Three-dimensional model surface reconstruction method

Abstract

The invention belongs to the technical field of image processing, and discloses a three-dimensional model surface reconstruction method, which comprises the following steps of: constructing a general cube reconstruction index table which comprises all combinations of six patches of a cube participating in reconstruction and corresponding patch reconstruction masks, vertex reconstruction masks and vertex indexes; the patch reconstruction mask is a six-bit binary number, and the patch corresponding to the bit with the value of 1 participates in reconstruction; the vertex reconstruction mask is an eight-bit binary number, and the vertex corresponding to the bit with the value of 1 participates in reconstruction; obtaining three-dimensional volume data, wherein each voxel has a three-dimensional index and a binary state; acquiring a patch reconstruction mask by a binary state of adjacent voxels, wherein the patch reconstruction mask is a reconstructed effective voxel when the patch reconstruction mask is greater than 0; obtaining a reconstruction effective voxel set and a patch reconstruction mask set, and obtaining a patch array set and a vertex coordinate set participating in reconstruction according to the reconstruction effective voxel set and the patch reconstruction mask set and the general cube reconstruction index table; the patch array is a global index of four vertexes forming the reconstructed patch; and constructing a reconstructed surface model. And the generation efficiency and speed of the dynamic surface model are improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of image processing, and particularly relates to a three-dimensional model surface reconstruction method. BACKGROUND

[0002] In many fields such as industrial detection, medical imaging and scientific computing, volumetric data has become an indispensable core data form. Industrial computed tomography (Industrial CT) can obtain high-precision three-dimensional structure information of the interior of an object without damage; medical CT, MRI and other imaging technologies generate detailed three-dimensional data describing human organs, tissues and lesions. These volumetric data are essentially three-dimensional regular grids containing scalar properties such as density and material. How to efficiently and accurately reconstruct a surface model from the volumetric data for visualization and interactive operation is a crucial and challenging issue.

[0003] Although traditional three-dimensional surface reconstruction algorithms (such as the classic Marching Cubes algorithm) are general, they often face efficiency bottlenecks when dealing with massive volumetric data, and it is difficult to meet the stringent performance requirements of real-time interactive applications. In particular, in real-time surgery simulation, industrial part online detection, virtual sculpting and editing, etc., users not only need to observe static models, but also need to perform dynamic and visual interactive operations (such as cutting, painting, deformation, Boolean operations, etc.). These operations require the system to immediately feedback the modification results, which puts extremely high requirements on the speed and flexibility of the underlying reconstruction algorithm. SUMMARY

[0004] The present application provides a three-dimensional model surface reconstruction method, which uses a pre-defined topological structure to generate a voxel cube reconstruction index table, and uses a lookup table to map the neighborhood state of a voxel to the surface patch set it should generate, so that the patch facing the external space is visible to the reconstructed surface, simplifying the reconstruction process and improving the generation efficiency and speed of the dynamic surface model.

[0005] To solve the above technical problems, the present application uses the following technical solutions:

[0006] A three-dimensional model surface reconstruction method, comprising the following steps.

[0007] A general cube reconstruction index table is constructed, which includes all combinations of six cube patches participating in reconstruction and corresponding patch reconstruction masks, vertex reconstruction masks and vertex indexes; the patch reconstruction mask is a six-bit binary number, and the bit with a value of 1 represents that the corresponding patch participates in reconstruction; the vertex reconstruction mask is an eight-bit binary number, and the bit with a value of 1 represents that the corresponding vertex participates in reconstruction;

[0008] acquiring three-dimensional volume data, which comprises a plurality of voxels; each of the voxels has a determined three-dimensional index, a binary state and a spatial size; the binary state comprises an active state and an inactive state;

[0009] acquiring, according to the binary states of the adjacent voxels of each of the voxels in the active state, a face patch reconstruction mask of the voxel, wherein when the value of the face patch reconstruction mask is greater than 0, the voxel is a reconstruction effective voxel; obtaining a reconstruction effective voxel set C and a corresponding face patch reconstruction mask set B;

[0010] obtaining, according to the reconstruction effective voxel set C, the face patch reconstruction mask set B and a general cubic reconstruction index table, a face patch array set participating in reconstruction and a vertex coordinate set; the face patch array is a global index of four vertices constituting a reconstructed face patch;

[0011] constructing a reconstructed surface model according to the face patch array set and the vertex coordinate set.

[0012] In some specific embodiments, when the reconstruction effective voxels are acquired, the three-dimensional volume data is divided into a plurality of spatial regions, and the reconstruction effective voxels of different spatial regions are acquired in parallel by using multi-threading.

[0013] In some specific embodiments, the adjacent voxels of each of the voxels are achieved by using a global index of each of the voxels, comprising:

[0014] The three-dimensional volume data has a determined three-dimensional dimension ; The three-dimensional index of each of the voxels is ;

[0015] The global index of each of the voxels is: ;

[0016] The global indexes of the adjacent voxels of each of the voxels are respectively: 、 、 、 、 、 .

[0017] In some specific embodiments, in the general cubic reconstruction index table, the vertex index participating in reconstruction is represented by a three-dimensional array; the first dimension is a face patch combination case serial number; the second dimension is a face patch serial number under the face patch combination case serial number; and the third dimension is a vertex serial number corresponding to the face patch serial number.

[0018] In some specific embodiments, the vertex coordinates of each of the effective voxels are obtained according to the three-dimensional index and the spatial size of the voxel, comprising:

[0019] According to the three-dimensional index of the voxel and a transformation matrix T of discrete index coordinates to world coordinates to obtain the center point coordinates of the voxel ;

[0020] According to the spatial size of the voxel and the center point coordinates obtain the coordinates of the eight vertices of the voxel respectively 、 、 、 、 、 、 、 ;

[0021] wherein,

[0022] is a voxel vertex coordinate function, whose parameters are the voxel vertex serial number;

[0023] , , , , , .

[0024] In some specific embodiments, according to the reconstructed effective voxel set, the patch reconstruction mask set and the general cubic reconstruction index table, the set of vertex coordinates participating in reconstruction includes:

[0025] re-indexing each of the reconstructed effective voxels in the reconstructed effective voxel set ;

[0026] voxel The number of vertices participating in the outer surface reconstruction is , and the coordinates of each vertex are calculated by the following formula:

[0027] ;

[0028] wherein, , are the numerical values (0 or 1) of each bit of the binary form of , the function converts the local vertex index of the voxel into the world coordinates of the vertex;

[0029] The set of vertex coordinates is:

[0030] .

[0031] In some specific embodiments, obtaining the set of arrays of patches involved in reconstruction from the set of reconstructed active voxels, the set of patch reconstruction masks and the general cube reconstruction index table comprises:

[0032] Obtaining patch offsets of each of the reconstructed active voxels and vertex offsets :

[0033] , ;

[0034] wherein, is a function of the number of patches involved in reconstruction of the reconstructed active voxel, parameterized by the patch reconstruction mask; is a function of the number of vertices involved in reconstruction of the reconstructed active voxel, parameterized by the vertex reconstruction mask;

[0035] Obtaining the set of arrays of patches:

[0036] ;

[0037] wherein, ; ; N is the number of the reconstructed active voxels; is a function of the serial number of each vertex involved in reconstruction of the reconstructed active voxel, parameterized by the three-dimensional array; , is a vertex global index function.

[0038] In some specific embodiments, the construction of the array of patches and the obtaining of the vertex coordinates are performed using multi-threading.

[0039] In some specific embodiments, an absolute tolerance is set, and vertices with a distance less than are merged, and the reconstructed surface model is optimized.

[0040] In some specific embodiments, all quadrilateral patches in the reconstructed surface model are decomposed into triangular patches to obtain a pure triangular mesh surface model;

[0041] A smoothing filter based on a windowed Sinc function is used to smooth the pure triangular mesh surface model to obtain a final smooth mesh model.

[0042] Compared with the prior art, the three-dimensional model surface reconstruction method has the advantages and positive effects that the three-dimensional model surface reconstruction method adopts the combination of the six face sheets of the cube for reconstruction to predefine and generate a general cube reconstruction index table of the corresponding topology structure, and obtains the reconstruction effective voxel by using the adjacency activation mechanism, obtains the face sheet array set and the vertex coordinate set for participating in the reconstruction of the effective voxel by using the fast lookup table mechanism, and is directly used for the reconstruction of the outer surface of the three-dimensional graph, which significantly reduces the calculation amount, simplifies the reconstruction process, improves the calculation efficiency, shortens the calculation time, and effectively solves the technical problems of low calculation efficiency and low three-dimensional surface reconstruction efficiency of the traditional method. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0044] Figure 1 is a three-dimensional model surface reconstruction method flowchart according to the embodiment;

[0045] Figure 2 is a voxel cube schematic diagram according to the embodiment;

[0046] Figure 3 is a schematic diagram of the connection structure of each voxel of the voxel according to the embodiment;

[0047] Figure 4 is a schematic diagram of the arrangement of each vertex number of the voxel according to the embodiment. DETAILED DESCRIPTION

[0048] The embodiments of the present application will be described in detail below, and the examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0049] Referring to Figure 1 , Figure 2 , Figure 3 The present application discloses a three-dimensional model surface reconstruction method, comprising:

[0050] S1, a general cubic reconstruction index table is constructed, which includes all combinations of six face patches of a cubic participating in reconstruction and corresponding face patch reconstruction masks, vertex reconstruction masks and vertex indexes; the face patch reconstruction mask is a six-bit binary number, a bit with a value of 1 represents that the corresponding face patch participates in reconstruction, and a bit with a value of 0 represents that the corresponding face patch does not participate in reconstruction; the vertex reconstruction mask is an eight-bit binary number, a bit with a value of 1 represents that the corresponding vertex participates in reconstruction, and a bit with a value of 0 represents that the corresponding vertex does not participate in reconstruction.

[0051] S2, three-dimensional body data is obtained, which includes a plurality of voxels; each voxel has a determined three-dimensional index, a spatial size and a binary state; the spatial size of each voxel is the same, and of course can be different; the binary state includes an activated state and a non-activated state.

[0052] S3, a face patch reconstruction mask of a voxel in an activated state is obtained according to the binary states of voxels directly adjacent to the voxel in the activated state, and when the value is greater than 0, the voxel is a reconstruction effective voxel; a reconstruction effective voxel set C and a corresponding face patch reconstruction mask set B are obtained.

[0053] S4, a face patch array set participating in reconstruction and a vertex coordinate set are obtained according to the reconstruction effective voxel set C, the face patch reconstruction mask set B and the general cubic reconstruction index table; the face patch array is the global index of four vertices constituting a reconstruction face patch.

[0054] S5, a reconstruction surface model is constructed according to the face patch array set and the vertex coordinate set.

[0055] S1 is constructed once and applied to the process of surface reconstruction of each three-dimensional model in three-dimensional image updating; S2 to S5 are repeatedly performed when the three-dimensional image is updated.

[0056] The three-dimensional model surface reconstruction method of the application adopts a general cubic reconstruction index table for predefining combinations of six face patches of a cubic participating in reconstruction and generating corresponding topological structures, and uses an adjacent activation mechanism to obtain reconstruction effective voxels, uses a fast lookup mechanism to obtain a face patch array set participating in reconstruction of the reconstruction effective voxels and a vertex coordinate set, and is directly used for reconstruction of an outer surface of a three-dimensional figure, thereby significantly reducing the amount of calculation, simplifying the reconstruction process, improving the calculation efficiency, shortening the calculation time and effectively solving the technical problems of low calculation efficiency and low three-dimensional surface reconstruction efficiency of traditional methods.

[0057] The specific process and principles of the three-dimensional model surface reconstruction method of the application will be described in detail below through specific embodiments.

[0058] In some specific embodiments, with reference to Figure 1 , Figure 2 , Figure 3 , the three-dimensional body data has a determined three-dimensional dimension ; 、 、 are the number of voxels in the x-axis direction, y-axis direction, and z-axis direction respectively; each voxel has a determined three-dimensional index ; ; ; .

[0059] The global index of a voxel is defined as:

[0060] ;

[0061] is the global index value of the voxel.

[0062] Each vertex of a voxel has a three-dimensional index ; the global index of the vertex of the voxel is defined as:

[0063] ;

[0064] is the global index value of the vertex of the voxel.

[0065] The global index values of the directly adjacent voxels of a voxel are respectively , 、 、 、 、 、 which are obtained by substituting , , into the global index definition of the voxel.

[0066] The positions of the adjacent voxels corresponding to each global index value relative to the voxel are shown in Table 1 below.

[0067] Table 1 Global linear index of adjacent voxels

[0068] The three-dimensional model surface reconstruction method of the present embodiment simplifies the retrieval of the data of the 1-dimensional continuous space storage of the three-dimensional body data in the memory by defining the global index of the voxel and obtaining the global index of each directly adjacent voxel through the global index, obtaining the data of each voxel stored continuously in the memory through the global index, such as the binary state of each adjacent voxel, improving the speed of obtaining the information of the directly adjacent voxels, and further improving the efficiency of obtaining the effective voxel set for reconstruction.

[0069] In some specific embodiments, the global index of a voxel is defined as: Figure 2The face reconstruction mask of the voxel in the active state is calculated to determine whether it is a reconstruction effective voxel, i.e., whether the voxel participates in the reconstruction of the three-dimensional surface model.

[0070] The face reconstruction mask is a six-bit binary number, and from low to high, it represents the x+ (right), x- (left), y+ (back), y- (front), z+ (up) and z- (down) direction face, and the face participating in the reconstruction corresponds to bit 1, otherwise 0. That is, the bit with a value of 1 represents that the corresponding face participates in the reconstruction; and when the voxel adjacent to the face participating in the reconstruction is a voxel in the non-active state, the corresponding face is a face on the three-dimensional surface, and the bit of the face mask corresponding to the face is 1.

[0071] Referring to Figure 4 from low to high, it represents the 0-7th vertex of the cube, and similarly, the vertex reconstruction mask is an eight-bit binary number, and the bit with a value of 1 represents that the corresponding vertex participates in the reconstruction; when the face of the vertex participates in the reconstruction, the bits of the vertex reconstruction masks corresponding to the four vertices are 1.

[0072] The face reconstruction mask b of the voxel is calculated as follows:

[0073] ;

[0074] wherein, is the Iverson bracket, which takes a value of 1 when the condition is true, and 0 otherwise;

[0075] that is, the directly adjacent voxels of the voxel in six directions, represents the adjacent voxel in the dth direction;

[0076] is the binary state of the voxel, and the active state is 1 and the non-active state is 0.

[0077] The set of voxels participating in the three-dimensional surface reconstruction is the reconstruction effective voxel set The corresponding face reconstruction mask set is , is the number of reconstruction effective voxels participating in the outer surface reconstruction.

[0078] In some specific embodiments, when the reconstruction effective voxels are obtained, the space of the three-dimensional body data is divided into multiple spatial regions, and different spatial regions are processed in multiple threads in parallel. The utilization rate of computing resources is effectively improved, the computing efficiency is greatly improved, the computing time is shortened, and the technical problems of low computing efficiency and long time of traditional methods are effectively solved.

[0079] In some specific embodiments, each voxel in the reconstruction effective voxel set is re-indexed globally , whose value is 0, 1, 2, …, N-1; N is the number of reconstructed active voxels in the reconstructed active voxel set. The reconstructed active voxels participating in the outer surface reconstruction are The mapping of individual voxels is established:

[0080] ;

[0081] The global index positions of the reconstructed patches and the reconstructed vertices of each reconstructed active voxel are calculated; that is, the position or number corresponding to the first reconstructed patch of the reconstructed active voxel, the position or number corresponding to the first reconstructed vertex of the reconstructed active voxel, which is specifically:

[0082] , ;

[0083] In the formula, is the reconstructed active voxel number, whose value is 0, 1, 2, …, N-1; and respectively represent the patch number and vertex number functions of the voxels participating in the reconstruction, and the parameters are respectively the patch mask and the vertex mask.

[0084] and The values of and can be obtained by adding the number of patches and vertices participating in the reconstruction in the general cubic reconstruction index table, and then looking up the table according to the patch mask of the reconstructed active voxel corresponding to the patch mask.

[0085] The total number of final reconstructed patches and the total number of reconstructed vertices are respectively:

[0086] , ;

[0087] Similarly, and respectively represent the reconstructed patch number and the reconstructed vertex number functions of the voxels corresponding to the patch mask , and the parameters are respectively the patch mask and the vertex mask of the reconstructed active voxel.

[0088] Similarly, and The values of and can be obtained by adding the number of patches and vertices participating in the reconstruction in the general cubic reconstruction index table, and then looking up the table according to the patch mask of the reconstructed active voxel corresponding to the patch mask.

[0089] The position or number corresponding to the first reconstructed vertex of the reconstructed active voxel is used to obtain the global index of the vertex; the total number of final reconstructed patches and the total number of reconstructed vertices Reserving storage space for storing a reconstructed surface model.

[0090] In some specific embodiments, the local index of each reconstructed vertex of a reconstructed active voxel in the voxel is represented by a three-dimensional array; the first dimension is the face combination case number, whose value is 0-63; the second dimension is the face number under the corresponding face combination case number, whose value is 0-6; and the third dimension is the vertex number corresponding to the face number, whose value is 0-7.

[0091] Then, the global index of the reconstructed vertex of the reconstructed active voxel is:

[0092] ;

[0093] wherein, ; ; N is the number of reconstructed active voxels; is a sequence function of each vertex participating in the reconstruction of the reconstructed active voxel, and the parameter is a three-dimensional array; is a vertex global index function. That is, the conversion function converts the vertex index local to the current voxel into the global vertex index.

[0094] Then, the reconstructed face array set is:

[0095] ;

[0096] wherein, ; ; the three-dimensional surface reconstruction model includes four quadrilateral faces; each face array is composed of 4 vertex indexes.

[0097] In some specific embodiments, referring to Figure 4 , the vertex coordinates of the voxel are obtained according to the three-dimensional index of the voxel, specifically:

[0098] According to the three-dimensional index of the voxel and the transformation matrix from the discrete index coordinates to the world coordinates , the center point coordinates of the voxel are obtained ;

[0099] According to the spatial size of the voxel and the point coordinates C, the coordinates of the eight vertices of the voxel are respectively , , , , , , , ;

[0100] wherein,

[0101] is the vertex coordinate function of a voxel, whose parameter is the vertex serial number of the voxel;

[0102] is the x-axis coordinate of the voxel center point offset by half unit length in the positive direction of the x-axis; is the x-axis coordinate of the voxel center point offset by half unit length in the negative direction of the x-axis; is the y-axis coordinate of the voxel center point offset by half unit length in the positive direction of the y-axis; is the y-axis coordinate of the voxel center point offset by half unit length in the negative direction of the y-axis; is the z-axis coordinate of the voxel center point offset by half unit length in the positive direction of the z-axis; is the z-axis coordinate of the voxel center point offset by half unit length in the negative direction of the z-axis. The mapping relationship of each vertex coordinate of a specific voxel is shown in Table 2 below, which is the coordinate in the world coordinate system. Table 2 Vertex coordinate expression

[0103] In some specific embodiments, the reconstructed vertex coordinate set of the effective voxel is:

[0104]

[0105] In some specific embodiments, the reconstructed vertex coordinate set of the effective voxel is:

[0106]

[0107] The number of vertices of the voxel participating in the outer surface reconstruction is , and the coordinate of each vertex is calculated by the formula:

[0108]

[0109] In the formula, , is the numerical value (0 or 1) of each bit of the binary form of , i.e. ; the function converts the local vertex index of the voxel into the world coordinate of the vertex. ​​​​​​​​​​​​​

[0110] In some specific embodiments, the specific steps of obtaining the vertex coordinate set and the patch array set are as follows:

[0111] Parallelly ;

[0112] Obtaining the voxel ;

[0113] Calculating the center coordinate ;

[0114] Generating 8 vertex coordinates ;

[0115] Reconstructing the mask according to the vertex Obtaining the vertex to be used; converting the local vertex index used into the global vertex index; According to the vertex index list

[0116] Generating the patch array.

[0117] That is, the obtaining of the reconstructed vertex coordinates and the construction of the patch array are both executed in multi-thread parallel mode, so as to improve the reconstruction efficiency.

[0118] In some specific embodiments, the absolute tolerance is set as ; the distance between each reconstructed vertex is obtained; and the vertices with a distance less than are merged, so as to optimize the three-dimensional surface model.

[0119] In some specific embodiments, all quadrilateral patches in the three-dimensional surface model are decomposed into triangular patches, so as to obtain a pure triangular mesh surface model.

[0120] The pure triangular mesh surface model is smoothed by using a smoothing filter based on a window Sinc function, so as to obtain a final smooth mesh model.

[0121] In the description of the present application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", etc. indicate the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0122] ​In addition, the terms "first", "second", etc. are used only for the purpose of description and do not imply or imply relative importance or imply the number of the technical features indicated. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise explicitly specified.

[0123] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements, unless otherwise explicitly limited. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0124] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. Moreover, the first feature "above", "above" and "above" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "below" and "below" the second feature can be directly below or obliquely below the first feature, or only indicate that the horizontal height of the first feature is less than that of the second feature.

[0125] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms is not necessarily for the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or features of different embodiments or examples described in the present application without contradiction.

[0126] Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and cannot be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A method for reconstructing the surface of a three-dimensional model, characterized in that, include: A general cube reconstruction index table is constructed, which includes all combinations of the six faces of the cube participating in reconstruction and the corresponding face reconstruction mask, vertex reconstruction mask, and vertex index; the face reconstruction mask is a six-bit binary number, where a bit with a value of 1 represents the corresponding face participating in reconstruction; the vertex reconstruction mask is an eight-bit binary number, where a bit with a value of 1 represents the corresponding vertex participating in reconstruction. Acquire three-dimensional volume data, which includes multiple voxels; each voxel has a defined three-dimensional index, binary state, and spatial size; the binary state includes an active state and an inactive state; The patch reconstruction mask is obtained based on the binary state of the adjacent voxels in the active state. When the value is greater than 0, the voxel is a valid voxel for reconstruction. The set of valid voxels for reconstruction C and the corresponding set of patch reconstruction masks B are obtained. The set of effective voxels for reconstruction C, the set of face reconstruction masks B, and the general cube reconstruction index table are used to obtain the set of face arrays and the set of vertex coordinates for reconstruction; the face array is the global index of the four vertices that make up the reconstructed face. The reconstructed surface model is constructed based on the set of face arrays and the set of vertex coordinates.

2. The three-dimensional model surface reconstruction method according to claim 1, characterized in that, When acquiring the reconstructed effective voxels, the space of the three-dimensional volume data is divided into multiple spatial regions, and the reconstructed effective voxels of different spatial regions are acquired in parallel using multi-threading.

3. The three-dimensional model surface reconstruction method according to claim 1, characterized in that, The adjacent voxels are implemented through a global index of each voxel, including: The three-dimensional volume data has a defined three-dimensional dimension. The three-dimensional index of each voxel is: ; The global index of the voxel is: ; The global indices of the adjacent voxels are as follows: , , , , , .

4. The three-dimensional model surface reconstruction method according to claim 3, characterized in that, In the general cube reconstruction index table, the vertex indices participating in the reconstruction are represented by a three-dimensional array; the first dimension is the sequence number of the face combination; the second dimension is the face number under the face combination sequence number; and the third dimension is the vertex number corresponding to the face number.

5. The three-dimensional model surface reconstruction method according to claim 4, characterized in that, The vertex coordinates of each effective voxel are obtained based on the voxel's 3D index and spatial dimensions, including: According to the three-dimensional index of the voxel and the transformation matrix from discrete index coordinates to world coordinates Obtain the center point coordinates of the voxel ; ; Based on the spatial dimensions of the voxel and the coordinates of the center point The coordinates of the eight vertices of the voxel are obtained as follows: , , , , , , , ; in, This is a function for the coordinates of voxel vertices, whose parameter is the voxel vertex index; , , , , , 。 6. The three-dimensional model surface reconstruction method according to claim 5, characterized in that, The set of vertex coordinates for reconstruction, obtained based on the set of effective voxels, the set of face reconstruction masks, and the general cube reconstruction index table, includes: Each of the reconstructed effective voxels in the reconstructed effective voxel set is renumbered and indexed. ; voxels The number of vertices involved in the outer surface reconstruction is Each vertex The coordinates are calculated using the following formula: ； In the formula, , for The binary representation of each bit (0 or 1), the function voxels The local vertex index is converted to the world coordinates of that vertex; The set of vertex coordinates is: 。 7. The three-dimensional model surface reconstruction method according to claim 6, characterized in that, The set of facet arrays participating in reconstruction, obtained based on the reconstructed effective voxel set, the facet reconstruction mask set, and the general cube reconstruction index table, includes: Obtain the patch offset of each of the reconstructed effective voxels. and vertex offset : ， ； in, The function is the number of patches involved in the reconstruction of the effective voxels, and its parameter is the patch reconstruction mask; The function is the number of vertices participating in the reconstruction of the effective voxels, and its parameter is the vertex reconstruction mask; Obtain the set of face arrays: ； in, ; N represents the number of effective voxels in the reconstruction. The function is the index of each vertex participating in the reconstruction of the effective voxel, with the parameter being the three-dimensional array; , This is the global indexing function for vertices.

8. The three-dimensional model surface reconstruction method according to claim 7, characterized in that, The construction of the face array and the acquisition of vertex coordinates are both performed using multi-threading.

9. The method for reconstructing the surface of a three-dimensional model according to any one of claims 1 to 8, characterized in that, Setting absolute tolerance Merging distance is less than The vertices are used to optimize the reconstructed surface model.

10. The three-dimensional model surface reconstruction method according to claim 9, characterized in that, All quadrilateral patches in the reconstructed surface model are decomposed into triangular patches to obtain a pure triangular mesh surface model. The pure triangular mesh surface model is smoothed using a smoothing filter based on the window Sinc function to obtain the final smooth mesh model.

Images