Three-dimensional model data processing system based on augmented reality

Abstract

The invention relates to the field of 3D image processing, in particular to an augmented reality-based three-dimensional model data processing system, which comprises a data acquisition module, a grid classification simplification module, a three-dimensional model simplification module and a display module, according to the three factors of the change of the normal vector of the triangular mesh surface, the geometric structure and the geometric structure distribution, subspace structure complexity is constructed; obtaining the complexity of subspace types; obtaining subspace feature relevance; obtaining the complexity of the subspace; completing the division of the subspaces; and simplifying the grid vertex of the three-dimensional model by utilizing a clustering algorithm according to a division result. Therefore, the simplification processing of the three-dimensional model data is realized, the problem of shape deformation or data loss of the three-dimensional model in the simplification process is greatly avoided, and the error between the simplified model and the original model is reduced.

Description

Technical Field

[0001] The present invention relates to the field of 3D image processing, and in particular to a three-dimensional model data processing system based on augmented reality. Background Art

[0002] Augmented Reality (AR) is a technology that combines virtual content with real-world scenes, superimposing virtual objects into the real environment through display devices (such as mobile phones, head-mounted displays, etc.).

[0003] Complex 3D models have a large number of faces, which increases the computational cost of rendering and drawing. Reducing the number of faces can speed up rendering and improve real-time interactive performance, especially in applications that require real-time rendering, such as games, virtual reality, and augmented reality. Large 3D models with high face counts can consume a large amount of memory. By simplifying the mesh, the model can be made easier to perform subsequent editing and processing, such as deformation, texture mapping, and UV unwrapping. Simplified models have fewer faces and vertices, making editing and processing more efficient.

[0004] When using the mesh vertex clustering algorithm to divide the subspace, due to the inaccuracy of the division space, some detail areas may be divided into the same subspace, which in turn causes the model to lose a lot of information and make the model inaccurate. Therefore, the present invention uses a mesh simplification algorithm based on mesh vertex clustering to simplify the three-dimensional mesh model of augmented reality. Summary of the Invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide a three-dimensional model data processing system based on augmented reality. The technical solutions adopted are as follows:

[0006] The present invention provides a three-dimensional model data processing system based on augmented reality, the system comprising:

[0007] A data acquisition module acquires a vertex set and a facet set; the vertex set records the vertex coordinates and corresponding pixel values ​​in the three-dimensional model; the facet set records the three vertex coordinates of each triangular mesh surface;

[0008] The mesh classification simplification module obtains the three-dimensional data model and obtains the first structural complexity of the subspace based on the normal vector distribution of the triangular mesh surface; obtains the triangular mesh surface vector based on the three vertex coordinates of the triangular mesh surface; obtains the second structural complexity of the subspace based on the clustering results of the triangular mesh face quantity and the vertex angle of the triangular mesh surface; obtains the geometric structure similarity of the triangular mesh surface based on the clustering results of the triangular mesh face quantity and the geometric distribution of the triangular mesh surface; obtains the third structural complexity of the subspace based on the geometric structure similarity of the triangular mesh surface; obtains the subspace structural complexity based on the first, second and third structural complexities of the subspace;

[0009] Obtain subspace type complexity based on the clustering results of triangle mesh vertex pixel values ​​in the subspace; obtain subspace feature relevance based on the clustering results of triangle mesh face quantities and subspace type complexity; obtain subspace complexity based on subspace structure complexity, subspace type complexity and subspace feature relevance; divide the subspace according to the subspace complexity;

[0010] The 3D model simplification module uses a clustering algorithm to simplify the vertices of the 3D model mesh based on the segmentation results;

[0011] The display module displays the simplified three-dimensional model through the display system.

[0012] Furthermore, the first structural complexity of the subspace is obtained according to the normal vector distribution of the triangular mesh surface, specifically:

[0013]

[0014] Where Sz represents the first structural complexity of the subspace, r represents the number of complete triangle meshes contained in the current space, f represents the number of adjacent triangle meshes of the triangle mesh, and α (i,j) Represents the angle value of the mesh surface normal vector corresponding to the i-th triangle mesh and its adjacent j-th triangle mesh in space, Represents the average value of the angle between the normal vectors of the i-th triangle mesh and its adjacent triangle meshes in space.

[0015] Furthermore, the subspace second structure complexity is obtained based on the clustering results of the triangle mesh face quantities and the triangle mesh face vertex angles. The specific expression is:

[0016] Calculate the sum of the absolute values ​​of the differences between the range of the three vertex angle values ​​in each triangular mesh surface in the cluster and the mean of the range of the three vertex angle values ​​in all triangular mesh surfaces in the cluster; and take the product of the mean of the sum values ​​of all clusters and the number of clusters as the second structural complexity.

[0017] Furthermore, the obtaining of geometric structure similarity of the triangular mesh surfaces according to the clustering results of the triangular mesh face quantities and the geometric distribution of the triangular mesh surfaces specifically includes:

[0018] The triangular mesh surface is marked according to the clustering results: if the current triangular mesh surface and its adjacent triangular mesh surface are in the same cluster, the adjacent triangular mesh surface is marked as 1, otherwise it is marked as 0, and the sum of the marking results of all triangular mesh surfaces adjacent to the current triangular mesh surface is recorded as the geometric structure similarity of the current triangular mesh surface.

[0019] Furthermore, obtaining the third structural complexity of the subspace according to the geometric structural similarity of the triangular mesh surface includes:

[0020] The sum of the geometric structure similarities of all triangular mesh surfaces in the current subspace is calculated; and the reciprocal of the sum is used as the third structural complexity of the subspace.

[0021] Furthermore, the subspace structural complexity is the product of the first, second, and third structural complexities of the subspace.

[0022] Furthermore, the subspace type complexity is obtained based on the clustering result of the pixel values ​​of the triangle mesh vertices in the subspace, and the expression is:

[0023]

[0024] Where Yz represents the subspace type complexity, W represents the number of clusters obtained based on the clustering of pixel values ​​corresponding to the subspace vertices, and Δg i Indicates the extreme difference of the pixel value of the i-th triangle mesh vertex in the subspace, Δg (i,j) It represents the range of pixel values ​​of the j-th triangular mesh surface adjacent to the i-th triangular mesh surface, r represents the number of complete triangular meshes contained in the current subspace, and norm() represents the normalization function.

[0025] Furthermore, the obtaining of subspace feature relevance based on the clustering results of the triangular mesh dimensions and the subspace type complexity includes:

[0026] According to the calculation steps of subspace type complexity, the type complexity of each cluster cluster obtained by triangular mesh surface vector clustering is obtained; the absolute value of the difference between the type complexity of each cluster cluster and the type complexity of the subspace in which it is located is calculated; and the ratio of the absolute value in the cluster cluster to the type complexity of each cluster cluster is accumulated as the subspace feature correlation.

[0027] Furthermore, the subspace complexity is obtained according to the subspace structure complexity, subspace type complexity and subspace feature correlation, specifically:

[0028] The product of the subspace structure complexity and the subspace type complexity is calculated; and the ratio of the product to the subspace feature correlation is used as the subspace complexity.

[0029] Furthermore, the subspace is divided according to the subspace complexity, specifically:

[0030] Set a complexity threshold; when the subspace complexity is greater than the complexity threshold, use octree to divide the subspace.

[0031] The present invention has at least the following beneficial effects:

[0032] The present invention analyzes three-dimensional mesh model data and constructs a spatial triangular mesh structure complexity based on spatial structure analysis, combining the spatial triangular mesh normal vectors, triangular mesh geometry, and corresponding distribution. It further constructs the spatial triangular mesh vertex type complexity based on the vertex pixel value analysis, analyzes the subspace feature correlation, and constructs a subspace complexity index. When using an octree to partition the subspace, this index is used as a condition to evaluate whether the subspace needs to be further divided. This adaptive partitioning is performed, making the partitioning result more consistent with the corresponding three-dimensional model structure, thereby avoiding the situation where the simplified mesh does not conform to the original model due to inaccurate subspace partitioning, and reducing the error between the simplified model and the original model.

[0033] Furthermore, it not only achieves the simplification of the three-dimensional model, reduces memory usage and speeds up rendering, but also displays the adjusted three-dimensional virtual scene in real time through the display system, improves the real-time interactive performance, and achieves the goal of a comfortable picture with sufficient details. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 A block diagram of a three-dimensional model data processing system based on augmented reality is provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0036] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a 3D model data processing system based on augmented reality proposed by the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0037] 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.

[0038] The following describes in detail a specific solution of a three-dimensional model data processing system based on augmented reality provided by the present invention with reference to the accompanying drawings.

[0039] See also Figure 1 , which shows a block diagram of a three-dimensional model data processing system based on augmented reality provided by an embodiment of the present invention. The system includes: a data acquisition module 101, a grid classification simplification module 102, a three-dimensional model simplification module 103, and a display module 104.

[0040] The data acquisition module 101 uses AR equipment to obtain three-dimensional data models and sends the obtained three-dimensional models in batches to the processing center.

[0041] The corresponding data model is mesh data, including a vertex set V and a face set F. The vertex set records the vertex coordinates and corresponding pixel values ​​in the three-dimensional model, and the face set records the three vertex coordinates corresponding to each triangle mesh.

[0042] The grid classification simplification module 102 obtains the subspace complexity and completes the subspace division according to the subspace complexity.

[0043] Specifically, in the augmented reality process, it is necessary to simplify the initially obtained complex detailed model in order to achieve the purpose of normal operation on some old devices. Therefore, this embodiment uses a mesh simplification algorithm based on mesh vertex clustering to simplify the augmented reality three-dimensional mesh model, and uses an octree algorithm to adaptively divide the bounding box. By analyzing the subspace after octree division, a complexity index is constructed to evaluate whether the subspace needs to be further divided, thereby avoiding the situation where the simplified mesh based on the division result does not match the original model due to inaccurate subspace division. The method for constructing the complexity is as follows:

[0044] 1) Structural complexity. A 3D model can be represented by individual mesh triangles, so the complexity of its structure can be determined by three factors: the change in the normal vector of the mesh triangle, the geometric structure of the triangle mesh surface, and the geometric structure distribution of the triangle mesh surface. The steps are as follows:

[0045] The difference between the normal vectors of each complete triangular mesh surface and its adjacent triangular mesh surfaces in the calculation space (the adjacent mesh surfaces are adjacent to each other along the same edge, that is, each triangular mesh surface has three adjacent triangular mesh surfaces). The greater the overall difference in the normal vectors in the calculated space and the greater the difference in the normal vectors of the adjacent triangular mesh surfaces, the higher the corresponding structural complexity in the current subspace. The first structural complexity of the subspace obtained based on this is calculated as follows:

[0046]

[0047] Where r represents the number of complete triangle meshes contained in the current space, f represents the number of adjacent triangle meshes of the triangle mesh, and α (i,j) Represents the angle value of the mesh surface normal vector corresponding to the i-th triangle mesh and its adjacent j-th triangle mesh in space, Represents the average value of the angle between the normal vectors of the i-th triangle mesh and its adjacent triangle meshes in space.

[0048] According to the vertex coordinates of the triangular mesh surface, the area y0 of the triangular mesh surface and the angle values ​​y1, y2, and y3 of the three vertices of the triangular mesh surface are calculated. It should be noted that y1, y2, and y3 represent the angle values ​​of the three vertices of the corresponding triangular mesh surface arranged from small to large. Construct the triangular mesh surface vector Y = {y0, y1, y2, y3}, and perform cluster analysis on it. It should be noted that the implementer of the clustering algorithm can obtain it by himself according to the actual situation. In this embodiment, the clustering algorithm uses the DBSCAN clustering algorithm. Further analysis is performed in combination with the relationship between the three angle values ​​of the triangular mesh surface vertices to construct the second structural complexity in the subspace:

[0049]

[0050] Where Jz represents the second structural complexity of the subspace, K represents the number of clusters obtained by clustering the triangular mesh surface vectors in the subspace, and T l represents the number of triangular mesh faces contained in the lth cluster, β (i,l) Indicates the range of the angle values ​​of the three vertices in the i-th triangular mesh in the l-th cluster, Represents the average range of the angles of the three vertices within the triangular mesh within the lth cluster. When clustering triangular mesh geometry, the greater the number of clusters obtained, and the greater the difference between the range of the angles of the triangular meshes within the corresponding cluster and the average range within the cluster, the more complex the structure of the triangular mesh within the current subspace.

[0051] According to the clustering results of the above process, the distribution similarity of the triangular mesh surfaces is analyzed: First, the triangular mesh surfaces are marked according to the clustering results. The marking statistical method is to determine whether the triangular mesh surfaces adjacent to the current triangular mesh surface are in the same cluster as the current triangular mesh surface. If so, it is recorded as 1, otherwise it is 0. The obtained marking results are accumulated and recorded as the geometric structure similarity degree E of the triangular mesh surface and its adjacent triangular mesh surfaces. The value is [0,3]. The larger the value, the greater the geometric structure similarity between the current triangular mesh surface and its adjacent triangular mesh surfaces. Based on the above clustering results and the similarity degree of the geometric structure distribution of the triangular mesh surfaces, the third structural complexity index in the subspace is constructed. The calculation method is as follows:

[0052]

[0053] Where, E i The geometric similarity between the i-th triangle mesh face and its adjacent triangle mesh faces in the subspace is expressed as , and r represents the number of complete triangle meshes contained in the current space. The greater the geometric similarity between the triangle mesh face in the subspace being sought and its adjacent triangle mesh faces, the less complex the current subspace structure is.

[0054] Finally, based on the above analysis, the calculation method of subspace structure complexity is as follows:

[0055] Tz=norm(Sz*Jz*Pz)

[0056] Where Tz is the subspace complexity, Sz is the primary structural complexity of the subspace, Jz is the secondary structural complexity of the subspace, and Pz is the tertiary structural complexity of the subspace. norm() represents the normalization function. The complexity obtained by analyzing the normal vectors of the subspace triangle mesh, the complexity obtained by analyzing the geometric structure of the triangle mesh, and the complexity obtained by analyzing the distribution of the geometric structure of the triangle mesh are combined to construct the subspace structural complexity index. The larger the values ​​of Sz, Jz, and Pz, the higher the structural complexity of the triangle mesh in the subspace.

[0057] 2) Constructing a type complexity index. Obtain the vertex pixel values ​​corresponding to each triangular mesh surface. Since different pixel values ​​represent different objects or different object features, the triangular mesh surfaces in the subspace can be analyzed based on the vertex corresponding pixel values. In this embodiment, the DBSCAN clustering algorithm is used to cluster the vertex pixel values ​​in the obtained subspace to construct a type-based subspace complexity index:

[0058]

[0059] Where Yz represents the subspace type complexity, W represents the number of clusters obtained based on the clustering of pixel values ​​corresponding to the subspace vertices, and Δg i Indicates the extreme difference of the pixel value of the i-th triangle mesh vertex in the subspace, Δg (i,j) represents the range of the pixel values ​​of the jth triangle mesh surface adjacent to the i-th triangle mesh surface, and r represents the number of complete triangle meshes contained in the current space. It should be noted that the more clusters are obtained based on the pixel values ​​of the triangle mesh surface vertices, the larger the range of the pixel values ​​of the corresponding triangle mesh surface vertices, and the more obvious the difference in the range of the pixel values ​​of the corresponding triangle mesh surface and its adjacent triangle mesh surfaces, the more complex the subspace features are, and the more complex the object types in the current subspace are.

[0060] 3) Subspace feature correlation analysis. After clustering the triangular mesh structure in the subspace, if the corresponding species complexity in each cluster decreases significantly, or the feature correlation between species complexity and structural complexity is strong, it means that the complexity in the current subspace is low. The subspace feature correlation is analyzed based on species complexity and structural complexity. The specific formula is:

[0061]

[0062] Where Xz is the subspace feature correlation, K represents the number of clusters obtained by clustering the geometric structure of the subspace triangular mesh surface, and Yz u Represents the species complexity of the uth cluster in the subspace triangular mesh geometry clustering. Analyzing the clusters derived from the subspace triangular mesh geometry, the smaller the cluster's species complexity, the greater the difference between the cluster's species complexity and the subspace's species complexity. This indicates a greater correlation between the subspace's geometry and species, and a lower corresponding subspace complexity.

[0063] 4) Subspace complexity analysis. Based on the structural complexity, type complexity, and the correlation between the two obtained in steps 1), 2), and 3), the subspace complexity Rz is constructed. The specific formula is as follows:

[0064]

[0065] Where Rz represents the subspace complexity, Tz represents the subspace structure complexity, Yz represents the subspace type complexity, and Xz represents the subspace feature correlation.

[0066] That is, the higher the structural complexity of the desired subspace, the higher the category complexity, and the lower the correlation between subspace features, the higher the subspace complexity. A complexity threshold η is set. When the desired subspace complexity Rz is greater than the threshold, the subspace is partitioned; otherwise, no further partitioning is performed. It should be noted that the complexity threshold can be set by the implementer based on actual circumstances. In this embodiment, the complexity threshold η = 0.5.

[0067] The initial subspace is partitioned using an octree. The depth of the partitioning tree can be determined by the implementer based on actual conditions. In this embodiment, the depth of the partitioning tree is h = 4, i.e., 4096 subspaces are initially partitioned. The subspace complexity is calculated for each subspace in turn according to the above method, and then a determination is made as to whether each subspace should be further partitioned. If the conditions are met, the corresponding subspace is partitioned into eight subspaces, and the above determination method is repeated until the complexity of all subspaces is less than the complexity threshold.

[0068] The 3D model simplification module 103 combines the bounding box division result of the adaptive octree and substitutes it into the mesh vertex clustering algorithm.

[0069] Specifically, when vertices are merged, the original vertex is deleted and the newly generated vertex is connected to the original mesh to form a new triangle mesh. The specific method is to find the original vertex, the set of vertices connected to the vertex, and the newly generated vertex. If the original vertex set has an edge connected to the vertex, then the new vertex is connected to the original vertex; otherwise, it is not connected. This completes the 3D model mesh vertex simplification.

[0070] Display module 104, after the model is simplified, when the input module of the intelligent display device inputs the device information, the image is displayed accordingly.

[0071] Specifically, the real environment in which it is located is obtained through the AR device; a three-dimensional virtual scene is obtained, and the three-dimensional virtual scene is displayed in real time through the display system; the three-dimensional virtual scene is manually adjusted according to the real environment, where the adjustment parameter is the complexity threshold mentioned above, and the adjusted three-dimensional virtual scene is displayed in real time through the display system.

[0072] In summary, the embodiment of the present invention analyzes the three-dimensional mesh model data, and based on the spatial structure analysis, combines the normal vectors of the triangular mesh in space, the geometric structure of the triangular mesh, and the corresponding distribution to construct the structural complexity of the triangular mesh in space. It further constructs the type complexity of the spatial triangular mesh vertex based on the vertex pixel value analysis in space, analyzes the correlation of subspace features, and constructs a subspace complexity index. When using the octree to divide the subspace, this index is used as a condition to evaluate whether the subspace needs to be further divided, and the adaptive division is completed so that the division result is more consistent with the corresponding three-dimensional model structure, thereby avoiding the situation where the simplified mesh does not match the original model due to inaccurate subspace division, and reducing the error between the simplified model and the original model.

[0073] Furthermore, it not only achieves the simplification of the three-dimensional model, reduces memory usage and speeds up rendering, but also displays the adjusted three-dimensional virtual scene in real time through the display system, improves the real-time interactive performance, and achieves the goal of a comfortable picture with sufficient details.

[0074] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0075] 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.

[0076] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A three-dimensional model data processing system based on augmented reality, characterized in that: The system comprises: A data acquisition module acquires a vertex set and a facet set; the vertex set records the vertex coordinates and corresponding pixel values ​​in the three-dimensional model; the facet set records the three vertex coordinates of each triangular mesh surface; The mesh classification simplification module obtains the three-dimensional data model and obtains the first structural complexity of the subspace based on the normal vector distribution of the triangular mesh surface; obtains the triangular mesh surface vector based on the three vertex coordinates of the triangular mesh surface; obtains the second structural complexity of the subspace based on the clustering results of the triangular mesh face quantity and the vertex angle of the triangular mesh surface; obtains the geometric structure similarity of the triangular mesh surface based on the clustering results of the triangular mesh face quantity and the geometric distribution of the triangular mesh surface; obtains the third structural complexity of the subspace based on the geometric structure similarity of the triangular mesh surface; obtains the subspace structural complexity based on the first, second and third structural complexities of the subspace; Obtain subspace type complexity based on the clustering results of triangle mesh vertex pixel values ​​in the subspace; obtain subspace feature relevance based on the clustering results of triangle mesh face quantities and subspace type complexity; obtain subspace complexity based on subspace structure complexity, subspace type complexity and subspace feature relevance; divide the subspace according to the subspace complexity; The 3D model simplification module uses a clustering algorithm to simplify the vertices of the 3D model mesh based on the segmentation results; The display module displays the simplified three-dimensional model through the display system.

2. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The first structural complexity of the subspace is obtained according to the normal vector distribution of the triangular mesh surface, specifically: Where Sz represents the first structural complexity of the subspace, r represents the number of complete triangle meshes contained in the current space, f represents the number of adjacent triangle meshes of the triangle mesh, and α (i,j) Represents the angle value of the mesh surface normal vector corresponding to the i-th triangle mesh and its adjacent j-th triangle mesh in space, Represents the average value of the angle between the normal vectors of the i-th triangle mesh and its adjacent triangle meshes in space.

3. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The second structural complexity of the subspace is obtained based on the clustering results of the triangular mesh face quantities and the vertex angles of the triangular mesh face. The specific expression is: Calculate the sum of the absolute values ​​of the differences between the range of the three vertex angle values ​​in each triangular mesh surface in the cluster and the mean of the range of the three vertex angle values ​​in all triangular mesh surfaces in the cluster; and take the product of the mean of the sum values ​​of all clusters and the number of clusters as the second structural complexity.

4. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The obtaining of the geometric structure similarity of the triangular mesh surface according to the clustering result of the triangular mesh surface quantity and the geometric distribution of the triangular mesh surface specifically includes: The triangular mesh surface is marked according to the clustering results: if the current triangular mesh surface and its adjacent triangular mesh surface are in the same cluster, the adjacent triangular mesh surface is marked as 1, otherwise it is marked as 0, and the sum of the marking results of all triangular mesh surfaces adjacent to the current triangular mesh surface is recorded as the geometric structure similarity of the current triangular mesh surface.

5. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The obtaining of the third structural complexity of the subspace according to the geometric structural similarity of the triangular mesh surface includes: The sum of the geometric structure similarities of all triangular mesh surfaces in the current subspace is calculated; and the reciprocal of the sum is used as the third structural complexity of the subspace.

6. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The subspace structural complexity is the product of the first, second and third structural complexities of the subspace.

7. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The subspace type complexity is obtained based on the clustering result of the pixel values ​​of the triangle mesh vertices in the subspace, and the expression is: Where Yz represents the subspace type complexity, W represents the number of clusters obtained based on the clustering of pixel values ​​corresponding to the subspace vertices, and Δg i Indicates the extreme difference of the pixel value of the i-th triangle mesh vertex in the subspace, Δg (i,j) It represents the range of pixel values ​​of the j-th triangular mesh surface adjacent to the i-th triangular mesh surface, r represents the number of complete triangular meshes contained in the current subspace, and norm() represents the normalization function.

8. The three-dimensional model data processing system based on augmented reality according to claim 7, characterized in that: The obtaining of subspace feature relevance based on the clustering results of the triangular mesh face quantities and the subspace type complexity includes: According to the calculation steps of subspace type complexity, the type complexity of each cluster cluster obtained by triangular mesh surface vector clustering is obtained; the absolute value of the difference between the type complexity of each cluster cluster and the type complexity of the subspace in which it is located is calculated; and the ratio of the absolute value in the cluster cluster to the type complexity of each cluster cluster is accumulated as the subspace feature correlation.

9. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The subspace complexity is obtained according to the subspace structure complexity, subspace type complexity and subspace feature correlation, specifically: The product of the subspace structure complexity and the subspace type complexity is calculated; and the ratio of the product to the subspace feature correlation is used as the subspace complexity.

10. The three-dimensional model data processing system based on augmented reality according to claim 1, characterized in that: The subspace is divided according to the subspace complexity, specifically: Set a complexity threshold; when the subspace complexity is greater than the complexity threshold, use octree to divide the subspace.

Images