A tooth partitioning method and apparatus based on collision detection

Abstract

The application discloses a tooth partitioning method and device based on collision detection, and the method comprises the following steps: introducing a tooth model, and setting an edge line of a region based on the tooth model; performing 2D triangulation processing based on the edge line of the region to obtain a region blocking grid; performing calculation based on the region blocking grid by using a ray intersection method to obtain a region internal seed point; performing region diffusion collision processing based on the region internal seed point to obtain an internal region patch; performing neighborhood filling processing on a missing region of the internal region patch to obtain an internal region patch after neighborhood filling processing; obtaining an outer patch based on the internal region patch after neighborhood filling processing by using a maximum connected region searching method, and outputting a tooth partitioning result based on the outer patch and the internal region patch after neighborhood filling processing. By using the method and device, the failure of the region partitioning result caused by the existence of an undesirable topological problem of the model can be avoided, and the accuracy of tooth segmentation is improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and apparatus for tooth partitioning based on collision detection. Background Technology

[0002] In the digital fabrication of teeth, it is necessary to determine the processing direction of the material. Dividing the tooth into defined regions allows for the calculation of the processing direction for specific areas. Current region segmentation methods are based on topology. However, during the layout design process in tooth typesetting software, the quality of the input model cannot be guaranteed; some models may have topological issues, such as subtle genus discrepancies. This leads to errors when applying topology methods, and the region search results spread throughout the entire model. Therefore, to address these issues, a collision detection method is proposed for tooth region segmentation. This method is independent of topology and avoids the failure of region segmentation results when the model has suboptimal topological issues. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a tooth partitioning method and device based on collision detection. By using the collision detection method to partition tooth regions, it can avoid the failure of region partitioning results when the model has undesirable topological problems, thereby improving the accuracy of tooth segmentation.

[0004] To address the aforementioned technical problems, embodiments of the present invention provide a tooth partitioning method based on collision detection, the method comprising:

[0005] Import the tooth model and set the edge lines of the region based on the tooth model;

[0006] 2D triangulation is performed based on the edge lines of the region to obtain the region baffle mesh;

[0007] Seed points within the region are obtained by using the ray intersection method based on the region's baffle mesh.

[0008] Based on the seed points within the region, a region diffusion collision process is performed to obtain an internal region patch.

[0009] The missing areas of the internal region patch are filled with neighborhoods to obtain the internal region patch after neighborhood filling.

[0010] The outer surface area is obtained using the method of finding the maximum connected region based on the inner region area after neighborhood filling, and the tooth partitioning result is output based on the outer surface area and the inner region area after neighborhood filling.

[0011] Optionally, importing the tooth model and setting the edge line of the region based on the tooth model includes:

[0012] Import the tooth model and select the position information of the edge lines based on the tooth model;

[0013] The edge line of the region is set based on the position information of the edge line.

[0014] Optionally, the step of performing 2D triangulation based on the edge lines of the region to obtain the region baffle mesh includes:

[0015] Centroid points are calculated based on the edge lines of the region to obtain a central triangle, and the central triangle is set as the initial triangular mesh. The edge lines of the region include discrete points of the contour lines.

[0016] Insert the discrete points of the contour line into the initial triangular mesh point by point to obtain a new triangular mesh;

[0017] The new triangular mesh is optimized based on the Delaunay optimization criterion to obtain the regional baffle mesh.

[0018] Optionally, the step of using the ray intersection method to calculate and obtain seed points within the region based on the region baffle mesh includes:

[0019] The region baffle mesh is searched to find the first triangle with the largest area in the region baffle mesh, and the center of the first triangle is calculated.

[0020] The intersection points of the regions are obtained by using the ray intersection method based on the center of the first triangle;

[0021] Seed points within the region are obtained based on the midpoint between the center of the first triangle and the intersection of the regions.

[0022] Optionally, the step of calculating the intersection points of the regions using the ray intersection method based on the center of the first triangle includes:

[0023] The center of the first triangle is selected as the starting point, and the opposite direction of tooth insertion is selected as the ray direction;

[0024] Based on the starting point and the direction of the ray, the position where the ray intersects with the internal region is determined, and the intersection point of the regions is obtained.

[0025] Optionally, the step of performing region diffusion collision processing based on seed points within the region to obtain internal region patches includes:

[0026] Select a seed point inside the region as the center of the sphere, and set the radius of the sphere based on the side length of the tooth model;

[0027] A sphere is generated within the region based on the sphere's center and radius.

[0028] Set the single diffusion step size and diffusion direction;

[0029] Based on the single diffusion step size and the diffusion direction, the sphere inside the region is subjected to regional diffusion collision processing to obtain an internal region patch.

[0030] Optionally, the step of performing neighborhood filling processing on the missing areas of the internal region patch to obtain the neighborhood-filled internal region patch includes:

[0031] Neighborhood features of internal region patches are extracted based on feature extraction methods;

[0032] Based on the internal region patch and the neighborhood features, the missing regions of the internal region patch are filled with neighborhood features to obtain the internal region patch after neighborhood filling.

[0033] Optionally, the inner region patch after neighborhood filling is used to obtain the outer surface patch using the maximum connected region search method, and the tooth partitioning result is output based on the outer surface patch and the inner region patch after neighborhood filling, including:

[0034] Return the internal region patches after neighborhood filling and set them to a disconnected state;

[0035] The outer layer mesh is obtained by finding the maximum connected region based on the disconnected state of the internal region mesh after the neighborhood filling process.

[0036] The tooth partitioning results are output based on the outer surface patch and the inner region patch after neighborhood filling processing.

[0037] Optionally, the disconnected state of the internal region patches after neighborhood filling is used to obtain the outer layer patches using the method of finding the maximum connected region, including:

[0038] Based on the disconnected state of the internal region patches after neighborhood filling, a grayscale information similarity algorithm is used to obtain connected regions.

[0039] The connectivity of the connected regions is analyzed to obtain the connectivity of the region pixels.

[0040] Region filtering is performed based on the connectivity of the pixels in the region to obtain the outer layer sheet.

[0041] In addition, embodiments of the present invention also provide a tooth partitioning device based on collision detection, the device comprising:

[0042] Edge Line Module: Imports a tooth model and sets the edge lines of the region based on the tooth model;

[0043] Regional baffle mesh module: Based on the edge lines of the region, 2D triangulation is performed to obtain the regional baffle mesh;

[0044] Seed point module within the region: Based on the region's baffle mesh, the ray intersection method is used to calculate and obtain the seed points within the region;

[0045] Internal region patch module: Based on the seed points inside the region, perform region diffusion collision processing to obtain internal region patches;

[0046] Neighborhood filling module: performs neighborhood filling processing on the missing areas of the internal region patch to obtain the internal region patch after neighborhood filling processing;

[0047] Outer surface slice and result module: Based on the inner region slice after neighborhood filling, the outer surface slice is obtained using the maximum connected region search method, and the tooth partitioning result is output based on the outer surface slice and the inner region slice after neighborhood filling.

[0048] In this embodiment of the invention, the collision detection method is used to divide the tooth region. This method is not dependent on the topology and can avoid the failure of the region division result when the model has an undesirable topological problem. At the same time, the segmented region can be obtained without inputting the type of boundary line. It can also achieve relatively ideal results in terms of efficiency for different tooth models. Furthermore, the collision detection method not only improves the accuracy and efficiency of tooth segmentation, but also avoids the problems caused by the inconsistent quality of imported models, improves the reliability of the segmentation results, reduces the pressure on the staff for subsequent processing of the tooth partitioning results, and enables more accurate subsequent tooth layout design. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart illustrating the tooth partitioning method based on collision detection in an embodiment of the present invention.

[0051] Figure 2 This is a schematic diagram of the structural composition of the tooth partitioning device based on collision detection in an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Example 1

[0054] Please see Figure 1 , Figure 1 This is a schematic flowchart of the tooth partitioning method based on collision detection in an embodiment of the present invention.

[0055] like Figure 1 As shown, a tooth partitioning method based on collision detection is described, the method comprising:

[0056] S11: Import the tooth model and set the edge line of the region based on the tooth model;

[0057] In a specific implementation of the present invention, the step of importing a tooth model and setting the edge line of the region based on the tooth model includes: importing a tooth model and selecting the position information of the edge line based on the tooth model; and setting the edge line of the region based on the position information of the edge line.

[0058] Specifically, first import the tooth model that needs to be partitioned, select the position of the edge line in the tooth model, which exists at the boundary of the hole area and the bottom area of ​​the tooth model; after obtaining the position of the edge line, the user specifies the edge line of the area, without needing to set the type of the boundary line.

[0059] S12: Perform 2D triangulation based on the edge lines of the region to obtain the region baffle mesh;

[0060] In a specific implementation of this invention, the step of performing 2D triangulation based on the edge lines of the region to obtain a region baffle mesh includes: calculating the centroid point based on the edge lines of the region to obtain a central triangle, and setting the central triangle as an initial triangular mesh, wherein the edge lines of the region include discrete points of the contour lines; inserting the discrete points of the contour lines point by point into the initial triangular mesh to obtain a new triangular mesh; and optimizing the new triangular mesh based on the Delaunay optimization criterion to obtain the region baffle mesh.

[0061] Specifically, the process involves inputting edge lines, which are discrete points on the contour line. Centroid points are calculated based on these discrete points. A central triangle is constructed using the three construction points closest to the centroid point, and this central triangle is set as the initial triangular mesh. Discrete points are then inserted into the initial triangular mesh point by point. A discrete point is selected from the set, and a triangle containing that point is found. This point is connected to all vertices of the triangle to generate a new triangular mesh. The edges of the new triangular mesh are added to the optimization queue. The new triangular mesh is then optimized based on the Delaunay optimization criterion. The optimization process involves processing the optimization queue after each point is added, until the optimization queue is empty, before adding the next data point. Once all data points have been added to the triangular mesh, redundant triangles are deleted, resulting in a mesh that conforms to the Delaunay optimization criterion. This yields a region-blocking mesh, which is used to prevent spheres inside a region from spreading outwards during detection in a specific region.

[0062] S13: Based on the region baffle mesh, the ray intersection method is used to calculate and obtain the seed points inside the region;

[0063] In a specific implementation of this invention, the step of using the ray intersection method to calculate and obtain seed points within the region based on the region baffle mesh includes: performing a search process on the region baffle mesh to find the first triangle with the largest area in the region baffle mesh and calculating the center of the first triangle; using the ray intersection method based on the center of the first triangle to calculate and obtain the region intersection point; and obtaining the seed point within the region based on the midpoint between the center of the first triangle and the region intersection point.

[0064] Furthermore, the step of calculating the intersection point of the region using the ray intersection method based on the center of the first triangle includes: selecting the center of the first triangle as the starting point and selecting the opposite direction of tooth insertion as the ray direction; emitting a ray based on the starting point and the ray direction, determining the position where the ray intersects with the internal region, and obtaining the region intersection point.

[0065] Specifically, the algorithm finds the triangle with the largest area in the region's baffle mesh. It first defines an initial maximum area, then iteratively compares each triangle; if the triangle's area is greater than the initial maximum area, the maximum area is updated to the triangle's area, and the comparison continues until the loop ends. Once the first triangle with the largest area in the region's baffle mesh is found, its center is calculated. The ray intersection method is then used to calculate the region's intersection point. Specifically, the center of the triangle is selected as the starting point of the ray, and the opposite direction of the tooth insertion is chosen as the ray direction. Rays are emitted based on the starting point and ray direction until they intersect the region. The intersection point determines whether the baffle mesh is the top or bottom region, thus obtaining the region's intersection point. Finally, the midpoint between the region's intersection point and the center of the first triangle is taken as the seed point within the region.

[0066] S14: Perform region diffusion collision processing based on the seed points inside the region to obtain the internal region patch;

[0067] In a specific implementation of the present invention, the step of performing regional diffusion collision processing based on the seed point inside the region to obtain an internal region patch includes: selecting the seed point inside the region as the center of a sphere and setting the sphere radius based on the side length of the tooth model; generating an internal sphere based on the center of the sphere and the sphere radius; setting a single diffusion step size and diffusion direction; and performing regional diffusion collision processing on the internal sphere based on the single diffusion step size and the diffusion direction to obtain an internal region patch.

[0068] Specifically, the seed point inside the region is used as the center of the sphere, and the radius of the sphere is set to 0.2mm according to the side length of the tooth model. The sphere inside the region for diffusion collision processing is generated according to the set center and radius. The single diffusion step size is set, and the diffusion direction is set to spread along the positive and negative x, y, and z directions, i.e., six directions. Since both accuracy and efficiency need to be considered in the collision partitioning process, when the accuracy is high, the radius of the sphere is small, which will increase the number of collision executions and cause the time to be too long. Conversely, when the accuracy is low, the radius of the sphere is large, the number of collision executions is reduced, but more scattered areas will not be selected. The accuracy is selected and set according to the specific situation. In this invention, low-precision collision can be selected. The sphere performs diffusion collision processing according to the single diffusion step size and diffusion direction. When the sphere collides with the tooth model, the diffusion stops, and the collided surface is recorded, which is the surface of the internal region.

[0069] S15: Perform neighborhood filling processing on the missing areas of the internal region patch to obtain the internal region patch after neighborhood filling processing;

[0070] In a specific implementation of the present invention, the step of performing neighborhood filling processing on the missing regions of the internal region patch to obtain the neighborhood-filled internal region patch includes: extracting neighborhood features of the internal region patch based on a feature extraction method; and performing neighborhood filling processing on the missing regions of the internal region patch based on the internal region patch and the neighborhood features to obtain the neighborhood-filled internal region patch.

[0071] Specifically, when low-precision collision is selected, many scattered areas are not selected. Therefore, post-processing of the collision results is required. First, neighborhood features of the internal region patches are extracted based on feature extraction methods, neighborhood curvature is defined, feature points and non-feature points in the neighborhood are distinguished by normal vector features, feature regions and non-feature regions are identified by Euclidean distance features, and neighborhood feature parameters are constructed. Finally, the curvature feature parameters are used to further extract features based on the neighborhood feature parameters to obtain the neighborhood features of the internal region patches. The missing regions of the internal region patches are selected based on the features of the internal region patches and their neighbors. The missing regions are filled using neighborhood filling processing according to the set filling parameters until there are no missing regions. After filling, the internal region patches after neighborhood filling processing are obtained. Neighborhood filling processing can accurately fill the selected regions and save the time of performing region filling.

[0072] S16: Based on the internal region patches after neighborhood filling, the outer surface patch is obtained using the method of finding the maximum connected region, and the tooth partitioning result is output based on the outer surface patch and the internal region patches after neighborhood filling.

[0073] In a specific implementation of this invention, the process of obtaining an outer surface area using the maximum connected component method based on the inner region area patch after neighborhood filling processing, and outputting tooth partitioning results based on the outer surface area patch and the inner region area patch after neighborhood filling processing, includes: returning the inner region area patch after neighborhood filling processing and setting the inner region area patch after neighborhood filling processing to a disconnected state; obtaining an outer surface area using the maximum connected component method based on the disconnected state of the inner region area patch after neighborhood filling processing; and outputting tooth partitioning results based on the outer surface area patch and the inner region area patch after neighborhood filling processing.

[0074] Furthermore, the method of obtaining the outer layer patch by finding the maximum connected region based on the disconnected state of the internal region patches after neighborhood filling includes: obtaining connected regions by using a grayscale information similarity algorithm based on the disconnected state of the internal region patches after neighborhood filling; performing connectivity analysis on the connected regions to obtain the connectivity of the region pixels; and performing region filtering based on the connectivity of the region pixels to obtain the outer layer patch.

[0075] Specifically, the process begins by retrieving the inner layer patches, since these patches are already defined. The inner region patches are then set to a disconnected state. Based on this disconnected state, the outer layer patches are obtained using the method of finding the maximum connected component. Regions with similar grayscale values ​​are then marked using a grayscale similarity algorithm, thus acquiring all connected components within the same region. For each region, the connectivity between pixels is analyzed to determine which pixels can be connected by paths. Based on the connectivity of the pixels, regions are filtered to find the largest connected component, which yields the outer layer patch. Finally, the outer layer patch and the processed inner region patches are output together to obtain the tooth partitioning result. The output result will vary depending on the imported model; for example, importing a long bridge or crown bridge model can also achieve a relatively ideal tooth partitioning result.

[0076] In this embodiment of the invention, the collision detection method is used to divide the tooth region. This method is not dependent on the topology and can avoid the failure of the region division result when the model has an undesirable topological problem. At the same time, the segmented region can be obtained without inputting the type of boundary line. It can also achieve relatively ideal results in terms of efficiency for different tooth models. Furthermore, the collision detection method not only improves the accuracy and efficiency of tooth segmentation, but also avoids the problems caused by the inconsistent quality of imported models, improves the reliability of the segmentation results, reduces the pressure on the staff for subsequent processing of the tooth partitioning results, and enables more accurate subsequent tooth layout design.

[0077] Example 2

[0078] Please see Figure 2 , Figure 2 This is a schematic diagram of the structural composition of the tooth partitioning device based on collision detection in an embodiment of the present invention.

[0079] like Figure 2 As shown, a tooth partitioning device based on collision detection includes:

[0080] Edge line module 21: Import the tooth model and set the edge line of the region based on the tooth model;

[0081] In a specific implementation of the present invention, the step of importing a tooth model and setting the edge line of the region based on the tooth model includes: importing a tooth model and selecting the position information of the edge line based on the tooth model; and setting the edge line of the region based on the position information of the edge line.

[0082] Specifically, first import the tooth model that needs to be partitioned, select the position of the edge line in the tooth model, which exists at the boundary of the hole area and the bottom area of ​​the tooth model; after obtaining the position of the edge line, the user specifies the edge line of the area, without needing to set the type of the boundary line.

[0083] Region baffle mesh module 22: Performs 2D triangulation processing based on the edge lines of the region to obtain the region baffle mesh;

[0084] In a specific implementation of this invention, the step of performing 2D triangulation based on the edge lines of the region to obtain a region baffle mesh includes: calculating the centroid point based on the edge lines of the region to obtain a central triangle, and setting the central triangle as an initial triangular mesh, wherein the edge lines of the region include discrete points of the contour lines; inserting the discrete points of the contour lines point by point into the initial triangular mesh to obtain a new triangular mesh; and optimizing the new triangular mesh based on the Delaunay optimization criterion to obtain the region baffle mesh.

[0085] Specifically, the process involves inputting edge lines, which are discrete points on the contour line. Centroid points are calculated based on these discrete points. A central triangle is constructed using the three construction points closest to the centroid point, and this central triangle is set as the initial triangular mesh. Discrete points are then inserted into the initial triangular mesh point by point. A discrete point is selected from the set, and a triangle containing that point is found. This point is connected to all vertices of the triangle to generate a new triangular mesh. The edges of the new triangular mesh are added to the optimization queue. The new triangular mesh is then optimized based on the Delaunay optimization criterion. The optimization process involves processing the optimization queue after each point is added, until the optimization queue is empty, before adding the next data point. Once all data points have been added to the triangular mesh, redundant triangles are deleted, resulting in a mesh that conforms to the Delaunay optimization criterion. This yields a region-blocking mesh, which is used to prevent spheres inside a region from spreading outwards during detection in a specific region.

[0086] Region Internal Seed Point Module 23: Based on the region baffle mesh, the ray intersection method is used to calculate and obtain the region internal seed points;

[0087] In a specific implementation of this invention, the step of using the ray intersection method to calculate and obtain seed points within the region based on the region baffle mesh includes: performing a search process on the region baffle mesh to find the first triangle with the largest area in the region baffle mesh and calculating the center of the first triangle; using the ray intersection method based on the center of the first triangle to calculate and obtain the region intersection point; and obtaining the seed point within the region based on the midpoint between the center of the first triangle and the region intersection point.

[0088] Furthermore, the step of calculating the intersection point of the region using the ray intersection method based on the center of the first triangle includes: selecting the center of the first triangle as the starting point and selecting the opposite direction of tooth insertion as the ray direction; emitting a ray based on the starting point and the ray direction, determining the position where the ray intersects with the internal region, and obtaining the region intersection point.

[0089] Specifically, the algorithm finds the triangle with the largest area in the region's baffle mesh. It first defines an initial maximum area, then iteratively compares each triangle; if the triangle's area is greater than the initial maximum area, the maximum area is updated to the triangle's area, and the comparison continues until the loop ends. Once the first triangle with the largest area in the region's baffle mesh is found, its center is calculated. The ray intersection method is then used to calculate the region's intersection point. Specifically, the center of the triangle is selected as the starting point of the ray, and the opposite direction of the tooth insertion is chosen as the ray direction. Rays are emitted based on the starting point and ray direction until they intersect the region. The intersection point determines whether the baffle mesh is the top or bottom region, thus obtaining the region's intersection point. Finally, the midpoint between the region's intersection point and the center of the first triangle is taken as the seed point within the region.

[0090] Internal region patch module 24: Performs region diffusion collision processing based on seed points within the region to obtain internal region patches;

[0091] In a specific implementation of the present invention, the step of performing regional diffusion collision processing based on the seed point inside the region to obtain an internal region patch includes: selecting the seed point inside the region as the center of a sphere and setting the sphere radius based on the side length of the tooth model; generating an internal sphere based on the center of the sphere and the sphere radius; setting a single diffusion step size and diffusion direction; and performing regional diffusion collision processing on the internal sphere based on the single diffusion step size and the diffusion direction to obtain an internal region patch.

[0092] Specifically, the seed point inside the region is used as the center of the sphere, and the radius of the sphere is set to 0.2mm according to the side length of the tooth model. The sphere inside the region for diffusion collision processing is generated according to the set center and radius. The single diffusion step size is set, and the diffusion direction is set to spread along the positive and negative x, y, and z directions, i.e., six directions. Since both accuracy and efficiency need to be considered in the collision partitioning process, when the accuracy is high, the radius of the sphere is small, which will increase the number of collision executions and cause the time to be too long. Conversely, when the accuracy is low, the radius of the sphere is large, the number of collision executions is reduced, but more scattered areas will not be selected. The accuracy is selected and set according to the specific situation. In this invention, low-precision collision can be selected. The sphere performs diffusion collision processing according to the single diffusion step size and diffusion direction. When the sphere collides with the tooth model, the diffusion stops, and the collided surface is recorded, which is the surface of the internal region.

[0093] Neighborhood filling module 25: performs neighborhood filling processing on the missing areas of the internal region patch to obtain the internal region patch after neighborhood filling processing;

[0094] In a specific implementation of the present invention, the step of performing neighborhood filling processing on the missing regions of the internal region patch to obtain the neighborhood-filled internal region patch includes: extracting neighborhood features of the internal region patch based on a feature extraction method; and performing neighborhood filling processing on the missing regions of the internal region patch based on the internal region patch and the neighborhood features to obtain the neighborhood-filled internal region patch.

[0095] Specifically, when low-precision collision is selected, many scattered areas are not selected. Therefore, post-processing of the collision results is required. First, neighborhood features of the internal region patches are extracted based on feature extraction methods, neighborhood curvature is defined, feature points and non-feature points in the neighborhood are distinguished by normal vector features, feature regions and non-feature regions are identified by Euclidean distance features, and neighborhood feature parameters are constructed. Finally, the curvature feature parameters are used to further extract features based on the neighborhood feature parameters to obtain the neighborhood features of the internal region patches. The missing regions of the internal region patches are selected based on the features of the internal region patches and their neighbors. The missing regions are filled using neighborhood filling processing according to the set filling parameters until there are no missing regions. After filling, the internal region patches after neighborhood filling processing are obtained. Neighborhood filling processing can accurately fill the selected regions and save the time of performing region filling.

[0096] Outer surface slice and result module 26: Based on the inner region slice after neighborhood filling, the outer surface slice is obtained using the maximum connected region search method, and the tooth partitioning result is output based on the outer surface slice and the inner region slice after neighborhood filling.

[0097] In a specific implementation of this invention, the process of obtaining an outer surface area using the maximum connected component method based on the inner region area patch after neighborhood filling processing, and outputting tooth partitioning results based on the outer surface area patch and the inner region area patch after neighborhood filling processing, includes: returning the inner region area patch after neighborhood filling processing and setting the inner region area patch after neighborhood filling processing to a disconnected state; obtaining an outer surface area using the maximum connected component method based on the disconnected state of the inner region area patch after neighborhood filling processing; and outputting tooth partitioning results based on the outer surface area patch and the inner region area patch after neighborhood filling processing.

[0098] Furthermore, the method of obtaining the outer layer patch by finding the maximum connected region based on the disconnected state of the internal region patches after neighborhood filling includes: obtaining connected regions by using a grayscale information similarity algorithm based on the disconnected state of the internal region patches after neighborhood filling; performing connectivity analysis on the connected regions to obtain the connectivity of the region pixels; and performing region filtering based on the connectivity of the region pixels to obtain the outer layer patch.

[0099] Specifically, the process begins by retrieving the inner layer patches, since these patches are already defined. The inner region patches are then set to a disconnected state. Based on this disconnected state, the outer layer patches are obtained using the method of finding the maximum connected component. Regions with similar grayscale values ​​are then marked using a grayscale similarity algorithm, thus acquiring all connected components within the same region. For each region, the connectivity between pixels is analyzed to determine which pixels can be connected by paths. Based on the connectivity of the pixels, regions are filtered to find the largest connected component, which yields the outer layer patch. Finally, the outer layer patch and the processed inner region patches are output together to obtain the tooth partitioning result. The output result will vary depending on the imported model; for example, importing a long bridge or crown bridge model can also achieve a relatively ideal tooth partitioning result.

[0100] In this embodiment of the invention, the collision detection method is used to divide the tooth region. This method is not dependent on the topology and can avoid the failure of the region division result when the model has an undesirable topological problem. At the same time, the segmented region can be obtained without inputting the type of boundary line. It can also achieve relatively ideal results in terms of efficiency for different tooth models. Furthermore, the collision detection method not only improves the accuracy and efficiency of tooth segmentation, but also avoids the problems caused by the inconsistent quality of imported models, improves the reliability of the segmentation results, reduces the pressure on the staff for subsequent processing of the tooth partitioning results, and enables more accurate subsequent tooth layout design.

[0101] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0102] Furthermore, the above provides a detailed description of a tooth partitioning method and apparatus based on collision detection provided by the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A tooth partitioning method based on collision detection, characterized in that, The method includes: Import the tooth model and set the edge lines of the region based on the tooth model; 2D triangulation is performed based on the edge lines of the region to obtain the region baffle mesh; Seed points within the region are obtained by using the ray intersection method based on the region's baffle mesh. Based on the seed points within the region, a region diffusion collision process is performed to obtain an internal region patch. The missing areas of the internal region patch are filled with neighborhoods to obtain the internal region patch after neighborhood filling. The outer surface area is obtained using the method of finding the maximum connected region based on the inner region area after neighborhood filling, and the tooth partitioning result is output based on the outer surface area and the inner region area after neighborhood filling.

2. The tooth partitioning method based on collision detection according to claim 1, characterized in that, The process of importing a tooth model and setting the edge lines of a region based on the tooth model includes: Import the tooth model and select the position information of the edge lines based on the tooth model; The edge line of the region is set based on the position information of the edge line.

3. The tooth partitioning method based on collision detection according to claim 1, characterized in that, The step of performing 2D triangulation based on the edge lines of the region to obtain the region baffle mesh includes: Centroid points are calculated based on the edge lines of the region to obtain a central triangle, and the central triangle is set as the initial triangular mesh. The edge lines of the region include discrete points of the contour lines. Insert the discrete points of the contour line into the initial triangular mesh point by point to obtain a new triangular mesh; The new triangular mesh is optimized based on the Delaunay optimization criterion to obtain the regional baffle mesh.

4. The tooth partitioning method based on collision detection according to claim 1, characterized in that, The calculation based on the region baffle mesh using the ray intersection method to obtain seed points within the region includes: The region baffle mesh is searched to find the first triangle with the largest area in the region baffle mesh, and the center of the first triangle is calculated. The intersection points of the regions are obtained by using the ray intersection method based on the center of the first triangle; Seed points within the region are obtained based on the midpoint between the center of the first triangle and the intersection of the regions.

5. A tooth partitioning method based on collision detection according to claim 4, characterized in that, The calculation based on the center of the first triangle using the ray intersection method to obtain the intersection points of the regions includes: The center of the first triangle is selected as the starting point, and the opposite direction of tooth insertion is selected as the ray direction; Based on the starting point and the direction of the ray, the position where the ray intersects with the internal region is determined, and the intersection point of the regions is obtained.

6. A tooth partitioning method based on collision detection according to claim 1, characterized in that, The process of performing region diffusion collision processing based on seed points within the region to obtain internal region patches includes: Select a seed point inside the region as the center of the sphere, and set the radius of the sphere based on the side length of the tooth model; A sphere is generated within the region based on the sphere's center and radius. Set the single diffusion step size and diffusion direction; Based on the single diffusion step size and the diffusion direction, the sphere inside the region is subjected to regional diffusion collision processing to obtain an internal region patch.

7. The tooth partitioning method based on collision detection according to claim 1, characterized in that, The step of performing neighborhood filling processing on the missing areas of the internal region patch to obtain the neighborhood-filled internal region patch includes: Neighborhood features of internal region patches are extracted based on feature extraction methods; Based on the internal region patch and the neighborhood features, the missing regions of the internal region patch are filled with neighborhood features to obtain the internal region patch after neighborhood filling.

8. A tooth partitioning method based on collision detection according to claim 1, characterized in that, The inner region patches after neighborhood filling are used to obtain outer surface patches using the maximum connected component search method. Based on the outer surface patches and the inner region patches after neighborhood filling, tooth partitioning results are output, including: Return the internal region patches after neighborhood filling and set them to a disconnected state; The outer layer mesh is obtained by finding the maximum connected region based on the disconnected state of the internal region mesh after the neighborhood filling process. The tooth partitioning results are output based on the outer surface patch and the inner region patch after neighborhood filling processing.

9. A tooth partitioning method based on collision detection according to claim 8, characterized in that, The disconnected state of the internal region patches after neighborhood filling is obtained by using the method of finding the maximum connected region to obtain the outer layer patches, including: Based on the disconnected state of the internal region patches after neighborhood filling, a grayscale information similarity algorithm is used to obtain connected regions. The connectivity of the connected regions is analyzed to obtain the connectivity of the region pixels. Region filtering is performed based on the connectivity of the pixels in the region to obtain the outer layer sheet.

10. A tooth partitioning device based on collision detection, characterized in that, The device includes: Edge Line Module: Imports a tooth model and sets the edge lines of the region based on the tooth model; Regional baffle mesh module: Based on the edge lines of the region, 2D triangulation is performed to obtain the regional baffle mesh; Seed point module within the region: Based on the region's baffle mesh, the ray intersection method is used to calculate and obtain the seed points within the region; Internal region patch module: Based on the seed points inside the region, perform region diffusion collision processing to obtain internal region patches; Neighborhood filling module: performs neighborhood filling processing on the missing areas of the internal region patch to obtain the internal region patch after neighborhood filling processing; Outer surface slice and result module: Based on the inner region slice after neighborhood filling, the outer surface slice is obtained using the maximum connected region search method, and the tooth partitioning result is output based on the outer surface slice and the inner region slice after neighborhood filling.

Images