efficiency of this manual modeling is extremely low, and it is difficult to ensure the authenticity and accuracy of the scene. For example, when constructing a virtual scene of a large city, it is necessary to accurately model every building, every street, every green area, etc. Manual operation not only has a huge workload but also is prone to omissions or errors, resulting in a deviation between the virtual scene and the actual scene.

[0004] Although some virtual scene construction technologies use three - dimensional scanning technology to obtain data, there are limitations in the data processing and scene construction processes. On the one hand, the data obtained from a single data source is often not comprehensive enough. For example, the point cloud data obtained only by lidar can accurately measure the spatial position information of objects, but it is insufficient in obtaining the texture and semantic information of objects; while only using oblique photography image data can obtain rich texture information, but there are certain errors in the accurate measurement of spatial coordinates. This makes it difficult to achieve an ideal effect in the fusion of geometric accuracy and texture details of the constructed virtual scene. On the other hand, when processing multi - source data, there are differences in data formats, accuracies, resolutions, etc. of different data sources. How to efficiently fuse these data and extract useful information has become a key problem restricting the quality of virtual scene construction.

[0005] In the process of virtual scene construction, there are also deficiencies in the dynamic evolution simulation of objects in the scene. Existing technologies often have difficulty accurately simulating the dynamic changes of objects in the scene under different conditions, and cannot meet some application scenarios with high requirements for scene real - time and interactivity. For example, when simulating an urban traffic scene, dynamic information such as the driving trajectories of vehicles, the movement paths of pedestrians, and the changes of buildings over time are difficult to accurately simulate, resulting in a lack of realism and immersion in the virtual scene.

[0006] In addition, existing virtual scene construction technologies also need to be strengthened in terms of scene consistency detection and optimization. Due to the error accumulation in the data acquisition and processing process, problems such as object position conflicts and unnatural texture splicing may exist in the constructed virtual scene, affecting the quality of the virtual scene and the user experience. At the same time, after the scene construction is completed, there is a lack of effective dynamic monitoring and real-time correction mechanisms, and it is impossible to respond in a timely manner to changes that may occur in the scene, such as the demolition and new construction of buildings, and environmental changes. Summary of the Invention

[0007] The purpose of the present invention is to provide a method for constructing a virtual scene driven by real-scene three-dimensional data to solve the problems raised in the above background technology.

[0008] To achieve the above purpose, the present invention provides the following technical solution: A method for constructing a virtual scene driven by real-scene three-dimensional data, the method includes: Obtain multi-source three-dimensional data of the target area, and extract the geometric feature parameters of each scene data unit from a preset three-dimensional modeling engine, where the geometric feature parameters include the spatial coordinate attributes, texture attributes, and semantic attributes of the scene data unit; For each scene data unit, calculate the topological feature index of the scene data unit based on the geometric feature parameters respectively, and generate a construction input set for the scene data unit according to multiple topological feature indexes; where the topological feature index is a comprehensive quantization parameter that fuses the spatial coordinate attributes and texture attributes for each scene data unit; the construction input set is a standardized data set composed of multiple selected topological feature indexes; For each scene data unit, load a pre-configured three-dimensional modeling algorithm, and determine the dynamic evolution parameters of the scene data unit during the construction process based on the spatial relationship between the multiple feature indexes included in the construction input set and the processing rules of the three-dimensional modeling algorithm; Annotate the dynamic evolution parameters to each scene data unit in the three-dimensional modeling engine to generate a three-dimensional scene atlas; For each scene data unit, perform spatial consistency detection according to the dynamic evolution parameters to obtain a detection result; Based on the detection result, construct a virtual scene to generate a virtual scene construction result.

[0009] Preferably, constructing a virtual scene based on the detection result to generate a virtual scene construction result includes: when the detection result indicates that there are conflicts among multiple dynamic evolution parameters included in the scene data unit, determining the scene data unit as an abnormal unit; obtaining the conflicting parameters of the abnormal unit as the correction object, and determining the multi-source data priorities corresponding to the conflicting parameters; and performing parameter fusion on the scene data unit based on the data priorities to generate a virtual scene construction result.

[0010] Preferably, the multi-source three-dimensional data includes lidar point cloud data and oblique photography image data. The data priority of the lidar point cloud data is the first priority, and the data priority of the oblique photography image data is the second priority. The performing parameter fusion on the scene data unit based on the data priorities to generate a virtual scene construction result includes: Adjusting the construction input set corresponding to the oblique photography image data according to the order of the first priority and the second priority; Based on the three-dimensional scene map, extracting the dynamic evolution parameters of the oblique photography image data in the adjusted construction input set; Combining the dynamic evolution parameters and the three-dimensional scene map for parameter fusion to generate a virtual scene construction result.

[0011] Preferably, for each scene data unit, calculating the topological feature index of the scene data unit based on the geometric feature parameters respectively, and generating the construction input set according to multiple topological feature indexes, includes: For each scene data unit, determining multiple adjacent units adjacent to it in space; Calculating the correlation degree between the geometric feature parameters of each adjacent unit and the texture attribute, and determining the first feature unit with the highest correlation degree from the adjacent units based on multiple correlation degrees; Based on the first feature unit, determining multiple extended units adjacent to the feature unit, calculating the correlation degree between the geometric feature parameters of each extended unit and the texture attribute, and determining the next feature unit with the highest correlation degree from the extended units based on multiple correlation degrees; Repeatedly executing the steps of determining multiple extended units adjacent to the feature unit, calculating the correlation degree of each extended unit, and determining the next feature unit with the highest correlation degree until all scene data units in the target area are traversed to obtain a set of topological feature indexes of the scene data unit; Generating the construction input set based on the set of topological feature indexes.

[0012] Preferably, the calculating the correlation degree between the geometric feature parameters of each adjacent unit and the texture attribute includes: Calculate the first correlation degree between the spatial coordinate attributes and the texture attributes of each of the adjacent units, and the second correlation degree between the spatial coordinate attributes and the semantic attributes of each of the adjacent units; For each of the adjacent units, use the weighted sum of the first correlation degree and the second correlation degree as the correlation degree of the adjacent unit.

[0013] Preferably, the determining the first feature unit with the highest correlation degree from the adjacent units based on the multiple correlation degrees includes: Store the multiple correlation degrees corresponding to the multiple adjacent units in a feature candidate list, and perform a descending order sorting on the correlation degrees in the feature candidate list to obtain a sorting result; Based on the sorting result, determine the adjacent unit corresponding to the highest correlation degree as the first feature unit.

[0014] Preferably, after generating the virtual scene construction result based on the detection result, it further includes: When the virtual scene construction result is generated, for each scene data unit, monitor the spatial difference value between the scene data unit and the adjacent unit in real time; When there is a spatial difference value exceeding a preset threshold, use the adjacent unit as an abnormal reference unit, and perform dynamic parameter correction on the scene data unit based on the abnormal reference unit to generate an updated virtual scene construction result.

[0015] Preferably, the training process of the pre-configured three-dimensional modeling algorithm includes: Obtain a historical three-dimensional data set, and extract the mapping relationship between the geometric feature parameters and the dynamic evolution parameters in the data set; Construct an initial three-dimensional modeling algorithm according to the mapping relationship, and use the grid division method to perform iterative optimization on the algorithm; When the error between the predicted parameters and the measured parameters output by the algorithm is lower than a preset threshold, determine that the algorithm is configured.

[0016] Preferably, the three-dimensional modeling engine further includes a three-dimensional modeling layer of spatial grid division, texture mapping rules and lighting models. When generating the three-dimensional scene atlas, superimpose the dynamic evolution parameters on the three-dimensional modeling layer for visual expression.

[0017] Preferably, the generation method of the three-dimensional modeling layer includes: Collect the terrain point cloud data, building surface data and environmental lighting data of the target area; Perform grid processing on the terrain point cloud data to generate a spatial grid division layer, perform texture mapping on the building surface data to generate a texture mapping rule layer, and simulate the environmental light data to generate a lighting model layer; Perform spatial overlay analysis on the spatial grid division layer, the texture mapping rule layer, and the lighting model layer to form the three-dimensional modeling layer.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: In terms of data utilization, the method obtains multi-source three-dimensional data of the target area, such as lidar point cloud data and oblique photography image data, etc., and gives full play to the advantages of different data sources. Lidar point cloud data can accurately provide spatial coordinate information, while oblique photography image data is rich in texture information. By extracting the geometric feature parameters of each scene data unit from the preset three-dimensional modeling engine, covering spatial coordinate attributes, texture attributes, and semantic attributes, in-depth mining and integration of multi-source data are realized. This is compared with the traditional single-data-source modeling method, which greatly enriches the basic data for scene construction, making the constructed virtual scene more real and accurate in terms of geometric accuracy and texture details. For example, when constructing a virtual scene of a historical and cultural block, not only can the external dimensions and spatial positions of the buildings be accurately restored, but also the traces of years and decorative details on the building surface can be shown through texture attributes, providing strong support for the digital protection and display of cultural heritage.

[0019] The process of calculating topological feature indicators and generating the construction input set further optimizes the data processing method. For each scene data unit, the topological feature indicators are calculated by fusing its spatial coordinate attributes and texture attributes. This comprehensive quantization parameter can better reflect the internal relationship between scene data units. By determining adjacent units, calculating the association degree, and screening feature units, the construction input set generated by the obtained topological feature indicator set is a standardized data set, providing a high-quality data basis for subsequent modeling. In the construction of virtual scenes in urban areas, this method can accurately capture the spatial relationships between buildings, between roads and buildings, making the scene layout more reasonable and conforming to the actual geographical space logic.

[0020] In the process of virtual scene construction, the pre-configured 3D modeling algorithm plays an important role. By obtaining historical 3D data sets, this algorithm extracts the mapping relationship between geometric feature parameters and dynamic evolution parameters, and uses the grid division method for iterative optimization. When the error between the predicted parameters and the measured parameters output by the algorithm is lower than the preset threshold, it is determined that the configuration is completed. This enables the algorithm to accurately determine the dynamic evolution parameters of scene data units during the construction process based on the spatial relationship between feature indicators in the construction input set. Taking the example of simulating the change scenario of urban traffic flow, the algorithm can accurately predict the driving trajectory and speed change of vehicles (dynamic evolution parameters) based on the traffic flow data of roads in different time periods (as part of the geometric feature parameters), realizing the dynamic simulation of traffic scenarios and improving the authenticity and interactivity of virtual scenes.

[0021] Generating a 3D scene atlas and performing spatial consistency detection effectively guarantees the quality of virtual scenes. In the 3D modeling engine, dynamic evolution parameters are labeled for each scene data unit to generate a 3D scene atlas, and the conflict problem of dynamic evolution parameters contained in scene data units is promptly discovered through spatial consistency detection. When a conflict is detected, parameter fusion is performed on the scene data units according to the priority of multi-source data. For example, when dealing with the conflict between lidar point cloud data and oblique photography image data, the construction input set is adjusted according to the set priority order, and dynamic evolution parameters are extracted and fused to ensure the consistency of information such as the position and texture of objects in the virtual scene, avoiding problems such as model overlap and texture disorder, and enhancing the visual effect and usability of virtual scenes.

[0022] In addition, after the virtual scene construction result is generated, the spatial difference value between the scene data unit and adjacent units is monitored in real time. When the difference value exceeds the preset threshold, dynamic parameter correction is performed on the scene data unit. This mechanism enables the virtual scene to adapt to the dynamic changes of the environment, such as the new construction or demolition of buildings in urban construction, and updates the virtual scene content in a timely manner to maintain the accuracy and timeliness of the scene. Whether in the field of urban planning, providing real-time updated virtual scenes for planners to assist in decision-making; or in game development, creating a more immersive and realistic game environment for players, this method has important application value. Brief Description of the Drawings

[0023] Figure 1 It is the working principle diagram of the virtual scene construction method driven by real scene 3D data described in the present invention; Figure 2 It is the working flow chart of constructing a virtual scene by processing oblique photography image data based on data priority; Figure 3 It is the working flow chart of calculating topological feature indicators and generating a construction input set; Figure 4Workflow diagram for calculating the correlation degree of adjacent units; Figure 5 Workflow diagram for generating a 3D modeling layer. Specific implementation manners

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0025] Please refer to Figures 1 - 5 , the present invention provides a method for constructing a virtual scene driven by real-scene 3D data, and the specific implementation steps are as follows: Obtain multi-source 3D data of the target area. These data come from a wide range of sources, such as point cloud data collected by lidar, image data obtained through oblique photography, etc. Extract the geometric feature parameters of each scene data unit from a preset 3D modeling engine, including the spatial coordinate attributes of the scene data unit, which are used to accurately determine its position in space; texture attributes, which can show the detailed features of the object surface; semantic attributes, which endow the data unit with specific meanings, such as distinguishing whether it is a building, a road or vegetation, etc.

[0026] For each scene data unit, calculate the topological feature index of the scene data unit respectively based on the obtained geometric feature parameters. The topological feature index is a comprehensive quantization parameter that integrates spatial coordinate attributes and texture attributes, and it can more comprehensively reflect the characteristics of the scene data unit. After calculating multiple topological feature indexes, select appropriate indexes from them to form a standardized data set, that is, construct an input set, which provides basic data for subsequent processing.

[0027] For each scene data unit, load a pre-configured 3D modeling algorithm. Based on the spatial relationship between the multiple feature indexes included in the constructed input set and the processing rules of the 3D modeling algorithm, determine the dynamic evolution parameters of the scene data unit during the construction process. These parameters determine the change situation of the scene data unit during the virtual scene construction process.

[0028] Mark the determined dynamic evolution parameters for each scene data unit in the 3D modeling engine, thereby generating a 3D scene atlas. This atlas integrates the information of each scene data unit, providing an intuitive visual expression and data support for virtual scene construction.

[0029] For each scene data unit, spatial consistency detection is performed according to the marked dynamic evolution parameters to obtain the detection result. This detection mainly determines whether the dynamic evolution parameters of the scene data unit are spatially coordinated, and whether there are conflicts or unreasonable situations.

[0030] Based on the detection result, a virtual scene is constructed to generate the virtual scene construction result. If the detection result indicates that there are no conflicts in the dynamic evolution parameters of the scene data unit, these parameters are directly used to construct the virtual scene; if there are conflicts, further processing is required to ensure the accuracy and reasonableness of the virtual scene.

[0031] The technical solution of the present invention will be further described below through specific embodiments: Embodiment 1: When constructing a virtual scene, if the detection result indicates that there are conflicts in multiple dynamic evolution parameters included in a scene data unit, the scene data unit is determined as an abnormal unit. The conflict parameters of the abnormal unit are obtained as the correction object, and the priorities of multi-source data corresponding to the conflict parameters are determined. For example, if the multi-source 3D data includes lidar point cloud data and oblique photography image data, and the data priority of the lidar point cloud data is the first priority, and the data priority of the oblique photography image data is the second priority. The parameters of the scene data unit are fused according to the data priority to generate the virtual scene construction result. Specifically, first, according to the priority order, the construction input set corresponding to the oblique photography image data is adjusted to better match the lidar point cloud data; then, based on the 3D scene atlas, the dynamic evolution parameters of the oblique photography image data in the adjusted construction input set are extracted; finally, these dynamic evolution parameters are combined with the 3D scene atlas for parameter fusion, so as to obtain a more accurate virtual scene construction result. This can make full use of the advantages of different data sources and improve the quality of the virtual scene.

[0032] Suppose a virtual scene of a city center area is to be constructed, which includes various scene elements such as multiple buildings, streets, parks, etc. Lidar is used to collect point cloud data, and oblique photography is used to obtain image data as the multi-source 3D data for constructing the virtual scene.

[0033] During the construction process, it is detected that there are conflicts in multiple dynamic evolution parameters in a certain scene data unit (suppose it is the scene data unit corresponding to a building located at a street corner). The scene data unit corresponding to this building is determined as an abnormal unit, and its conflict parameters are obtained as the correction object. It is known that the data priority of the lidar point cloud data is the first priority, and the data priority of the oblique photography image data is the second priority.

[0034] Adjust the constructed input set corresponding to the oblique photography image data according to the priority order. For example, lidar point cloud data can more accurately reflect the outline and height information of buildings, while oblique photography image data has advantages in texture details. When adjusting, first ensure that the spatial coordinate information of the oblique photography image data is consistent with the lidar point cloud data. According to the approximate position and shape of the building determined by the lidar point cloud data, fine-tune the spatial position and angle of the part of the oblique photography image data corresponding to the building, so that it can be better integrated with the lidar point cloud data.

[0035] Based on the adjusted constructed input set, extract the dynamic evolution parameters of the oblique photography image data in the three-dimensional scene atlas. The three-dimensional scene atlas records various information of each scene data unit. Through a specific extraction algorithm, obtain the dynamic evolution parameters of the oblique photography image data under the current adjusted constructed input set, such as texture changes and lighting effects of the building.

[0036] Combine these dynamic evolution parameters with the three-dimensional scene atlas for parameter fusion. For example, combine the geometric structure of the building determined by the lidar point cloud data with the texture information of the oblique photography image data, so that the building has an accurate shape and realistic texture in the virtual scene. At the same time, considering the lighting effect, comprehensively calculate the lighting-related parameters from the two types of data to determine the final lighting parameters, making the lighting effect of the building in the virtual scene more natural. After these steps, the virtual scene construction result can be generated, and a more realistic and accurate street corner building model can be constructed, which together with the surrounding scene elements constitutes a complete virtual scene of the urban central area.

[0037] Example 2: When calculating the topological feature indicators for each scene data unit and generating the construction input set, first determine multiple adjacent units that are spatially adjacent to it. Then calculate the correlation degree between the geometric feature parameters and the texture attributes of each adjacent unit. The calculation of the correlation degree is carried out by calculating the first correlation degree between the spatial coordinate attributes and the texture attributes of each adjacent unit respectively, and the second correlation degree between the spatial coordinate attributes and the semantic attributes of each adjacent unit, and then taking the weighted sum of the two as the correlation degree of the adjacent unit. Based on multiple correlation degrees, determine the first feature unit with the highest correlation degree from the adjacent units. The specific method is to store the multiple correlation degrees corresponding to the multiple adjacent units in the feature candidate list, and sort the correlation degrees in the feature candidate list in descending order, and determine the adjacent unit corresponding to the highest correlation degree as the first feature unit. Based on the first feature unit, determine multiple extended units adjacent to this feature unit, calculate the correlation degree between the geometric feature parameters and the texture attributes of each extended unit, and determine the next feature unit with the highest correlation degree from the extended units based on multiple correlation degrees. Repeat the above steps until all scene data units in the target area are traversed, obtain the topological feature indicator set of the scene data units, and then generate the construction input set based on this set. This method can more comprehensively and accurately reflect the relationship between scene data units and provide more reliable data for subsequent virtual scene construction.

[0038] Suppose we want to construct a virtual scene of a large campus, which contains rich and diverse scene elements such as teaching buildings, libraries, playgrounds, greening areas, etc. These elements are all divided into individual scene data units.

[0039] When calculating the topological feature indicators for each scene data unit and generating the construction input set, take a certain scene data unit of one of the teaching buildings (assuming it is the scene data unit A corresponding to the entrance of the teaching building) as an example for illustration.

[0040] First, determine multiple spatially adjacent units of the scene data unit A. These adjacent units may include the corresponding scene data units of the stairs, flower beds, roads, etc. near the entrance of the teaching building. Next, calculate the correlation degree between the geometric feature parameters and the texture attributes of each adjacent unit. For example, for the adjacent unit B corresponding to the stairs, calculate the first correlation degree between its spatial coordinate attributes and texture attributes, and the second correlation degree between its spatial coordinate attributes and semantic attributes respectively. The calculation of the correlation degree between spatial coordinate attributes and texture attributes can be considered from the closeness between the two when describing the characteristics of the stairs. If the texture of the stairs shows a pattern related to the direction of the steps, and this pattern has a high consistency with the changes in the spatial coordinates of the stairs (such as height changes, position directions), then the first correlation degree is relatively high; while for the calculation of the second correlation degree between spatial coordinate attributes and semantic attributes, it depends on the connection between the spatial position of the stairs and its semantic concept of "stairs". If the position of the stairs conforms to the normal traffic logic in the campus and is highly relevant to the semantics of "used for teachers and students to go up and down the stairs", the second correlation degree will also be relatively high. Weighted sum these two correlation degrees to obtain the correlation degree of the adjacent unit B. Calculate the correlation degrees of other adjacent units (such as the units corresponding to flower beds and roads) in the same way.

[0041] Store the correlation degrees corresponding to all adjacent units into the feature candidate list, and then sort the correlation degrees in the list in descending order. Suppose after sorting, it is found that the correlation degree of the adjacent unit B corresponding to the stairs is the highest, then determine the adjacent unit B as the first feature unit.

[0042] Based on the first feature unit B, determine multiple extended units adjacent to it. These extended units may include the corresponding scene data units of the floors, corridors, etc. connected to the stairs. Then calculate the correlation degree between the geometric feature parameters and the texture attributes of each extended unit. For example, for the extended unit C corresponding to a certain floor connected to the stairs, calculate its correlation degree according to the previous method. After that, determine the next feature unit with the highest correlation degree from the extended units based on multiple correlation degrees. Suppose the correlation degree of unit C is the highest among these extended units, then determine unit C as the next feature unit.

[0043] Repeat the above steps of determining multiple extended units adjacent to the feature unit, calculating the correlation degree, and determining the next feature unit with the highest correlation degree. With continuous repetition, gradually traverse all the scene data units of the entire campus, so as to obtain the topological feature index set of the scene data unit A. This set covers a series of indexes that are closely related to the scene data unit A in terms of space and features, comprehensively reflecting its connection with the surrounding scene elements.

[0044] Finally, generate a construction input set based on the obtained topological feature index set. This construction input set serves as the key data for subsequent virtual scene construction, providing strong support for accurately constructing the virtual scenes at the entrance of the teaching building and the entire campus, enabling the constructed virtual scenes to more realistically reflect the actual layout and characteristics of the campus.

[0045] Embodiment 3: After the virtual scene construction is completed, for each scene data unit, the spatial difference value between this scene data unit and adjacent units is monitored in real time. When there is a spatial difference value exceeding the preset threshold, the adjacent unit is used as an abnormal reference unit, and the dynamic parameters of the scene data unit are corrected based on the abnormal reference unit to generate an updated virtual scene construction result. This process can promptly detect possible problems in the virtual scene and continuously optimize the virtual scene through dynamic parameter correction to make it more in line with the actual situation. During the monitoring process, by continuously comparing information such as the spatial position and geometric features of the scene data unit and adjacent units, once a difference exceeding the preset range is found, the correction program is immediately started to ensure the accuracy and stability of the virtual scene.

[0046] Suppose a virtual scene of a large commercial plaza is being constructed, which contains multiple commercial buildings, parking lots, fountain landscapes and other elements inside the scene, and these elements are divided into numerous scene data units.

[0047] After the virtual scene construction is completed, each scene data unit is monitored in real time. For example, select the scene data unit A corresponding to a landmark tower in the commercial plaza, and the scene data unit B at the entrance of the adjacent parking lot.

[0048] Set a calculation method for the spatial difference value. Assume the spatial difference value The calculation formula is: , where represents the spatial coordinates of the scene data unit A, represents the spatial coordinates of the scene data unit B. This formula measures the spatial difference degree between the two scene data units by calculating the Euclidean distance in the three-dimensional space.

[0049] Preset a threshold , assume (unit: meter, which can be set according to the actual scene accuracy requirements). During the real-time monitoring process, continuously calculate the spatial difference value between the scene data unit A and the adjacent unit B.

[0050] At a certain moment, due to possible data updates or model adjustments in the virtual scene, the calculated spatial difference value meters, exceeding the preset threshold meter. At this time, the adjacent unit B is used as an abnormal reference unit.

[0051] Based on the abnormal reference unit B, dynamic parameter correction is performed on the scene data unit A. For example, it is found through inspection that there is a deviation in the spatial coordinates of the scene data unit A during the data update process, resulting in an excessive spatial difference from the adjacent unit B. According to the accurate spatial position information of the scene data unit B and the overall layout logic of the commercial plaza, the spatial coordinate parameters of the scene data unit A are corrected. At the same time, considering that the scene data unit A is a tower building, the relationships such as light and shadow, occlusion with the surrounding environment will also change due to the change in spatial position, and the dynamic evolution parameters such as lighting and shadow are adjusted accordingly.

[0052] After completing the parameter correction of the scene data unit A, the relationships and parameters between the scene data units in the entire virtual scene are recalculated to generate an updated virtual scene construction result. After such processing, the spatial relationship between the tower building and the parking lot entrance in the virtual scene becomes reasonable, and the virtual scene of the entire commercial plaza is more in line with the actual situation, improving the accuracy and realism of the virtual scene.

[0053] Embodiment 4: When the pre-configured 3D modeling algorithm is trained, first, a historical 3D data set is obtained, and the mapping relationship between geometric feature parameters and dynamic evolution parameters is extracted from the data set. An initial 3D modeling algorithm is constructed according to this mapping relationship, and the grid division method is used to iteratively optimize the algorithm. During the iteration process, the parameters and structure of the algorithm are continuously adjusted to gradually reduce the error between the predicted parameters output by the algorithm and the measured parameters. When the error between the predicted parameters output by the algorithm and the measured parameters is lower than the preset threshold, it is determined that the algorithm configuration is completed. The 3D modeling algorithm obtained through such training can more accurately process the construction input set of the scene data unit, determine more reasonable dynamic evolution parameters, thereby improving the quality and efficiency of virtual scene construction.

[0054] Suppose it is necessary to construct a virtual scene for a historical block with a typical architectural style. To achieve this goal, a precise 3D modeling algorithm needs to be trained.

[0055] Collect a large number of historical 3D datasets of this historical block. These data come from a wide range of sources, such as the previous block surveying data, high-precision image materials taken in the past, etc. Extract the mapping relationship between geometric feature parameters and dynamic evolution parameters from these datasets. For example, for an ancient temple in the block, the geometric feature parameters include the length, width and height of the temple, the dimensions and spatial positions of each building component (such as eaves, brackets, etc.); the dynamic evolution parameters include the changes in the appearance effect of the temple under the light and shadow changes over time, the changes in the surface texture wear of the building caused by natural erosion, etc. By analyzing the data of a large number of similar temples and other buildings in the block, summarize the corresponding relationship between these geometric feature parameters and dynamic evolution parameters. For example, it is found that the orientation and height of a building will affect its light and shadow effects at different times, and then affect the light-related parameters in the dynamic evolution parameters.

[0056] Construct an initial 3D modeling algorithm according to the extracted mapping relationship. This algorithm initially sets the rules for calculating and predicting dynamic evolution parameters based on geometric feature parameters. Then, use the grid division method to iteratively optimize the algorithm. Taking the area where the temple is located as an example, divide the space within a certain range around the temple and its surrounding into grids, and each grid is used as an independent analysis unit. In each iteration process, compare the dynamic evolution parameters predicted by the algorithm in each grid with the actually measured parameters. For example, if there is a difference between the predicted light intensity on the building surface in a certain grid at a specific time and the actually measured light intensity, adjust the parameters and rules related to light calculation in the algorithm according to this difference. Through continuous iteration, the error between the predicted parameters output by the algorithm and the measured parameters gradually decreases.

[0057] In the iterative optimization process, set a preset threshold. For example, an error rate within 5% is considered an acceptable accuracy range. When the error between the predicted parameters output by the algorithm and the measured parameters is lower than this preset threshold, it is determined that the algorithm configuration is completed. The 3D modeling algorithm obtained through such training can more accurately determine reasonable dynamic evolution parameters according to the input geometric feature parameters when processing the scene data units of the historical block. For example, when constructing a virtual scene, for buildings with different positions and structures in the block, it can accurately predict the appearance changes under different times and different natural conditions based on their geometric features, thus providing reliable technical support for constructing a highly restored and dynamically changing historical block virtual scene, enabling people to more realistically feel the style of the historical block in the virtual scene.

[0058] Example 5: The 3D modeling engine includes 3D modeling layers of spatial grid division, texture mapping rules and illumination model. When generating a 3D scene atlas, the dynamic evolution parameters are superimposed on the 3D modeling layer for visual expression. The method for generating the 3D modeling layer is to first collect the terrain point cloud data, building surface data and ambient lighting data of the target area. The terrain point cloud data is gridded to generate a spatial grid division layer. Through this processing, the terrain data can be converted into a grid form that is easy to process and analyze; the building surface data is texture mapped to generate a texture mapping rule layer, so that the building surface can present a real texture effect; the ambient lighting data is simulated to generate an illumination model layer to simulate the lighting conditions in the scene. Finally, the spatial grid division layer, texture mapping rule layer and illumination model layer are spatially superimposed and analyzed to form a 3D modeling layer. The 3D modeling layer generated in this way can provide rich information for the construction of virtual scenes, making the virtual scenes more realistic and vivid.

[0059] Taking the construction of a virtual scene of a large theme park as an example, when constructing the virtual scene of a theme park, it is necessary to use a 3D modeling engine to generate a 3D modeling layer containing a variety of information to achieve a more realistic and vivid virtual scene display.

[0060] First, data collection is carried out to collect the terrain point cloud data, building surface data and ambient lighting data of the theme park. For the terrain point cloud data, professional LiDAR equipment is used to scan the terrain of the entire theme park to obtain accurate three-dimensional coordinate information of each terrain in the park, such as the undulating terrain in the roller coaster area and the terrain changes around the artificial lake. The building surface data is obtained through field photography and measurement. For various characteristic buildings in the theme park, such as fairy tale castles and pirate ship-shaped entertainment facilities, the surface texture details of the buildings are photographed in all directions, and the dimensions of each part of the building are measured. The collection of ambient lighting data uses light sensors to measure information such as light intensity and light angle in various areas of the park at different time periods and under different weather conditions.

[0061] After the data is collected, the terrain point cloud data is gridded to generate a spatial grid division layer. The large amount of terrain point cloud data collected is divided into small grids according to certain rules. For example, 1 meter × 1 meter is a grid unit, and each grid is assigned a corresponding terrain height value, thus forming a spatial grid division layer that can clearly show the undulating terrain of the park. Through this layer, you can intuitively see which areas in the theme park are high and which are low-lying areas, providing a terrain basic framework for subsequent scene construction.

[0062] Generate a texture mapping rule layer by performing texture mapping on the building surface data. Taking a fairy tale castle as an example, process the high-definition photos of the castle surface taken before, and accurately map the textures on the photos to the corresponding building model surface according to the actual structure and size of the castle. Determine the rules of texture mapping, such as the arrangement direction of the brick textures on the castle walls, the correspondence between the positions of doors and windows and the actual photos, etc., so that the castle can present a lifelike appearance in the virtual scene as in reality. The generated texture mapping rule layer endows the buildings in the virtual scene with rich details and a sense of reality.

[0063] Generate a lighting model layer by simulating the environmental lighting data. According to the collected lighting data under different time periods and weather conditions, use professional lighting simulation software to simulate the corresponding lighting effects in the virtual scene. For example, at noon on a sunny day, simulate strong direct sunlight to make the building produce obvious shadows; at dusk, simulate soft warm-colored light to create a warm atmosphere. By adjusting parameters such as the intensity, angle, and color of the lighting, generate a lighting model layer that conforms to the actual situation, making the light and shadow effects in the virtual scene more natural.

[0064] Perform a spatial overlay analysis on the spatial grid division layer, the texture mapping rule layer, and the lighting model layer to form a 3D modeling layer. In this process, integrate the information of each layer to make the terrain, building, and lighting information related to each other. For example, the undulations of the terrain will affect the illumination angle of the light and the distribution of shadows, the position and shape of the building will block the light, and the texture will present different effects under the light. Through this spatial overlay analysis, the generated 3D modeling layer contains rich information, providing a comprehensive basis for subsequent generation of 3D scene maps and construction of virtual scenes. When generating the 3D scene map, overlay the dynamic evolution parameters of the scene data unit onto this 3D modeling layer for visual expression, enabling virtual scene constructors to more intuitively see the position, characteristics, and dynamic changes of each scene data unit in the virtual scene of the entire theme park, thereby constructing a highly restored and vivid virtual scene of the theme park.

[0065] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, article, or device.

[0066] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a virtual scene driven by real - scene three - dimensional data, characterized in that, The method includes: Obtaining multi-source three-dimensional data of a target area, and extracting geometric feature parameters of each scene data unit from a preset three-dimensional modeling engine, where the geometric feature parameters include spatial coordinate attributes, texture attributes, and semantic attributes of the scene data unit; For each scene data unit, calculating topological feature indexes of the scene data unit respectively based on the geometric feature parameters, and generating a construction input set according to the multiple topological feature indexes; where the topological feature index is a comprehensive quantization parameter for each scene data unit, integrating its spatial coordinate attributes and texture attributes; the construction input set is a standardized data set composed of multiple selected topological feature indexes; For each scene data unit, loading a pre-configured three-dimensional modeling algorithm, and determining dynamic evolution parameters of the scene data unit during the construction process based on the spatial relationship between the multiple feature indexes included in the construction input set and the processing rules of the three-dimensional modeling algorithm; Annotating the dynamic evolution parameters to each scene data unit in the three-dimensional modeling engine to generate a three-dimensional scene atlas; For each scene data unit, performing spatial consistency detection according to the dynamic evolution parameters to obtain a detection result; Based on the detection result, performing virtual scene construction to generate a virtual scene construction result.

2. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, characterized in that, The performing virtual scene construction based on the detection result to generate a virtual scene construction result includes: when the detection result indicates that there are conflicts among multiple dynamic evolution parameters included in the scene data unit, determining the scene data unit as an abnormal unit; obtaining the conflict parameters of the abnormal unit as the correction object, and determining the multi-source data priority corresponding to the conflict parameters; performing parameter fusion on the scene data unit based on the data priority to generate a virtual scene construction result.

3. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, characterized in that, The multi-source three-dimensional data includes lidar point cloud data and oblique photography image data. The data priority of the lidar point cloud data is the first priority, and the data priority of the oblique photography image data is the second priority; the performing parameter fusion on the scene data unit based on the data priority to generate a virtual scene construction result includes: Adjusting the construction input set corresponding to the oblique photography image data according to the order of the first priority and the second priority; Based on the three-dimensional scene atlas, extracting the dynamic evolution parameters of the oblique photography image data in the adjusted construction input set; Combining the dynamic evolution parameters and the three-dimensional scene atlas to perform parameter fusion to generate a virtual scene construction result.

4. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, characterized in that, The for each scene data unit, calculating topological feature indexes of the scene data unit respectively based on the geometric feature parameters, and generating the construction input set according to the multiple topological feature indexes includes: For each scene data unit, determining multiple adjacent units adjacent to it in space; Calculating the correlation degree between the geometric feature parameters and texture attributes of each adjacent unit, and determining the first feature unit with the highest correlation degree from the adjacent units based on the multiple correlation degrees; Based on the first feature unit, determine a plurality of extended units adjacent to the feature unit, calculate the correlation degree between the geometric feature parameters and the texture attributes of each extended unit, and determine the next feature unit with the highest correlation degree from the extended units based on the plurality of correlation degrees; Repeat the steps of determining a plurality of extended units adjacent to the feature unit, calculating the correlation degree of each extended unit, and determining the next feature unit with the highest correlation degree until all scene data units in the target area are traversed to obtain a set of topological feature indicators for the scene data units; Generate the construction input set based on the set of topological feature indicators.

5. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 3, characterized in that, The calculation of the correlation degree between the geometric feature parameters and the texture attributes of each adjacent unit includes: Calculate the first correlation degree between the spatial coordinate attributes and the texture attributes of each adjacent unit, and the second correlation degree between the spatial coordinate attributes and the semantic attributes of each adjacent unit respectively; For each adjacent unit, use the weighted sum of the first correlation degree and the second correlation degree as the correlation degree of the adjacent unit.

6. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 4, characterized in that, The determination of the first feature unit with the highest correlation degree from the adjacent units based on the plurality of correlation degrees includes: Store the plurality of correlation degrees corresponding to the plurality of adjacent units into a feature candidate list, and sort the correlation degrees in the feature candidate list in descending order to obtain a sorting result; Based on the sorting result, determine the adjacent unit corresponding to the highest correlation degree as the first feature unit.

7. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, characterized in that, After generating the virtual scene construction result based on the detection result, it further includes: When the virtual scene construction result is generated, for each scene data unit, real-time monitor the spatial difference value between the scene data unit and the adjacent unit; When there is a spatial difference value exceeding a preset threshold, use the adjacent unit as an abnormal reference unit, and perform dynamic parameter correction on the scene data unit based on the abnormal reference unit to generate an updated virtual scene construction result.

8. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, characterized in that, The training process of the pre-configured 3D modeling algorithm includes: Obtain a historical 3D data set, and extract the mapping relationship between the geometric feature parameters and the dynamic evolution parameters in the data set; Construct an initial 3D modeling algorithm according to the mapping relationship, and use the grid division method to iteratively optimize the algorithm; When the error between the predicted parameters and the measured parameters output by the algorithm is lower than a preset threshold, determine that the algorithm configuration is completed.

9. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 1, wherein, The 3D modeling engine also includes a 3D modeling layer for spatial grid division, texture mapping rules, and lighting models. When generating the 3D scene atlas, superimpose the dynamic evolution parameters onto the 3D modeling layer for visual expression.

10. A method for constructing a virtual scene driven by real - scene three - dimensional data according to claim 8, characterized in that, The generation method of the 3D modeling layer includes: Collect the terrain point cloud data, building surface data, and environmental light data of the target area; Perform grid processing on the terrain point cloud data to generate a spatial grid division layer, perform texture mapping on the building surface data to generate a texture mapping rule layer, and simulate the environmental light data to generate a lighting model layer; Perform spatial overlay analysis on the spatial grid division layer, texture mapping rule layer, and lighting model layer to form the 3D modeling layer.

Images