WO2011097947A1

WO2011097947A1 - Method for separating triangular mesh surface of tooth from triangular mesh surface of dental

Info

Publication number: WO2011097947A1
Application number: PCT/CN2011/000163
Authority: WO
Inventors: 钮叶新; 范然; 金小刚; 顾书华
Original assignee: 常州高新技术产业开发区三维工业技术研究所有限公司
Priority date: 2010-02-09
Filing date: 2011-01-30
Publication date: 2011-08-18
Also published as: CN102147935B; CN102147935A

Abstract

A method for separating the triangular mesh surface of a tooth from the triangular mesh surface of a dental is provided by the present invention, which integrates three advantages of accuracy, efficiency and smooth boundary, through the operations: picking the dental region to be separated interactively, preliminarily separating of the region growing, generating a space control curve automatically, positioning precisely, generating a separate line projection, optimizing of the smoothing of the separate line and so on, the separation of the surface triangular mesh of the tooth is finally realized.

Description

Method for segmenting dental triangle mesh surface from dental triangle mesh surface

Technical field

The present invention relates to a model segmentation technique in digital geometry processing, and more particularly to a method for segmenting a tooth triangular mesh surface for orthodontic, repair, and implantation.

Background technique

In recent years, oral CAD orthodontic systems have developed rapidly, and various new types of oral computer-aided treatment systems, such as scanning technology and digital geometric processing, have emerged in an endless stream. The two most typical systems are the OrthoCAD and Invi sal ign invisible correction systems. The typical mode of the oral computer-assisted treatment system is to scan the plaster model by optical method to obtain the three-dimensional data of the dental jaw, and then use the digital geometric processing technology to separate the dental gingival data, repair the teeth, and finally design the correction. Dental segmentation is the basic function of the computerized computer-assisted therapeutic system. Subsequent functions such as the design of the correction plan require the teeth to be used as an independent model for rigid body motion and interference detection. Since teeth have different shapes and there are large differences between different individuals, it is difficult to separate a single tooth quickly and accurately.

In 2004, Kondo used two distance images to identify the splitting boundary of two adjacent teeth. This method transforms the three-dimensional problem into an image problem, but the method ends up cutting the tooth data with a plane, and cannot accurately locate the tooth gingival boundary. See Toshiaki Kondo , S, H. Ong, Ke lvin WC Foong. Tooth segmentat ion of dental s tudy mode 1 s us ing range images. IEEE Transact ions on Medical Imag ing, 2004, 23 (3) : 350-362. Hao Guodong et al. Calculate the curvature of the vertices of the triangular mesh, select the feature regions according to the curvature values of the vertices, connect the feature regions interactively by the user, and then perform morphological operations on the feature regions to obtain the skeleton lines of the feature regions. According to the obtained skeleton lines, the teeth can be accurately separated. come out. The main idea of this method to separate the tooth model is to find the boundary line between the teeth. The main method is based on the automatic feature calculation of the curvature, but the usually obtained feature line does not completely conform to the tooth boundary, so it is impossible to perform precise cutting. The method has very high requirements on the model, and the requirements for the doctor or the technician to prepare the dental jaw or the dental impression and the modified modified plaster model are very high, the workload is relatively large, and the conditions for practical application are not satisfied, and it is not suitable for industrial use. See Hao Guodong, Cheng Yusheng, Dai Ning, Yu Qing. Morphological model for interactive segmentation of dental models[J] . China Manufacturing Informatization, 2008, (01) : 36- 39. Therefore, it is necessary to provide a more accurate method of boundary determination. A variety of commonly used 3D model segmentation algorithms can be used to design the tooth segmentation system. Chen et al. have no segmentation algorithm suitable for all kinds of models, see Xiaobai Chen, Aleksey Golovinskiy and Thomas Funkhouser. A benchmark for 3D mesh Segmentat ion. ACM Transact ions on Graphics, 2009, 28 (3) .

In general, the existing methods for segmenting the tooth triangle mesh surface from the dental triangle mesh surface generally have two ideas: (1) determining the li domain to determine the boundary; (2) determining the boundary directly according to the boundary segmentation. The former cannot obtain smooth and precise segmentation boundaries, and the latter requires too much user interaction and inefficiency. Summary of the invention

The present invention provides a method for segmenting a tooth triangular mesh surface from a dental mandrel triangular mesh surface. The method uses a variety of techniques in digital geometry processing and proposes a new segmentation line localization algorithm. The topology tracking is used to project the space control curve to the triangular mesh surface, realizing the combination of fast automatic segmentation and boundary precise adjustment. Makes teeth cut fast, precise, and smooth.

The invention mainly comprises six steps, and the process is as shown in Fig. 1:

A. Select the region of the tooth to be segmented containing the target tooth triangle mesh surface, on which the foreground line is marked to identify the tooth area, and the sketched background line identifies the gum area. This step provides a novel means of interaction for determining the target tooth area.

(1) Using the scene picking geometry as a space bounding box to pick up the segmented jaw region to improve the operating speed, the segmented jaw region includes the target tooth triangle mesh surface and its surrounding triangular mesh surface. Scene picking geometry mainly includes but is not limited to: cube, cuboid, cylinder. Scene picking geometry allows you to pan, rotate, scale, and combine operations.

(2) Sketch the foreground line and the background line on the above-mentioned area to be divided. The foreground line is a set of triangular patches on the target tooth triangle mesh surface selected by the user to identify the target tooth region; the background line is a set of triangular patches on the triangular mesh surface of the target tooth selected by the user, and the target tooth is identified. The surrounding gum area.

B. Calculate the area of the tooth region to be segmented that belongs to the target tooth and determine the initial boundary of the target tooth. The foreground line and the background line information in step A are used as the seed area of the area growing algorithm, and the greedy algorithm is used to perform regional growth, so that the target tooth area is approximately separated, and the initial boundary of the target tooth is obtained, which provides a follow-up operation. The distance formula of the greedy algorithm is based on the principle of visual minimization:

Where ⁽ Ρ, ^ distance metric, is the Euclidean distance of the path, ^ is the length of the path Gauss map, w and yes Weights. It is a measure of the negative curvature on the surface. According to the visual principle, the boundary between the parts of the object is generally on the minimum negative curvature line. The purpose of the item is to increase the distance influence of the negative curvature area on the surface. The definition is as follows:

Where _{gW is} defined as a function with a large growth rate such as or etc.

C. The initial boundary acquisition node is uniformly sampled according to the specified number, and the space control curve is calculated, and the spatial control curve is interactively adjusted to the target position according to the boundary feature of the teeth and the gums.

Since the tooth-neck model has complex features and rich details, the initial boundary of the target tooth obtained in step B is usually not directly used as the segmentation boundary. The present invention automatically generates a space control curve by using a computer and performs interactive adjustment to obtain the boundary target position.

First, the number of samples is specified by the user. According to the number, the computer automatically and uniformly samples the initial boundary acquisition nodes, and generates a spatial control curve based on the node information. The structure of the dental model is complex and the structure itself is not significant. At the same time, due to the limitations of existing data acquisition devices, the digital dental model is not a regular continuous feature area in the geometrical sense of the gum line, and in most cases. Underneath, the number of human oral teeth is between 28 and 32. Dividing a triangular mesh surface for each tooth of each benefiter is actually a rather time consuming and complicated task. The traditional method relies entirely on the human eye to observe the feature area (ie, the area where the gum line is located). The way to construct the space control curve interactively requires a large amount of manual observation of the complex structure, and the meshing surface is divided into all the teeth of each patient, and the process workload is constructed. Large, inefficient, not suitable for industrial production. By means of the method of the invention, the teeth and the gums can be accurately distinguished from each other, and the space control curve can be quickly constructed, thereby greatly reducing the workload in the later stage, thereby improving the efficiency of the entire design process and reducing the labor load of the designer. Suitable for industrial production. The space control curve can be a spline curve of 2 or more times. The present invention uses a 3 B-spline curve as a space control curve. A split line is a polygon embedded in a triangular mesh whose shape is changed by the positional adjustment of the vertices or edges of the polygon. The position adjustment is also limited to the triangular mesh surface. Since the number of vertices and sides of the dividing line polygon is huge, if the direct adjustment is performed, the interactive operation workload is very large, the efficiency is very low, and it is difficult to control the shape of the dividing line polygon as a whole, so the present invention adjusts the segmentation indirectly by adjusting the shape of the space control curve. line. This method has the advantage of being flexible and precise relative to the former. The shape adjustment of the space control curve is achieved by the addition, deletion, and positional changes of the nodes on the triangular mesh surface of the region to be divided. By interactively adjusting the space control curve to the target position, the boundary target position is precisely specified, and the actual situation of the complicated and variable tooth boundary is adapted.

D. Project the adjusted spatial control curve as a dividing line embedded on the surface of the dental triangle mesh. The dividing line is coplanar with the passing triangular face piece.

To segment the toothed triangle mesh surface, you first need to determine the segmentation line embedded on the surface, and the space control curve is close to the shape of the segmentation line but does not coincide. Therefore, the space control curve needs to be projected on the surface of the tooth triangle. The method for transforming the space control curve projection into the dividing line embedded on the tooth surface of the dental jaw triangle adopts two algorithms: (1) spatial control curve projection based on topology tracking; (2) spatial control based on nearest point Curve projection.

(1) Space control curve projection based on topology tracking

As shown in Figure 2, let the space control curve parameter be expressed as "«, which is a point on the space control curve, 9 ⁽ 0 is the corresponding projection point of the point on the toothed triangle mesh surface, every point on the space control curve The projection line vector on the toothed triangle mesh surface is ^ = P ⁽ 0-9 ⁽ 0, with the space control curve as the guideline to construct the ruled surface ^, =« ⁽ 0 ₊ , then the space control curve projection problem can be converted For the ruled surface, the problem is solved with the triangular mesh surface.

The invention discretizes the space control curve into a polygon, and the problem of the intersection of the ruled surface can be transformed into the problem of tracking the projection point by using the local topology relationship. In the concrete realization, the space control curve is approximated by a polygon, and the space control curve L is composed of vertices { , a, A, ej, and the nodes are included in the point set. Let F be the current triangular patch, and set Proj (Plane, Q) as a point-to-plane projection operation, which is equivalent to finding the closest point to the plane.

The core idea of the topology tracking algorithm is to consider only the projection of the vertices of the control curve L on F. If the projection point is not in the face F, the neighborhood triangle of the current face F is searched, the patch of the projection point is calculated and Set to the current plane, perform the above operations on each vertex of L, and the space control curve projection operation can be efficiently realized. The biggest advantage of this algorithm is to minimize the search range and make full use of the local topology information on the triangular mesh surface to maximize the path search efficiency.

The algorithm steps for each vertex of the space control curve L are as follows:

Stepl: Calculate Proj(F, Q) to determine whether the projection is in the current plane;

Step2: If the projection point does not leave the current plane, determine the next vertex of L;

Step3: If the projection exceeds the current plane, search for the F neighborhood triangle patch, if no algorithm is found to exit;

Step4: If the projection point is found within a patch in the neighborhood, set to the current triangle patch. Topology tracking typically presents inaccuracies in areas of greater curvature, and the present invention solves this problem with a control curve projection based on the nearest point. The basic idea is to sample the projection point, that is, the point on the sampling space control curve, and find the nearest point on the triangular mesh surface. The sampling projection point should be dense enough so that two adjacent projection sampling points are at most one on the grid. Intersect. But because triangle mesh surfaces are usually It is reconstructed after scanning, and even the subsequent optimization operation can change the mesh quality. The density of the sample cannot be well controlled. The nearest point projection sampling in this system is only used as a means of automatically encrypting the control node. After the user issues an automatic encryption control node command, the system encrypts the sampling point on the control curve to double it, calculates the nearest point of the sample point to the triangle mesh, and recalculates the control curve with these nearest points as control vertices to increase the topology tracking projection. Stability.

(2) Space control curve projection based on the nearest point

In this system, the projection of the point on the space control curve to the triangular mesh surface is accelerated by the uniform spatial partitioning structure to find the nearest point.

Let q be the surface of the surface ^{5 (} ", ^, the Euclidean distance of the spatial point p to q is «? il, then the shortest distance from p to the surface is defined as ^β = minWA9) \ ie S} _o

Z) (The necessary condition for _{P to} get the extreme value is ou dv

Is the unit normal on the surface. If the surface is smooth, then ⁽ p - that is, the vector of the nearest point to the space point is parallel to the normal at the nearest point. The surface is discretized into a triangular mesh, and the system uses the uniform space splitting structure to accelerate the nearest Click to find.

The algorithm steps are as follows:

Stepl: divide the target triangle mesh into uniform cells, and store the triangular mesh vertex indexes in the cells respectively;

Step2: Search the triangular patches near the spatial point p by using the neighborhood topology relationship, and calculate the projection point of the plane from the plane of the triangle to the triangle;

Step3: If the projection point is in the triangle patch, mark it as the nearest point candidate point and record the distance; Step4: Take the projection point with the smallest distance as the nearest point.

The above algorithm makes full use of geometric neighborhood information and topology neighborhood information. The combination of the two methods can quickly determine the shape of the dividing line embedded in the triangular mesh surface.

Since the spatial control curve is close to the tooth gingival segmentation boundary, calculating the nearest point of the apex on the spatial control curve on the triangular mesh surface is equivalent to calculating the projection of the vertex on the triangular mesh surface. The projection point obtained by this method is used as a segmentation node in the topology tracking projection algorithm to increase the stability of the topology tracking projection.

E. Smoothing optimized dividing line

The present invention utilizes the topology tracking algorithm and the nearest point projection algorithm to solve the intersection of the ruled surface and the dentition triangle mesh surface generated by the spatial control curve. Although the efficient and flexible curve control is realized, the embedded cannot be guaranteed. The split lines that are placed on the grid remain smooth. We use the active contour algorithm on the toothed triangle mesh surface as a follow-up optimization method to ensure that the segmentation boundary remains smooth. This process is achieved by the motion of the dividing line vertices on the edges of the triangular mesh.

The principle of the active contour algorithm is to use the energy function to optimize the shape of the curve to make it smoother or more closely controlled to the user-specified shape. It has the advantage of moving directly to the ideal shape through the curve embedded in the mesh, which is intuitive and can be controlled according to the scalar field on the triangular mesh surface. The system uses this algorithm as a subsequent optimization operation to increase the smoothness of the curve shape.

The split line energy formula embedded in the triangular mesh surface is as follows:

Among them, £ _ta "TM." determines the smoothness of the curve, ^ is the external constraint acting on the triangular mesh. According to the variational principle, the system uses the explicit Euler iteration to solve the equation for embedding on the grid. Each vertex of the split line causes it to move along the edge to reduce the energy value, and repeating this process ultimately keeps the split line smooth.

F. Dividing the triangular patches on the surface of the dental triangle mesh to obtain a smooth triangular mesh surface with a smooth boundary

Use the dividing line in step E to split the ivory triangle mesh surface to obtain the target tooth triangle mesh surface. The dividing line is a spatial polygon whose vertex is embedded on the edge of the triangular surface of the toothed triangle mesh surface. After the splitting, the dividing line becomes the boundary of the target tooth triangle mesh surface. The dividing line embedded on the triangular mesh surface is coplanar with the passing triangular patch, and can have one or two intersection points with the passing triangular patch, and is divided into two or three triangles after splitting. After the split, the tooth triangle mesh surface is separated from the toothed triangle mesh surface.

The invention relates to a method for segmenting a tooth triangle mesh surface from a dental mandrel triangle mesh surface, which can not only process the dental triangle mesh surface, but also can process any triangular mesh surface, especially for the insignificant Or fast and precise cropping of triangular mesh surfaces with irregular feature regions. This method is a highly efficient and versatile triangular mesh surface segmentation method.

On the other hand, the method of the present invention provides a convenient modeling method for orthodontics, repair and implantation. The invention further provides the above-mentioned method for segmenting a tooth triangular mesh surface from a gingival triangular mesh surface in orthodontic and/or repairing and/or planting, especially in orthodontics, restoration and planting. . DRAWINGS Figure 1 Flow chart of segmenting the tooth triangle mesh surface from the dental triangle mesh surface

Figure 2 Schematic diagram of spatial control curve projection based on topology tracking

Figure 3 Upper maxillary triangular mesh surface in the case of right maxillary triangular mesh surface segmentation from the upper maxillary triangle mesh surface in orthodontics

Fig. 4 In the orthodontic treatment, the upper jaw region of the right incisor triangle mesh surface is segmented from the upper jaw triangle mesh surface

Figure 5 Sketched foreground and background lines in the right incisor triangle mesh surface case from the upper jaw triangle mesh surface in orthodontics

Figure 6 The initial boundary of the right incisor in the right incisor triangle mesh surface case from the upper maxillary triangle mesh surface in orthodontics

Figure 7 Spatial control curve adjusted to the target position in the case of the right incisor triangle mesh surface segmentation from the upper jaw triangle mesh surface

Fig. 8 The dividing line generated by the spatial control curve projection in the case of the right incisor triangle mesh surface segmentation from the upper jaw triangle mesh surface in orthodontics

Fig. 9 The smoothing-optimized dividing line in the case of right-handed triangular mesh surface segmentation from the upper maxillary triangle mesh surface in orthodontics

Fig.10 Tooth gingival view of the right incisor triangle mesh surface in the right incisor triangle mesh surface case from the upper maxillary triangle mesh surface in orthodontics

Figure 11 Orthodontic orthodontic segmentation of the right incisor triangle mesh surface from the upper jaw triangle mesh surface case

Figure 12 Oral four-unit bridge restoration in the upper maxillary base crown triangle mesh surface segmentation of the right first molar The base crown triangle mesh surface case in the upper jaw base crown triangle mesh surface

Figure 13 The upper first molar is divided from the upper maxillary base crown triangle mesh surface in the oral four-unit bridge repair. The selected upper jaw region is selected in the base crown triangle mesh surface case.

Figure 14 Oral four-unit bridge repair in the upper first molar from the upper jaw base crown triangle mesh surface segmentation of the right first molar in the base crown triangle mesh surface case sketched foreground and background lines

Figure 15 The initial boundary of the right first molar base crown in the case of the upper quadrilateral sulcus

Figure 16 Spatial control curve adjusted to the target position in the case of the upper quadrilateral base crown triangle mesh surface segmentation in the oral four-unit bridge repair

Figure 17 The right first molar is divided from the upper jaw base crown triangle mesh surface in the oral four-unit bridge repair The dividing line generated by the space control curve projection in the base crown triangle mesh surface case

Figure 18 Segmentation of the right first molar from the upper maxillary base crown triangle mesh surface in the four-unit bridge restoration of the oral cavity. The smoothed optimization of the dividing line in the case of the base crown triangle mesh surface

Figure 19 Oral four-unit bridge repair from the upper jaw base crown triangle mesh surface segmentation right first molar base crown triangle mesh surface case in the right first molar base crown triangle mesh surface tooth gingival diagram Figure 20 oral Individual unit diagram of the right first molar base crown triangle mesh surface in the case of the quadrilateral bridge body repair from the upper jaw base crown triangle mesh surface segmentation right first molar base crown triangle mesh surface case

DETAILED DESCRIPTION OF THE INVENTION The present invention is not intended to limit the scope of the invention.

Embodiment 1 In the orthodontic treatment, the right incisor triangle mesh surface is segmented from the upper jaw triangle mesh surface.

1. Open the computer software and import the upper jaw triangle mesh surface data, as shown in the attached figure: Figure 3 Oral orthodontics from the upper jaw triangle mesh surface segmentation of the right incisor triangle mesh surface case in the upper jaw triangle case Grid surface.

2. Use the scene picking cube to select the upper jaw region including the right incisor triangle mesh surface to improve the subsequent operation speed, as shown in the attached figure: Figure 4 Segmentation from the upper jaw triangle mesh surface in orthodontics The selected upper jaw region to be segmented in the right incisor triangle mesh surface case.

3. Select a set of triangular patches on the target tooth triangle mesh surface to be divided into the upper jaw region as the foreground line to identify the tooth region, and select the triangle mesh surface around the target tooth to be divided into the upper jaw region. A set of triangular patches, used as background lines to identify the gingival area, as shown in the attached figure: Figure 5 Sketched foreground line in the case of right-toothed triangular mesh surface segmentation from the upper jaw triangle mesh surface in orthodontics And background lines.

4. Using the foreground line and background line information in the third step as the seed area, use the greedy algorithm to calculate by the base area growth.

The initial boundary of the right incisor, as shown in the attached figure: Figure 6 The initial boundary of the right incisor in the right incisor triangle mesh surface case from the upper maxillary triangle mesh surface in orthodontics.

5. The number of interactive designation is 30, the computer software automatically samples the initial boundary in the fourth step according to the number, and uses the 3 times B-spline curve to produce the space control curve. According to the boundary feature of the right incisor and the gum, the positional variation of the node on the triangular mesh surface of the tooth region to be segmented is interactively adjusted, the spatial control curve is interactively adjusted to the target position, and the dividing line is indirectly accurately positioned. As shown in the attached drawings: Figure 7 Spatial control curve adjusted to the target position in the case of the right-toothed triangular mesh surface segmentation from the upper-triangular triangular mesh surface in orthodontics.

6. The computer software automatically calls the space control curve projection based on topology tracking and the space-based curve projection based on the nearest point, and projects the space control curve in the fifth step into the upper jaw triangle mesh surface. The dividing line, as shown in the specification drawing: Figure 8 The dividing line generated by the space control curve projection in the case of the right incisor triangle mesh surface segmentation from the upper jaw triangle mesh surface in orthodontics.

7. The computer software automatically calls the movable wheel gallery algorithm on the surface of the dental triangle mesh, smoothing the dividing line in the sixth step to achieve smooth segmentation, as shown in the attached figure: Figure 9 From the upper teeth in orthodontics The segmentation line after smoothing optimization in the case of the right triangle triangle mesh surface segmentation of the jaw triangle mesh surface.

8. The computer software automatically follows the smoothing line after the smoothing in the seventh step, and divides the triangular face piece that the dividing line traverses to obtain the right-cut tooth triangle mesh surface with smooth boundary, as shown in the specification: Figure 10 Orthodontic orthodontics The tooth gingival view of the right incisor triangle mesh surface in the right incisor triangle mesh surface case is segmented from the upper jaw triangle mesh surface; Figure 11 is the right incisor from the upper jaw triangle mesh surface in orthodontics An individual figure of the tooth of the right incisor triangle mesh surface in the triangular mesh surface case.

Example 2 In the oral four-unit bridge repair, the right first molar base crown triangle mesh surface is segmented from the upper jaw base crown triangle mesh surface.

1. Open the computer software and import the upper mesh base triangle mesh surface data, as shown in the attached figure: Figure 12 Oral four unit bridge repair from the upper jaw base crown triangle mesh surface segmentation right first molar base crown triangle The upper jaw base crown triangle mesh surface in the mesh surface case.

2. Use the scene picking cube to select the upper jaw region including the right first molar base crown triangle mesh surface to improve the subsequent operation speed, as shown in the attached figure: Figure 13 Oral four unit bridge repair from the upper teeth The maxillary basal crown triangle mesh surface segmentation is selected from the upper first molar base crown triangle mesh surface case to be divided into the upper jaw region.

3. Select a set of triangular patches on the target base crown triangle mesh surface to be segmented in the upper basal crown region as the foreground line to identify the base crown region, and the target base crown in the upper jaw base region to be segmented. A set of triangular patches is selected on the surrounding triangular mesh surface as the background line to identify the non-basal crown region, as shown in the specification drawing: Figure 14 Oral four-unit bridge repair from the upper jaw base crown triangle mesh surface segmentation right The foreground and background lines of the sketch in the first molar base crown triangle mesh surface case. 4. Using the foreground line and background line information in the third step as the seed area, use the greedy algorithm, and pass the distance formula based on the visual minimum principle ^ _Γ (Ρ^) = !γ^ + Μ> \ds + w \f (k _D )ds for segmentation region growth, the computer software automatically calculates the region of the right first molar molar base crown in the triangular mesh surface of the upper jaw base, and determines the initial boundary of the right first molar base crown, as shown in the specification: Fig. 15 The initial boundary of the right first molar base crown in the case of the right first molar base basal crown triangle mesh surface in the oral four unit bridge repair.

5. The number of interactively designated samples is 20, the computer software automatically samples the initial boundary in the fourth step according to the number, and uses the 3 times B-spline curve to produce the space control curve, according to the right first molar base crown. The boundary feature with the non-base crown, through the addition, deletion of nodes and the positional changes of the nodes on the triangular mesh surface of the base region to be divided, interactively adjust the spatial control curve to the target position, and indirectly accurately locate the dividing line, as in the specification Figure: Figure 16 Oral four-unit bridge repair in the space control curve from the upper-finished base crown triangle mesh surface segmentation to the right first molar base crown triangle mesh surface case adjustment to the target position.

6. The computer software automatically calls the space control curve projection based on topology tracking and the space-based curve projection based on the nearest point. The spatial control curve in the fifth step is projected as a triangle mesh embedded in the upper collar base. The dividing line on the surface, as shown in the drawing of the specification: Figure 17 The four-unit bridge in the oral cavity is generated from the upper jaw base crown triangle mesh surface segmentation in the right first molar base crown triangle mesh surface case generated by the space control curve projection split line.

7. The computer software automatically calls the active contour algorithm on the triangular mesh surface of the upper jaw base crown, smoothing the dividing line in the sixth step to achieve smooth segmentation, as shown in the attached figure: Figure 18 Oral four unit bridge In the repair, the smoothing-optimized dividing line in the case of the right first molar base crown triangle mesh surface is segmented from the upper crown base crown triangle mesh surface.

8. The computer software automatically follows the smoothing-optimized dividing line in the seventh step, and divides the triangular face piece that the dividing line traverses to obtain a smooth right-cutting triangular mesh surface, as shown in the specification drawing: Figure 19 Oral four units In the pontic repair, the tooth gingival view of the right first molar base crown triangle mesh surface in the right first molar base crown triangle mesh surface case is divided from the upper tooth base base crown triangle mesh surface; Figure 20 oral four unit bridge body In the repair, the individual figure of the right first molar base crown triangle mesh surface in the right first molar base crown triangle mesh surface case is segmented from the upper jaw base crown triangle mesh surface.

Claims

Rights request

1. A method for segmenting a tooth triangular mesh surface from a dental triangle mesh surface, the method comprising the steps of:

(A) selecting a region of the tooth to be segmented including a triangular mesh surface of the target tooth, and sketching a foreground line on the region to identify a tooth region, the foreground line consisting of a set of triangular patches of the selected target tooth region; Sketching the background line identifies the gingival area, the background line consisting of a set of triangular patches of the selected target tooth peripheral area;

(B) using the foreground line and the background line specified in the step (A) to initially calculate the area of the tooth surface of the dental jaw triangle mesh to determine the initial boundary of the target tooth;

(C). On the initial boundary determined in step (B), the acquisition node is uniformly sampled according to the number of specified samples and the space control curve is calculated, and the spatial control curve is interactively adjusted to the target position according to the boundary feature of the teeth and the gums;

(D) Projecting the adjusted spatial control curve in step (C) into a dividing line embedded in the triangular surface of the dental jaw triangle, the dividing line being coplanar with the passing triangular surface;

(E). Smoothing optimized dividing line;

(F). According to the dividing line in step (E), the triangular patches on the surface of the dental triangle mesh are segmented to obtain a tooth triangular mesh surface with a smooth boundary.

2. A method for segmenting a tooth triangular mesh model from a dental triangle mesh surface according to claim 1, characterized in that the scene picking geometry is used to select a triangular mesh surface containing the target tooth and its surrounding triangular mesh. The area of the tooth to be divided inside the surface.

3. A method for segmenting a tooth triangular mesh surface from a dental triangle mesh surface according to claim 2, wherein the scene picking geometry comprises: a cube, a cuboid, and a cylinder.

4. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 1, wherein the input foreground line and background line information are used as seed regions of the region growing algorithm. The greedy algorithm performs regional growth, and initially calculates the area of the tooth region to be segmented that belongs to the target tooth.

5. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 4, wherein the distance formula of the greedy algorithm is based on a visual minimum principle:

^ (Ρ> 9) = ί ίώ + w \ ώ + w \ f{k ^ )ds where the distance metric is the Euclidean distance of the path, 'is the length of the path Gaussian map, w is the weight, is on the surface a measure of negative curvature, used to achieve minimal principles, ie human vision for objects The shape is generally identified by a line of minimal negative curvature as a boundary between the parts.

6. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 1, characterized in that the spatial control curve generated in step (C) is a spline greater than or equal to 2 times. curve.

7. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 1, characterized in that the node of the spatial control curve generated in step (C) is in the region of the tooth to be segmented On the triangular mesh surface.

8. A method for segmenting a tooth triangular mesh surface from a dental collar triangular mesh surface according to claim 1, wherein the spatial control curve is integrated with the triangular mesh surface projection of the tooth to be segmented. The method is implemented as follows: (1) Based on the projection method of topology tracking, the continuity of the projection of the space control curve on the triangular mesh surface is used to project the extension direction in the neighborhood of the vertex; (2) the projection method based on the nearest point search Use the spatial splitting structure to quickly find the closest point of the node on the grid.

9. A method for segmenting a tooth triangulated mesh surface from a dental mandibular triangular mesh surface according to claim 1, wherein the smoothing optimized segmentation line utilizes an active contour algorithm on a triangular mesh surface,

Where !u represents the smoothness of the split line, and J" ^£ «^ _rain ^ represents the effect of external factors on the shape of the split line.

10. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 1, wherein the dividing line is divided into triangular faces on the dental triangle mesh surface, The topology is subdivided into two or three triangular patches.

11. A method for segmenting a tooth triangular mesh surface from a dental mandibular triangular mesh surface according to claim 1, wherein the segmented operational object dental triangle mesh surface can be a full or partial jaw of the tooth Triangular mesh surface.

12. Use of the method of claim 1 in orthodontics and/or repair and/or implantation.

13. The method of claim 1 for use in segmenting a triangular mesh surface of a random feature region.