WO2024066891A1 - 曲面融合方法、装置和医疗影像设备 - Google Patents

曲面融合方法、装置和医疗影像设备 Download PDF

Info

Publication number
WO2024066891A1
WO2024066891A1 PCT/CN2023/115899 CN2023115899W WO2024066891A1 WO 2024066891 A1 WO2024066891 A1 WO 2024066891A1 CN 2023115899 W CN2023115899 W CN 2023115899W WO 2024066891 A1 WO2024066891 A1 WO 2024066891A1
Authority
WO
WIPO (PCT)
Prior art keywords
tsdf
value
mesh
fused
corner point
Prior art date
Application number
PCT/CN2023/115899
Other languages
English (en)
French (fr)
Inventor
王瑜
宋诚谦
Original Assignee
合肥美亚光电技术股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 合肥美亚光电技术股份有限公司 filed Critical 合肥美亚光电技术股份有限公司
Publication of WO2024066891A1 publication Critical patent/WO2024066891A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T15/00—3D [Three Dimensional] image rendering
    • G06T15/10—Geometric effects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation

Definitions

  • the present disclosure relates to the field of medical technology, and in particular to a surface fusion method, device and medical imaging equipment.
  • Intraoral scanners need to reconstruct, align, and fuse at the same time during the scanning process, and the fused surface should be displayed in real time.
  • Conventional methods require aligning and fusion before generating the surface, which makes conventional methods either fail to meet the accuracy requirements or cannot keep up with the speed.
  • the first purpose of the present disclosure is to propose a surface fusion method, which can effectively fuse the surface to be fused and the target surface without the need to reconstruct the surface separately, thereby greatly reducing the amount of calculation.
  • the third objective of the present disclosure is to provide a medical imaging device.
  • the first aspect of the present disclosure proposes a surface fusion method, which includes: calculating a first TSDF value corresponding to each corner point of the surface to be fused; obtaining a second TSDF value corresponding to each corner point of the target surface; determining a third TSDF value corresponding to each corner point of the fused surface based on the first TSDF value and the second TSDF value; determining the vertex coordinates of the triangular mesh of the fused surface and the topological structure of the triangular mesh based on the third TSDF value and the coordinates of the corner point; determining the fused surface based on the vertex coordinates of the triangular mesh of the fused surface and the topological structure of the triangular mesh.
  • the first TSDF value of each corner point corresponding to the surface to be fused is calculated, including: determining the envelope mesh group of the surface to be fused, and for each corner point in the envelope mesh group, calculating the minimum distance from the corner point to all triangular meshes on the surface to be fused; and using the minimum distance as the TSDF distance in the first TSDF value corresponding to each corner point of the surface to be fused.
  • an envelope grid group of the surface to be fused is determined, and For each corner point in the group, the minimum distance from the corner point to all triangular meshes on the surface to be fused is calculated, including: for all triangular meshes on the surface to be fused, an envelope mesh group that envelops all triangular meshes is determined, recorded as the first envelope mesh group, and the first envelope mesh group is the envelope mesh group of the surface to be fused; for each corner point of each cube in the first envelope mesh group, the projection point from the corner point to all triangular meshes on the surface to be fused is determined, and the first distance from the corner point to each of its projection points is calculated; the minimum value of the first distance of each corner point is used as the minimum distance from the corner point to all triangular meshes on the surface to be fused;
  • each triangular mesh on the surface to be fused determines the envelope mesh group that envelops the triangular mesh, recorded as the second envelope mesh group, and all second envelope mesh groups constitute the envelope mesh group of the surface to be fused; for each corner point of each cube in each second envelope mesh group, determine the projection point of the corner point to the corresponding triangular mesh, and calculate the first distance from the corner point to each of its projection points; take the minimum value of the first distances corresponding to the same corner point as the minimum distance from the corresponding corner point to all triangular meshes on the surface to be fused; wherein the projection point used to calculate the first distance is the projection point that falls within the corresponding triangular mesh.
  • a method for determining projection points includes at least one of the following methods: Method 1: For each corner point, the centroid method and the perpendicular vector method are used to obtain the projection point of the corner point to all triangular meshes on the surface to be fused; Method 2: For each mesh corresponding to the surface to be fused, the intersection of the center line of the opposite surface of the mesh and the corresponding triangular mesh on the surface to be fused is determined, and the intersection is used as the projection point of the 8 corner points of the mesh to the corresponding triangular mesh; Method 3: For each triangular mesh on the surface to be fused, the intersection of the mesh surface of the envelope mesh group of the triangular mesh and the triangular mesh is determined, and the intersection is used as the projection point of the 8 corner points of each mesh in the envelope mesh group of the triangular mesh to the triangular mesh.
  • the projection points of the corner points adjacent to the corner point are used as the projection points of the corner point.
  • the third TSDF value of each corner point corresponding to the fused surface is determined according to the first TSDF value and the second TSDF value, including: determining the envelope grid group of the target surface, recorded as the third envelope grid group, and recording the envelope grid group of the surface to be fused as the fourth envelope grid group; for each corner point of the common part of the third envelope grid group and the fourth envelope grid group, obtaining the TSDF distance in the third TSDF value of the corner point according to the TSDF distance in the first TSDF value of the corner point, the TSDF distance in the second TSDF value and the TSDF weight, and obtaining the TSDF weight in the third TSDF value of the corner point according to the TSDF distance in the first TSDF value of the corner point and the TSDF weight; for each corner point of the newly added part of the fourth envelope grid group relative to the third envelope grid group, using the first TSDF value of the corner point as the TSDF distance in the third TSDF value of the
  • the TSDF weight in the third TSDF value is obtained according to the TSDF distance and the TSDF weight in the first TSDF value, including: when the TSDF distance in the first TSDF value is less than the preset cutoff distance, according to the second The TSDF weight in the TSDF value is used to obtain the TSDF weight in the third TSDF value; when the TSDF distance in the first TSDF value is greater than or equal to the preset cutoff distance, the TSDF weight in the second TSDF value is used as the TSDF weight in the third TSDF value.
  • the TSDF distance in the third TSDF value is obtained according to the TSDF distance in the first TSDF value, the TSDF distance in the second TSDF value and the TSDF weight, including: calculating the product of the TSDF distance in the second TSDF value and the TSDF weight to obtain a first product; calculating the product of the TSDF distance in the first TSDF value and the unit threshold of TSDF to obtain a second product; calculating the sum of the first product and the second product to obtain a first sum value; calculating the sum of the TSDF weight in the second TSDF value and the unit threshold to obtain a second sum value; calculating the ratio of the first sum value to the second sum value to obtain the TSDF distance in the third TSDF value; wherein, obtaining the TSDF weight in the third TSDF value according to the TSDF weight in the second TSDF value includes: using the second sum value as the TSDF weight in the third TSDF value.
  • the vertex coordinates of the triangular mesh of the fused surface and the topological structure of the triangular mesh are determined according to the third TSDF value of the corner point and the corner point coordinates, including: determining the vertex coordinates of the triangular mesh of the fused surface according to the corner point coordinates and the TSDF distance in the third TSDF value; determining the nMC value corresponding to the cube mesh according to the TSDF distance in the third TSDF value, and determining the topological structure of the triangular mesh of the fused surface according to the nMC value and the vertex coordinates of the triangular mesh.
  • the surface fusion method of the disclosed embodiment calculates the first TSDF value of the surface to be fused through the TSDF algorithm, obtains the second TSDF value of the fused surface, and calculates the third TSDF value of each corner point corresponding to the fused surface based on the first TSDF value and the second TSDF value.
  • the corner point position is used instead of the voxel position, which reduces the amount of calculation for surface fusion.
  • the coordinates of the triangular mesh vertices of the fused surface and the topological structure of the triangular mesh are determined, thereby determining the fused surface.
  • the process of surface fusion calculation is the process of determining the triangular mesh vertices and the triangular mesh topological structure of the fused surface, which is equivalent to merging surface fusion and surface reconstruction, without the need to reconstruct the surface separately, greatly reducing the amount of calculation.
  • a second aspect of the present disclosure proposes a surface fusion device, including a memory, a processor, and a computer program stored in the memory.
  • the computer program is executed by the processor, the surface fusion method as described above is implemented.
  • a third aspect of the present disclosure provides a medical imaging device, including the curved surface fusion device as described above.
  • FIG1 is a flow chart of a surface fusion method according to an embodiment of the present disclosure
  • FIG2 is a flowchart of calculating the first TSDF value of each corner point corresponding to the surface to be fused according to an embodiment of the present disclosure
  • FIG3( a ) is a flowchart of calculating the minimum distance from a corner point to all triangular meshes on a surface to be fused according to an embodiment of the present disclosure
  • FIG3( b ) is a flowchart of calculating the minimum distance from a corner point to all triangular meshes on a surface to be fused according to another embodiment of the present disclosure
  • FIG4( a ) is a schematic diagram of calculating a triangular mesh projection point according to an embodiment of the present disclosure
  • FIG4( b ) is a schematic diagram of calculating triangular mesh projection points according to another embodiment of the present disclosure.
  • FIG5 is a flowchart of determining the third TSDF value of each corner point corresponding to the fused surface according to an embodiment of the present disclosure
  • FIG6 is a flowchart of calculating the TSDF distance in the third TSDF value according to an embodiment of the present disclosure
  • FIG7 is a flow chart of calculating the coordinates of the triangle mesh vertices of the fused surface and the topological structure of the triangle mesh according to an embodiment of the present disclosure
  • FIG8 is a schematic diagram of corresponding marks of edges and corners of a grid block according to an embodiment of the present disclosure.
  • FIG9 is a schematic structural diagram of a curved surface fusion device according to an embodiment of the present disclosure.
  • FIG. 10 is a schematic diagram of the structure of a medical imaging device according to an embodiment of the present disclosure.
  • the surface generated by the first scan is used as the target surface
  • the surface generated by the second scan is used as the surface to be registered and fused.
  • the surface to be registered and fused is fused to the target surface to generate a new target surface, and the surface scanned later is then fused and registered to the new target surface. In this way, multiple rounds of registration and fusion are performed to generate a larger surface, thereby completing the scanning and imaging of the three-dimensional object.
  • the surface fusion method of the embodiment of the present disclosure can be applied to the above fusion process.
  • the specific fields in which the embodiments of the present disclosure can be applied are not limited to the above-mentioned scanning and imaging fields of three-dimensional objects. In other fields, as long as the corresponding conditions in the fusion process can be met, the surface fusion method of the embodiments of the present disclosure can be used to fuse any two surfaces.
  • FIG1 is a flow chart of a surface fusion method according to an embodiment of the present disclosure.
  • the surface fusion method includes:
  • the present disclosure uses the TSDF (Truncated Signed Distance Function, based on the truncated zone signed distance function) method to process the surface scanned by the camera (such as medical imaging equipment such as oral digital impression device).
  • the TSDF method divides the space into equilateral and infinitely extendable grid blocks (i.e., cubes).
  • the TSDF value of the corner point includes the TSDF distance, i.e., the value of the corresponding truncated distance function, and the TSDF weight, i.e., the corresponding weight, which can represent the sampling quality of the corresponding point.
  • the triangle mesh position of the corresponding surface can be obtained by linear interpolation, which can be represented by the vertex coordinates of the triangle mesh and the topological structure of the triangle mesh.
  • the surface to be fused is the first surface
  • the surface to be fused is used as the target surface
  • the MC (Marching cubes) mesh division of the target surface is used as the division of the basic model mesh block, and the basic mesh is extended to form the basic mesh, and the first TSDF value of each corner point of all cube meshes corresponding to the target surface can be calculated.
  • the surface to be fused is not the first surface, the surface to be fused is processed according to the above-mentioned basic model network block to obtain the first TSDF value of each corner point of all cubes corresponding to the surface to be fused.
  • all the cubic meshes corresponding to the surface are composed of all the cubes in the basic mesh that envelop the corresponding surface, that is, the envelope mesh group of the surface, or all the cubic meshes corresponding to the surface are composed of the envelope mesh group of each triangular mesh in the basic mesh, that is, the envelope mesh group of the surface, wherein each triangular mesh refers to each triangular mesh in the surface, and the envelope mesh group of the triangular mesh is composed of all the cubes in the basic mesh that envelop the triangular mesh.
  • the target surface is the fused surface, which is the surface reconstructed according to the MC algorithm, and the surface to be fused is composed of triangular meshes, but it is not necessarily the surface reconstructed according to the MC algorithm.
  • calculating the first TSDF value of each corner point corresponding to the surface to be fused may include:
  • the envelope may refer to the minimum range that contains the target determined in three mutually perpendicular normal directions of the cube surface.
  • the minimum range in these three normal directions determines the envelope grid group corresponding to the target, and the determined range is an integer number of cubes.
  • the entire spatial grid group TotalGrid is used to envelope the surface to be fused, and a first envelope grid group CurSurGrid of the surface to be fused is obtained (the first envelope grid group CurSurGrid occupies a part of the entire spatial grid group TotalGrid).
  • the first envelope grid group CurSurGrid can be used as the envelope grid group of the surface to be fused. Therefore, only each corner point in the first envelope grid group CurSurGrid of the surface to be fused is calculated, which can reduce the subsequent calculation amount.
  • the entire spatial grid group TotalGrid can be used to treat each triangulated mesh of the fused surface.
  • the second grid group TriGrid corresponding to each triangular grid of the surface to be fused is enveloping to obtain the second enveloping grid group TriGrid (the second enveloping grid group TriGrid occupies a part of the entire spatial grid group TotalGrid), and the second grid group TriGrid corresponding to each triangular grid on the surface to be fused constitutes the enveloping grid group of the surface to be fused.
  • the number of corner points to be calculated can be further reduced, thereby reducing the amount of subsequent calculations.
  • CUDA Computer Unified Device Architecture
  • a projection method may be used to calculate the minimum distance from the corner point to all triangular meshes on the surface to be fused.
  • determining an envelope mesh group of a surface to be fused, and calculating, for each corner point in the envelope mesh group, a minimum distance from the corner point to all triangular meshes on the surface to be fused may include:
  • S121(a) for all the triangular meshes of the surface to be fused, determine an envelope mesh group that envelops all the triangular meshes, recorded as a first envelope mesh group, and the first envelope mesh group is the envelope mesh group of the surface to be fused.
  • the first envelope mesh group includes multiple cubes, each of which contains 8 corner points.
  • the projection point is calculated for each corner point of the cube, and the projection points of all corner points in the first envelope mesh group are calculated. Then the first distance from each corner point to its respective projection point is calculated, so that the minimum value of the first distance value of each corner point can be calculated, that is, the minimum distance from the corner point to all triangular meshes on the surface to be fused.
  • determining an envelope mesh group of the surface to be fused, and calculating, for each corner point in the envelope mesh group, a minimum distance from the corner point to all triangular meshes on the surface to be fused may include:
  • S121(b) for each triangular mesh on the surface to be fused, determine an envelope mesh group that envelops the triangular mesh, recorded as a second envelope mesh group, and all second envelope mesh groups constitute the envelope mesh group of the surface to be fused.
  • the corresponding triangular mesh refers to the triangular mesh enclosed by the second enveloping mesh group.
  • each corner point in the mesh group corresponding to each triangular mesh correspond to the triangle, and each corner point may correspond to one triangular mesh or may correspond to multiple triangular meshes.
  • the corner points of each cube in the second envelope grid group corresponding to each triangular mesh are the corner points of the cube on the second envelope grid group TriGrid, and the projection points of the corner points to all triangular meshes on the surface to be fused are determined.
  • the surface to be fused contains multiple triangular meshes, each of which corresponds to a second envelope mesh group. All the second envelope mesh groups obtained here constitute the envelope mesh group of the surface to be fused.
  • the first distance from the corner point to the projection point of the triangular mesh enclosed by the second envelope mesh group is calculated, that is, the first distance of each corner point of the cube in the envelope mesh group of the surface to be fused is calculated, and the minimum distance among the first distances of the same corner point is the minimum distance from the corner point to all the triangular meshes on the surface to be fused, that is, the distance value in the TSDF value corresponding to each corner point in the second envelope mesh group is calculated.
  • a cube contains 8 corner points, and each corner point is projected onto a triangular mesh.
  • a corner point may have corresponding projection points in multiple triangular meshes.
  • the first distances from the corner point to all its projection points are calculated.
  • the shortest first distance is the minimum distance from the corner point to all triangular meshes on the surface to be fused.
  • the method of determining the projection point in the above step S121 may include at least one of the following three methods:
  • Method 1 For each corner point, use the centroid method and the perpendicular vector method to obtain the projection point of the corner point to all the triangular meshes on the surface to be fused.
  • a1 v3 ⁇ v1
  • b1 v1 ⁇ v1
  • c1 v2 ⁇ v1
  • a2 v3 ⁇ v2
  • c2 v2 ⁇ v2
  • vector v3 P-A.
  • the calculated t1 and t2 satisfy the above conditions, it means that the projection point of the corner point is within the triangular mesh. Then, the coordinates of the projection point of the corner point are calculated according to formula (1), so that the first distance from the corner point to its projection point can be obtained, and then the minimum distance can be obtained among all the first distances.
  • Method 2 For each mesh (i.e., cube) corresponding to the surface to be fused, determine the intersection of the center line of the opposite faces of the mesh and the corresponding triangular mesh on the surface to be fused, and use the intersection as the projection point of the 8 corner points of the mesh to the corresponding triangular mesh.
  • mesh i.e., cube
  • G z can be calculated according to the above formula.
  • G y can be calculated; if G z and G y are known, G x can be calculated.
  • the calculation result is that the center point of the XY, YZ, and ZX planes in the grid moves along the Z, X, and Y axes to the intersection of the triangular mesh of the surface to be fused.
  • the 8 corner points in a grid only require one projection point, and the values of the projection points are assigned to the 8 corner points, reducing the amount of calculation.
  • the above method must also satisfy the three conditions of 0 ⁇ t1 ⁇ 1, 0 ⁇ t2 ⁇ 1, and t1+t2 ⁇ 1 at the same time. Only when these three conditions are met, the projection point is inside the triangular mesh and is a valid point. Otherwise, it is outside the triangular mesh and is an invalid projection point.
  • Method 3 For each triangular mesh on the surface to be fused, determine the intersection of the mesh surface of the enveloping mesh group (i.e., the second enveloping mesh group) of the triangular mesh and the triangular mesh, and use the intersection as the projection point of the eight corner points of each cube in the enveloping mesh group of the triangular mesh to the triangular mesh.
  • the mesh surface of the enveloping mesh group refers to the surface of each cube in its enveloping mesh group.
  • each triangular mesh on the surface to be fused and the mesh surface of its enveloping mesh group is obtained, as shown in Figure 4(b).
  • the coordinates of the three vertices of each triangular mesh on the surface to be fused are v1(x1, y1, z1), v2(x2, y2, z2), and v3(x3, y3, z3), respectively, and the intersection point w(wx, wy, wz) on the edge can be obtained.
  • the specific solution method is as follows:
  • a point that satisfies y1 ⁇ wy ⁇ y2,z1 ⁇ wz ⁇ z2 is a valid w point, otherwise it is an invalid point and will not be saved.
  • intersection points Z1, Z2, and Z3 are obtained. These points are the projection points of the eight corner points of each grid of the envelope grid group of the surface to be fused, i.e., the second envelope grid group TriGrid, to the triangular grid.
  • the projection points corresponding to the eight corner points of each grid of the second envelope grid group TriGrid which is the grid group of the envelope of the surface to be fused, and the distance between each corner point and the valid projection point are obtained through any one of the projection methods in the above embodiments.
  • the shortest distance is the TSDF distance in the first TSDF value of the corner point.
  • the color value of the projection point of each corner point of each cube in the envelope mesh group of the surface to be fused can be the color average of the three vertices of the triangle mesh corresponding to the projection point, and the normal coordinate of the projection is the unit normal of the triangle.
  • the color value and normal can be used as the color value and normal of the corner point.
  • the projection points of the corner points adjacent to the corner point can be used as the projection points of the corner point.
  • the entire space grid group TotalGrid is used to envelope each triangular mesh of the fused surface to obtain the enveloped grid group, namely the second enveloped grid group TriGrid.
  • the corner points p(x,y,z) that have no valid projection points in the area, their projection points are replaced by the projection points of the adjacent corner points (p1(x- ⁇ ,y,z), p2(x+ ⁇ ,y,z), p3(x,y- ⁇ ,z), p4(x,y+ ⁇ ,z), p5(x,y,z- ⁇ ), p6(x,y,z+ ⁇ ), where ⁇ is the side length of the mesh).
  • the entire fusion process is completed in a fixed world coordinate system, that is, a defined entire space grid group TotalGrid (i.e., the aforementioned basic grid), which is composed of many small grids (i.e., cubes) with equal side lengths.
  • TotalGrid i.e., the aforementioned basic grid
  • small grids i.e., cubes
  • this method is applied to an oral digital impression device.
  • the target surface in the world coordinate system is the overall surface TotalSurface, that is, the surface after each fusion is completed.
  • a new surface to be fused is obtained during the scanning process.
  • the surface to be fused is the current surface CurrentSurface.
  • the surface to be fused needs to be fused to the overall surface, and then a new overall surface is generated to achieve scanning imaging of the teeth.
  • the second TSDF value of each corner point corresponding to the target surface has been stored after the last fusion is completed, and can be used directly when the current surface is fused.
  • the third TSDF value of each corner point corresponding to the fused surface can be obtained based on the first TSDF value and the second TSDF value.
  • determining a third TSDF value of each corner point corresponding to the fused surface according to the first TSDF value and the second TSDF value may include:
  • the target surface can be a fused surface, which can be recorded as the total surface TotalSurface, and the target surface is enveloped with the entire spatial grid group TotalGrid to obtain the target surface's enveloping grid group goalGrid, which is recorded as the third enveloping grid group; the surface to be fused is the surface to be fused to the total surface, which can be recorded as the current surface CurrentSurface.
  • the current surface CurrentSurface is enveloped with the entire spatial grid group TotalGrid to obtain the current surface's enveloping grid group CurSurGrid, which is recorded as the fourth enveloping grid group.
  • the TSDF distance in the third TSDF value of the corner point is obtained according to the TSDF distance in the first TSDF value of the corner point, the TSDF distance in the second TSDF value and the TSDF weight, and the TSDF weight in the third TSDF value of the corner point is obtained according to the TSDF distance in the first TSDF value of the corner point and the TSDF weight.
  • obtaining the TSDF distance in the third TSDF value according to the TSDF distance in the first TSDF value, the TSDF distance in the second TSDF value, and the TSDF weight may include:
  • Add the two products to get the first sum calculate the sum of the TSDF weight nweight2 and the unit threshold in the second TSDF value to get the second sum, calculate the ratio of the first sum to the second sum, and get the TSDF distance fdistance in the third TSDF value.
  • the algorithm formula is as follows:
  • a TSDF weight in a third TSDF value is obtained based on the TSDF distance and the TSDF weight in the first TSDF value, including: when the TSDF distance in the first TSDF value is less than a preset cutoff distance, the TSDF weight in the third TSDF value is obtained based on the TSDF weight in the second TSDF value; when the TSDF distance in the first TSDF value is greater than or equal to the preset truncation distance, the TSDF weight in the second TSDF value is used as the TSDF weight in the third TSDF value.
  • the TSDF distance of the first TSDF value of the corner point is used as the TSDF distance in the third TSDF value of the corner point, and the TSDF weight in the third TSDF value of the corner point is obtained as a preset value, which can be a unit threshold.
  • the TSDF distance fdistance in the third TSDF value of the corner point is the TSDF distance fmindistance in the first TSDF value
  • the TSDF weight nweight of the corner point is the preset threshold, which can be the unit threshold 1.
  • the TSDF value of the fused surface is calculated according to the above method.
  • the coordinates of the triangle mesh vertices of the fused surface and the topological structure of the triangle mesh need to be calculated.
  • the second TSDF value of the remaining corner point is used as its corresponding third TSDF value, and the coordinates of the corresponding corner point remain unchanged.
  • the coordinates of the triangular mesh vertices and their topological structures can be obtained for the common parts and non-common parts (i.e., the newly added parts) of the third envelope mesh group and the fourth envelope mesh group, so as to obtain the coordinates of the triangular mesh vertices and their topological structures of the fused surface.
  • the coordinates of the triangular mesh vertices and the topological structures of the triangular mesh of the fused surface can be calculated based on the third TSDF value and the corner point coordinates, which may include:
  • the edge is determined to be a valid edge, and the coordinates and color values of the triangular mesh vertices of the fused surface are obtained according to the valid edge.
  • the formula can be as follows:
  • ncorner[n 1 ] and ncorner[n 2 ] are two corner points on an edge of the cube.
  • the identifiers of valid edges can be stored in nedge[].
  • the corresponding identifiers of edges and corners can be arranged as shown in Figure 8.
  • the determination of valid edges and the determination of coordinates and color values of triangle mesh vertices are the same as those of the common part above. The difference is that the identifiers of the two corner points of the valid edge, ncorner[2], need to be stored, and then the identifier of the valid edge can be written into the corresponding nedge[].
  • the nMC of the mesh is calculated.
  • nMC is initially 0. If the fdistance of the first corner point ncorner[0] is less than 0, nMC
  • 1; if the fdistance of the second corner point ncorner[1] (i.e., the TSDF distance in the third TSDF value) is less than 0, nMC
  • 2; if the fdistance of the third corner point ncorner[2] is less than 0, nMC
  • 4; if the fdistance of the fourth corner point ncorner[3] is less than 0, nMC
  • 8; if the fdistance of the fifth corner point ncorner[4] is less than 0, nMC
  • 16; if the fdistance of the sixth corner point ncorner[5] is less than 0, nMC
  • 32; if the fdistance of the seventh corner point ncorner[6] is less than 0, nMC
  • 64; if the fdistance of the eighth
  • the processing method is similar to the processing method of the above common part, the difference is that the identifiers of its 8 corner points ncorner[8] also need to be stored.
  • each grid in this step can be calculated in parallel by CUDA, and its coordinates can be directly obtained according to the quantitative position, that is, it is an integer multiple of the side length, and its color value is the color value of its corresponding point.
  • the triangle mesh vertices and the topological structure of the triangle mesh can be represented by the following stored information, which can be stored: corner point identifiers, TSDF values and coordinates; the eight corner point identifiers corresponding to the cube, the corresponding nMC values, the corresponding edge identifiers, and the attributes of the cube; the two corner point identifiers corresponding to the edge, the corresponding triangle mesh vertex coordinates, and the attributes of the edge.
  • the attributes of the cube and the attributes of the edge both indicate whether the corresponding triangle mesh exists.
  • the fused surface is determined based on the vertex coordinates of the triangle mesh of the fused surface and the topological structure of the triangle mesh, including: determining the vertex coordinates of the triangle mesh on the edge corresponding to the triangle mesh according to the attributes of the edge, finding the cube corresponding to the triangle mesh according to the attributes of the cube, and finding the index row number of the mesh through the nMC value corresponding to the cube, finding the edge corresponding to the three vertices of each triangle mesh in the cube through the edge identifier of the row, and corresponding to the vertices in the first step, that is, determining the edge in the cube where the three vertices of each triangle mesh are located, and its coordinates, thereby determining each triangle mesh.
  • the corresponding three-dimensional surface composed of each triangle mesh can be obtained.
  • the fused three-dimensional surface can be determined based on the triangle mesh vertices of the fused surface and the topological result of the triangle mesh.
  • the surface fusion method of the disclosed embodiment calculates the first TSDF value of the surface to be fused through the TSDF algorithm, obtains the second TSDF value of the fused surface, and calculates the third TSDF value of each corner point corresponding to the fused surface based on the first TSDF value and the second TSDF value.
  • the corner point position is used instead of the voxel position, which reduces the amount of calculation for surface fusion.
  • the triangular mesh vertex coordinates of the fused surface and the topological structure of the triangular mesh are calculated to determine the fused surface.
  • the present disclosure proposes a surface fusion device.
  • a surface fusion device 100 includes a memory 10 , a processor 20 , and a computer program stored in the memory 10 .
  • the computer program is executed by the processor 20 , the surface fusion method described above is implemented.
  • the present disclosure also proposes a medical imaging device.
  • a medical imaging device 1000 includes the curved surface fusion device 100 as described above.
  • the surface fusion method, device and medical imaging equipment of the disclosed embodiments calculate the first TSDF value of the surface to be fused through the TSDF algorithm, obtain the second TSDF value of the fused surface, and calculate the third TSDF value of each corner point corresponding to the fused surface based on the first TSDF value and the second TSDF value.
  • the corner point position is used instead of the voxel position, which reduces the amount of calculation for surface fusion.
  • the coordinates of the triangular mesh vertices of the fused surface and the topological structure of the triangular mesh are determined, thereby determining the fused surface.
  • the process of surface fusion calculation is the process of determining the triangular mesh vertices and the triangular mesh topological structure of the fused surface, which is equivalent to merging surface fusion and surface reconstruction, without the need to reconstruct the surface separately, greatly reducing the amount of calculation.
  • computer-readable media include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM).
  • the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.
  • first and second are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.
  • a feature defined as “first” or “second” may explicitly or implicitly include at least one of the features.
  • “plurality” means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
  • the terms “installed”, “connected”, “connected”, “fixed” and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined.
  • installed can be a fixed connection, a detachable connection, or an integral connection
  • it can be a mechanical connection or an electrical connection
  • it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined.
  • the specific meanings of the above terms in the present disclosure can be understood according to specific circumstances.
  • a first feature being “above” or “below” a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium.
  • a first feature being “above”, “above” or “above” a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature.
  • a first feature being “below”, “below” or “below” a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Geometry (AREA)
  • Computer Graphics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

一种曲面融合方法、装置和医疗影像设备，涉及医疗技术领域，曲面融合方法包括：计算待融合曲面对应的每个角点的第一TSDF值；获取目标曲面对应的每个角点的第二TSDF值；根据第一TSDF值和第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值；根据第三TSDF值和角点坐标，确定融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构；根据融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构确定融合后曲面。

Description

曲面融合方法、装置和医疗影像设备
相关申请的交叉引用
本公开要求于2022年09月30日提交的申请号为202211212199.5、名称为“曲面融合方法、装置和医疗影像设备”的中国专利申请的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开涉及医疗技术领域，尤其涉及一种曲面融合方法、装置和医疗影像设备。
背景技术
口内扫描仪在扫描过程中需要边重建、边配准、边融合，且把融合后的曲面实时显示出来。常规方法需要先配准融合再生成曲面，使得常规的方法要么精度达不到要求，要么速度跟不上。
发明内容
本公开的第一个目的在于提出一种曲面融合方法，可以有效地将待融合曲面和目标曲面进行融合，无需单独重新构建曲面，大大减少了计算量。
本公开的第二个目的在于提出一种曲面融合装置。
本公开的第三个目的在于提出一种医疗影像设备。
为达到上述目的，本公开第一方面实施例提出了一种曲面融合方法，所述方法包括：计算待融合曲面对应的每个角点的第一TSDF值；获取目标曲面对应的每个角点的第二TSDF值；根据所述第一TSDF值和所述第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值；根据角点的第三TSDF值和角点坐标，确定所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构；根据所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构确定所述融合后曲面。
另外，本公开上述实施例提出的曲面融合方法还可以具有如下附加的技术特征：
根据本公开的一个实施例，计算待融合曲面对应的每个角点的第一TSDF值，包括：确定所述待融合曲面的包络网格组，并针对所述包络网格组中的每个角点，计算该角点到所述待融合曲面上所有三角网格的最小距离；将所述最小距离作为所述待融合曲面对应的每个角点的第一TSDF值中的TSDF距离。
根据本公开的一个实施例，确定所述待融合曲面的包络网格组，并针对所述包络网格 组中的每个角点，计算该角点到所述待融合曲面上所有三角网格的最小距离，包括：针对所述待融合曲面的所有三角网格，确定包络所有三角网格的包络网格组，记为第一包络网格组，所述第一包络网格组即为所述待融合曲面的包络网格组；针对所述第一包络网格组中每个立方体的每个角点，确定该角点到所述待融合曲面上所有三角网格的投影点，并计算该角点到其各投影点的第一距离；将每个角点的第一距离中的最小值作为该角点到所述待融合曲面上所有三角网格的最小距离；
或，针对所述待融合曲面上的每个三角网格，确定包络该三角网格的包络网格组，记为第二包络网格组，所有第二包络网格组即组成所述待融合曲面的包络网格组；针对每个第二包络网格组中每个立方体的每个角点，确定该角点到对应三角网格的投影点，并计算该角点到其各投影点的第一距离；将同一角点对应的第一距离中的最小值作为对应角点到所述待融合曲面上所有三角网格的最小距离；其中，用于计算所述第一距离的投影点为落入相应三角网格内的投影点。
根据本公开的一个实施例，确定投影点的方式包括下述方式中的至少一者：方式一：针对每个角点，采用重心法和垂直向量法得到该角点到所述待融合曲面上所有三角网格的投影点；方式二：针对所述待融合曲面对应的每个网格，确定该网格相对面的中心连线与所述待融合曲面上相应三角网格的交点，并将该交点作为该网格8个角点到所述相应三角网格的投影点；方式三：针对所述待融合曲面上每个三角网格，确定该三角网格的包络网格组的网格面与该三角网格的交点，并将交点作为该三角网格的包络网格组中各网格的8个角点到该三角网格的投影点。
根据本公开的一个实施例，当存在到所述待融合曲面上所有三角网格均没有投影点的角点时，将该角点相邻角点的投影点作为该角点的投影点。
根据本公开的一个实施例，根据所述第一TSDF值和所述第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值，包括：确定所述目标曲面的包络网格组，记为第三包络网格组，并记所述待融合曲面的包络网格组为第四包络网格组；针对所述第三包络网格组和所述第四包络网格组的公共部分的每个角点，根据该角点的第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF距离，并根据该角点的第一TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF权重；针对所述第四包络网格组相对所述第三包络网格组的新增部分的每个角点，将该角点的第一TSDF值作为该角点的第三TSDF值中的TSDF距离，并得到该角点的第三TSDF值中的TSDF权重为预设值。
根据本公开的一个实施例，根据第一TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF权重，包括：在第一TSDF值中的TSDF距离小于预设截断距离时，根据第二 TSDF值中的TSDF权重得到第三TSDF值中的TSDF权重；在第一TSDF值中的TSDF距离大于或等于所述预设截断距离时，将第二TSDF值中的TSDF权重作为第三TSDF值中的TSDF权重。
根据本公开的一个实施例，根据第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF距离，包括：计算所述第二TSDF值中的TSDF距离和TSDF权重的乘积，得到第一乘积；计算所述第一TSDF值中的TSDF距离与TSDF的单位阈值的乘积，得到第二乘积；计算所述第一乘积与所述第二乘积的加和，得到第一和值；计算所述第二TSDF值中的TSDF权重与所述单位阈值的加和，得到第二和值；计算所述第一和值与所述第二和值的比值，得到所述第三TSDF值中的TSDF距离；其中，根据第二TSDF值中的TSDF权重得到第三TSDF值中的TSDF权重，包括：将所述第二和值作为所述第三TSDF值中的TSDF权重。
根据本公开的一个实施例，根据角点的第三TSDF值和角点坐标，确定所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，包括：根据所述角点坐标和所述第三TSDF值中的TSDF距离，确定所述融合后曲面的三角网格顶点坐标；根据所述第三TSDF值中的TSDF距离确定立方体网格对应的nMC值，根据nMC值和三角网格的顶点坐标，确定所述融合后曲面的三角网格的拓扑结构。
本公开实施例的曲面融合方法，通过TSDF算法计算待融合曲面的第一TSDF值，获取已融合的曲面的第二TSDF值，根据第一TSDF值和第二TSDF值，计算融合后曲面对应的每个角点的第三TSDF值。用角点位置代替体素位置，较少了曲面融合的计算量。且根据第三TSDF值和角点坐标，确定融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，从而确定了融合后曲面。曲面融合计算的过程就是确定融合后曲面的三角网格顶点及三角网格拓扑结构的过程，相当于将曲面融合和曲面重建合并，无需单独重新构建曲面，大大减少计算量。
为达到上述目的，本公开第二方面实施例提出了一种曲面融合装置，包括存储器、处理器以及存储在存储器上的计算机程序，计算机程序被处理器执行时，实现如上述的曲面融合方法。
为达到上述目的，本公开第三方面实施例提出了一种医疗影像设备，包括如上述的曲面融合装置。
本公开附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本公开的实践了解到。
附图说明
图1是本公开一个实施例的曲面融合方法的流程图；
图2是本公开一个实施例的计算待融合曲面对应的每个角点的第一TSDF值的流程图；
图3(a)是本公开一个实施例的计算角点到待融合曲面上所有三角网格的最小距离的流程图；
图3(b)是本公开另一个实施例的计算角点到待融合曲面上所有三角网格的最小距离的流程图；
图4(a)是本公开一个实施例的计算三角网格投影点的示意图；
图4(b)是本公开另一个实施例的计算三角网格投影点的示意图；
图5是本公开一个实施例的确定融合后曲面对应的每个角点的第三TSDF值的流程图；
图6是本公开一个实施例的计算第三TSDF值中的TSDF距离的流程图；
图7是本公开一个实施例的计算融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构的流程图；
图8是本公开一个实施例的网格块的边和角对应标识的示意图；
图9是本公开一个实施例的曲面融合装置的结构示意图；
图10是本公开一个实施例的医疗影像设备的结构示意图。
具体实施方式
下面详细描述本公开的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本公开，而不能理解为对本公开的限制。
下面将结合说明书附图以及具体的实施方式对本公开实施例的曲面融合方法、装置和医疗影像设备进行详细地说明。
为了便于理解本公开的应用场景，举例来说，在三维物体的扫描成像过程中，如口腔数字印模仪对口腔内牙齿进行扫描成像的过程，相对来说，先扫描生成的曲面作为目标曲面，后扫描生成的曲面作为待配准融合曲面，待配准融合曲面配准融合到目标曲面则可以生成新的目标曲面，后面扫描的曲面再融合配准到新的目标曲面上，这样进行多轮的配准融合，生成更大的曲面，从而可以完成三维物体的扫描成像。本公开实施例的曲面融合方法即可应用于上述融合过程。
需要注意的是，本公开的实施例可以应用的具体领域并不限于上述三维物体的扫描成像领域，在其他领域且可以满足融合过程中相应条件的情况下，都可以采用本公开实施例的曲面融合方法对任意两个曲面进行融合。
图1是本公开一个实施例的曲面融合方法的流程图。
在本公开的一个实施例中，如图1所示，曲面融合方法包括：
S1，计算待融合曲面对应的每个角点的第一TSDF值。
具体地，本公开采用TSDF(Truncated Signed Distance Function，基于截断地带符号距离函数)方法处理相机(如口腔数字印模仪等医疗影像设备)扫描出的曲面，TSDF方法是将空间划分为等边的可无限延伸的网格块(即立方体)。角点的TSDF值包括TSDF距离即对应截断距离函数的值和TSDF权重即相对应的权重，权重可以表示对应点的采样质量。
计算曲面对应每个角点的TSDF值后，可通过线性插值的方法求出相应曲面的三角网格位置，可以用三角网格顶点坐标和该三角网格的拓扑结构表示。
若待融合曲面是第一个曲面，则将该待融合曲面作为目标曲面，并根据该目标曲面的MC(Marching cubes，移动立方体)网格划分作为基础模型网格块的划分，并进行延拓形成基础网格，并且可以计算目标曲面对应的所有立方体网格的每个角点的第一TSDF值。若待融合曲面不是第一个曲面，则根据上述的基础模型网络块对待融合曲面进行处理，得到待融合曲面对应的所有立方体的每个角点的第一TSDF值。
需要说明的是，曲面对应的所有立方体网格由基础网格中包络相应曲面的所有立方体组成即曲面的包络网格组，或者曲面对应的所有立方体网格由基础网格中各个三角网格的包络网格组组成也即曲面的包络网格组，其中，各个三角网格是指曲面中的各个三角网格，三角网格的包络网格组是由基础网格中包络该三角网格的所有立方体组成。
另外，需要说明的是，目标曲面是融合后的曲面，是根据MC算法重建的曲面，待融合曲面是由三角网格组成，但其不一定是按照MC算法重建的曲面。
在本公开的一个实施例中，如图2所示，计算待融合曲面对应的每个角点的第一TSDF值，可包括：
S11，确定待融合曲面的包络网格组。
具体地，包络可以是指从立方体表面的相互垂直的三个法线方向上，分别确定包含目标的最小范围，这三个法线方向上的最小范围确定了目标所对应的包络网格组，所确定的范围为整数个立方体。
一种实施方式中，在对包络网格组中的每个角点计算之前，用整个空间网格组TotalGrid对待融合曲面进行包络，得到待融合曲面的第一包络网格组CurSurGrid(第一包络网格组CurSurGrid占整个空间网格组TotalGrid的一部分)，第一包络网格组CurSurGrid可以作为待融合曲面的包络网格组，由此，可只对待融合曲面的第一包络网格组CurSurGrid中每个角点进行计算，可以减少后续的计算量。
另一种实施方式中，还可以用整个空间网格组TotalGrid对待融合曲面的每个三角网 格进行包络，得到待融合曲面的每个三角网格对应的第二包络网格组TriGrid(第二包络网格组TriGrid占整个空间网格组TotalGrid的一部分)，待融合曲面上的每个三角网格对应的第二包括网格组TriGrid组成待融合曲面的包络网格组。由此，可以进一步减少所要计算的角点数量，从而可以减少后续的计算量。
待融合曲面的所有三角网格之间可以并行计算，该过程可以用CUDA(Compute Unified Device Architecture，统一计算设备架构)加速，CUDA是一种并行计算平台和编程模型，通过利用图形处理器(GPU)的处理能力，可以大幅提升计算性能。
S12，针对包络网格组中的每个角点，计算该角点到待融合曲面上所有三角网格的最小距离。
具体地，针对待融合曲面的包络网格组中的每个角点，可采用投影法计算该角点到待融合曲面上所有三角网格的最小距离。
在本公开的一个实施例中，如图3(a)所示，确定待融合曲面的包络网格组，并针对包络网格组中的每个角点，计算角点到待融合曲面上所有三角网格的最小距离，可包括：
S121(a)，针对待融合曲面的所有三角网格，确定包络所有三角网格的包络网格组，记为第一包络网格组，第一包络网格组即为待融合曲面的包络网格组。
S122(a)，针对第一包络网格组中每个立方体的每个角点，确定该角点到待融合曲面上所有三角网格的投影点，并计算该角点到其各投影点的第一距离，其中，用于计算第一距离的投影点为落入相应三角网格内的投影点。
S123(a)，将每个角点的第一距离中的最小值作为该角点到待融合曲面上所有三角网格的最小距离。
第一包络网格组中包括多个立方体，每个立方体又包含8个角点，针对立方体的每个角点计算投影点，则计算出了第一包络网格组中所有角点的投影点。然后计算每个角点到各自的投影点的第一距离，从而可计算出每个角点第一距离值中的最小值，即该角点到待融合曲面上所有三角网格的最小距离。
另一种实施例中，如图3(b)所示，确定待融合曲面的包络网格组，并针对包络网格组中的每个角点，计算角点到待融合曲面上所有三角网格的最小距离，可包括：
S121(b)，针对待融合曲面上的每个三角网格，确定包络该三角网格的包络网格组，记为第二包络网格组，所有第二包络网格组即组成所述待融合曲面的包络网格组。
S122(b)，针对每个第二包络网格组中每个立方体的每个角点，确定该角点到对应三角网格的投影点，并计算该角点到其各投影点的第一距离。
S123(b)，将同一角点对应的第一距离中的最小值作为对应角点到待融合曲面上所有三角网格的最小距离，其中，用于计算第一距离的投影点为落入相应三角网格内的投影点。 对应三角网格是指该第二包络网格组所包络的三角网格。
需要说明的是，每个三角网格对应的网格组中的角点与这个三角形对应，每个角点可能会对应一个三角网格也可能对应多个三角网格。
具体地，每个三角网格对应的第二包络网格组中的每个立方体的角点为第二包络网格组TriGrid上的立方体的角点，并确定角点到待融合曲面上所有三角网格的投影点。
待融合曲面包含多个三角网格，每个三角网格对应一个第二包络网格组，这里得到的所有第二包络网格组组成了待融合曲面的包络网格组。针对每个第二包络网格组中的每个角点，来计算该角点到该第二包络网格组所包络的三角网格的投影点第一距离，即计算出了待融合曲面的包络网格组中立方体的每个角点的第一距离，同一角点的第一距离中最小距离即为该角点到所述待融合曲面上所有三角网格的最小距离，即计算出第二包络网格组中每个角点对应的TSDF值中的距离值。
具体地，一个立方体中包含8个角点，对每一个角点进行到三角网格的投影。先判断投影点是否落在三角网格内，如果落在三角网格内，再求取投影点，如果不落在三角网格内，则没有对应的投影点。通过先判断后求解，可以减少计算量，加快计算速度。
进一步具体地，一个角点可在多个三角网格中有其对应的投影点，计算角点到其所有投影点的第一距离，在所有第一距离中，距离最短的第一距离即为该角点的到待融合曲面上所有三角网格的最小距离。
需要说明的是，由于每个角点的计算可以并行进行，每个立方体的计算也可以并行进行，该过程可以用CUDA方法加速处理。
在本公开的一个实施例中，上述步骤S121中确定投影点的方式可包括下述三种方式中的至少一者：
方式一：针对每个角点，采用重心法和垂直向量法得到该角点到待融合曲面上所有三角网格的投影点。
具体地，假设三角形(即三角网格)三个顶点分别是A,B,C，平面内由A点起始的两个向量分别为：v1＝B-A，v2＝C-A，根据重心法得知，三角形内任意一点PP：
PP＝A+t1×v1+t2×v2        (1)
其中t1、t2为向量系数，且当同时满足0≤t1≤1,0≤t2≤1,t1+t2＜1三个条件时，PP在三角形ABC内。如P的投影点为PP，那么向量v＝P-PP与三角形ABC垂直，依据垂直向量法两个垂直的向量点积为0，可得：
v·v1＝0，v·v2＝0，
以上公式推导可得：
其中，a1＝v3·v1，b1＝v1·v1，c1＝v2·v1，a2＝v3·v2，c2＝v2·v2，向量v3＝P-A。
如果通过计算得到的t1,t2满足上述条件，则表明该角点的投影点在三角网格内，再根据公式(1)计算得到该角点的投影点的坐标，从而可得到该角点到其投影点的第一距离，进而可在所有第一距离中得到最小距离。
方式二：针对待融合曲面对应的每个网格(即立方体)，确定该网格相对面的中心连线与待融合曲面上相应三角网格的交点，并将该交点作为该网格8个角点到相应三角网格的投影点。
具体地，如图4(a)所示，曲面上任意三角形ABC上任一点G，有：
Gx＝Bx+t1×CBx+t2×ABx
Gy＝By+t1×CBy+t2×ABy
Gz＝Bz+t1×CBz+t2×ABz
其中，
进行叠加计算，Gx是XY平面上X方向上的大小，已知Gx＝((nNumx+0.5)*每个网格的边长)，Gy是XY平面上Y方向上的大小，Gy＝((nNumy+0.5)*每个网格的边长)，nNumx和nNumy为相应方向的从原点开始的角点序号，可以根据Gx和Gy的方程求出如下：
其中，
则可以根据上述公式求出Gz。
同理，已知Gx，Gz，可求出Gy；已知Gz，Gy，可求出Gx。
其计算结果为网格中XY、YZ、ZX平面的中心点以Z、X、Y轴移动到待融合曲面的三角网格的交点，这样一个网格中的8个角点只需求一次投影点即可，得到的投影点的值赋给8个角点，减少了计算量。
需要说明的是，上述方式也必须同时满足0≤t1≤1,0≤t2≤1,t1+t2＜1三个条件时，只有满足这三个条件时，该投影点才在三角网格内，为有效的点，否则在三角网格外，为无效的投影点。
方式三：针对待融合曲面上每个三角网格，确定该三角网格的包络网格组(即第二包络网格组)的网格面与该三角网格的交点，并将交点作为该三角网格的包络网格组中各立方体的8个角点到该三角网格的投影点。其中，包络网格组的网格面是指其包络网格组中各个立方体的面。
具体地，求取待融合曲面上每个三角网格和其包络网格组的网格面的交线，如图4(b)所示，待融合曲面上每个三角网格三个顶点坐标分别为v1(x1,y1,z1)，v2(x2,y2,z2)，v3(x3,y3,z3)，可求得边上交点w(wx,wy,wz)。具体求解方法如下：
对待融合曲面的每个三角网格包络的第二包络网格组TriGrid的每个面进行计算，以面S(sx＝x)为例，交点w为：


且同时满足y1≤wy≤y2,z1≤wz≤z2的点是有效的w点，否则是无效点不保存。
如图4(b)所示，X轴向每个面计算得到结果线段w2-w4(x＝wx1)，w1-w7(图中w7和v2同点，x＝wx2)，y轴向每个面计算得到结果线段w3-w6(y＝wy1)，w5-w8(图中w8和v3同点，y＝wy2)，z轴向每个面计算得到结果为没有有效点。
求取待融合曲面上每个三角网格和其包络网格组的网格面的交线之后，求w点组成的线段的交点z，具体计算过程如下：
以X轴向每个面的线段为例L1(L1x＝wx)和y轴向每个面计算得到结果L2((x1,y1,z1)-(x2,y2,z2))为例求在其上的交点：
以图4(b)为例，求得交点Z1,Z2,Z3。该点即为待融合曲面的包络的网格组即第二包络网格组TriGrid的各网格的8个角点到该三角网格的投影点。
S13，将最小距离作为待融合曲面对应的每个角点的第一TSDF值中的TSDF距离。
具体地，通过上述实施例的任意一种投影法得到待融合曲面的包络的网格组即第二包络网格组TriGrid的各网格的8个角点对应的投影点，以及每个角点与有效投影点之间的距离，在角点和其多个投影点的所有第一距离中，距离最短的即为该角点的第一TSDF值中的TSDF距离。
需要说明的是，待融合曲面的包络网格组中每个立方体的每个角点的投影点的颜色值可为该投影点对应三角网格三个顶点的颜色均值，投影的法线坐标为三角形的单位法线。其中，该颜色值和法线可以作为角点的颜色值和法线。
在整个过程中，待融合曲面的上有可能存在极少数角点一个有效投影点都没有。在本公开的一个实施例中，当存在到待融合曲面上所有三角网格均没有投影点的角点时，可将该角点相邻角点的投影点作为该角点的投影点。
具体的，用整个空间网格组TotalGrid对待融合曲面的每个三角网格进行包络，得到包络的网格组即第二包络网格组TriGrid，对该区域内没有有效投影点的角点p(x,y,z)，其投影点用相邻角点(p1(x-Δ,y,z),p2(x+Δ,y,z),p3(x,y-Δ,z),p4(x,y+Δ,z),p5(x,y,z-Δ),p6(x,y,z+Δ)，其中Δ为网格的边长)的投影点代替。
S2，获取目标曲面对应的每个角点的第二TSDF值。
具体的，整个融合过程是在一个固定的世界坐标系下完成的，即一个划定好的整个空间网格组TotalGrid(即前述基础网格)，其由许多边长相等的小网格(即立方体)组合而成。
例如，该方式应用于口腔数字印模仪上，在世界坐标系下的目标曲面为总体曲面TotalSurface，即每次融合完成后的曲面，扫描过程中获得新的待融合曲面，待融合曲面为当前曲面CurrentSurface，待融合曲面需要融合到总体曲面上，然后生成新的总体曲面，从而实现对牙齿的扫描成像。
其中，目标曲面对应的每个角点的第二TSDF值，均已在上一次融合完成后储存，在当前曲面融合时，可直接使用。
S3，根据第一TSDF值和第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值。
具体的，按照上述方法得到待融合曲面每个角点的第一TSDF值和目标曲面的第二TSDF值后，可根据第一TSDF值和第二TSDF值，得到融合后的曲面对应的每个角点的第三TSDF值。
在本公开的一个实施例中，如图5所示，根据第一TSDF值和第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值，可包括：
S31，确定目标曲面的包络网格组，记为第三包络网格组，并记待融合曲面的包络网格组为第四包络网格组。
具体的，目标曲面可以为已融合的曲面，可记为总体曲面TotalSurface，用整个空间网格组TotalGrid对目标曲面进行包络，得到目标曲面的包络网格组goalGrid，记为第三包络网格组；待融合曲面为将要融合到总体的曲面，可记为当前曲面CurrentSurface。用整个空间网格组TotalGrid对当前曲面CurrentSurface进行包络，得到当前曲面的包络网格组CurSurGrid，记为第四包络网格组。
S32，针对第三包络网格组和第四包络网格组的公共部分(也即目标曲面和待融合曲面的公共部分)的每个角点，根据该角点的第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF距离，并根据该角点的第一TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF权重。
在本公开的一个实施例中，如图6所示，根据第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF距离，可包括：
S321，计算第二TSDF值中的TSDF距离和TSDF权重的乘积，得到第一乘积。
S322，计算第一TSDF值中的TSDF距离与TSDF的单位阈值的乘积，得到第二乘积。
S323，计算第一乘积与第二乘积的加和，得到第一和值。
S324，计算第二TSDF值中的TSDF权重与单位阈值的加和，得到第二和值。
S425，计算第一和值与第二和值的比值，得到第三TSDF值中的TSDF距离。
具体的，先计算第二TSDF值中的TSDF距离fdistance2与TSDF权重nweight2的乘积，并计算第一TSDF值中的TSDF距离fmindistance与TSDF的单位阈值的乘积，这里的TSDF的单位阈值为1，其中，每个角点的初始TSDF权重nweight都是0。将两个乘积加和得到第一和值，计算第二TSDF值中的TSDF权重nweight2与单位阈值的加和，得到第二和值，计算第一和值与第二和值的比值，得到第三TSDF值中的TSDF距离fdistance。算法公式如下：
在本公开的一个实施例中，根据第一TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF权重，包括：在第一TSDF值中的TSDF距离小于预设截断距离时，根据第二TSDF值中的TSDF权重得到第三TSDF值中的TSDF权重；在第一TSDF值中的TSDF距离大于或等于预设截断距离时，将第二TSDF值中的TSDF权重作为第三TSDF值中的TSDF权重。
具体的，待融合曲面的第一TSDF值中的TSDF距离小于预设截断距离时，第三TSDF值中的TSDF权重为目标曲面的第二TSDF值中的TSDF权重加单位阈值得到，如nweight＝nweight2+1；在第一TSDF值中的TSDF距离大于或等于预设截断距离时，第三TSDF值中的TSDF权重为目标曲面的第二TSDF值中的TSDF权重，即nweight＝nweight2。
S33，针对第四包络网格组相对第三包络网格组的新增部分(也即待融合曲面相对目标曲面的新增部分)的每个角点，将该角点的第一TSDF值的TSDF距离作为该角点的第三TSDF值中的TSDF距离，并得到该角点的第三TSDF值中的TSDF权重为预设值，该预设值可以为单位阈值。
具体的，针对第四包络网格组即待融合曲面的包络网格组CurSurGrid相对第三包络网格组即目标曲面的包络网络组goalGrid的新增部分每个角点，该角点的第三TSDF值中的TSDF距离fdistance为第一TSDF值中的TSDF距离fmindistance，该角点的TSDF权重nweight为预设阈值，预设阈值可为单元阈值1。算法公式如下：
fdis tance＝f min dis tan ce
nweight＝1
根据上述方法计算融合后曲面的TSDF值。融合后曲面除需要融合后曲面的TSDF值，还需计算融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构。
第三包络网格组中若存在去除上述公共部分的其余部分，则仍然按照目标曲面原有部分对应的数据进行确定。也就是说，将其余部分角点的第二TSDF值作为其对应的第三TSDF值，相应角点的坐标也未变。
S4，根据第三TSDF值和角点坐标，确定融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构。
具体的，可针对第三包络网格组和第四包络网格组的公共部分和非公共部分(即新增部分)进行三角网格顶点坐标及其拓扑结构的求取，得到融合后曲面的三角网格顶点坐标及其拓扑结构。在本公开的一个实施例中，如图7所示，根据第三TSDF值和角点坐标，计算融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，可包括：
S41，根据角点坐标和第三TSDF值中的TSDF距离，确定融合后曲面的三角网格顶点坐标。
具体的，针对第三包络网格组和第四包络网格组的公共部分的每条边，如果该边的两个角点都有效(即存在第三TSDF值)，且两个角点的fdistance(即第三TSDF值中的TSDF距离)的符号相反(例如，立方体上同一条边的两个角点的TSDF距离符号相反)，则确定该边为有效边，根据该有效边求取融合后曲面的三角网格顶点的坐标、颜色值，公式可如下：





其中，ncorner[n1]和ncorner[n2]为立方体一条边上的两个角点。
同时，有效边的标识可以存入nedge[]中。其中，边和角点的对应标识可按照如图8所示排布。
针对第四包络网格组相对第三包络网格组的新增部分的每条边，有效边的判定和三角网格顶点的坐标、颜色值的求取与上述公共部分的相同。区别在于，还需存储有效边2个角点的标识ncorner[2]，进而才能把该有效边的标识写入的对应nedge[]中。
S42，根据第三TSDF值中的TSDF距离确定立方体网格对应的nMC值，根据nMC值和三角网格的顶点坐标，确定融合后曲面的三角网格的拓扑结构。
具体的，针对第三包络网格组和第四包络网格组的公共部分的每个网格，如果该网格的8个角点都有效，则求取该网格的nMC。
例如，nMC初始为0，如果第一个角点ncorner[0]的fdistance小于0，则nMC|＝1；如果第二个角点ncorner[1]的fdistance(即第三TSDF值中的TSDF距离)小于0，则nMC|＝2；如果第三个角点ncorner[2]的fdistance小于0，则nMC|＝4；如果第四个角点ncorner[3]的fdistance小于0，则nMC|＝8；如果第五个角点ncorner[4]的fdistance小于0，则nMC|＝16；如果第六个角点ncorner[5]的fdistance小于0，则nMC|＝32；如果第七个角点ncorner[6]的fdistance小于0，则nMC|＝64；如果第八个角点ncorner[7]的 fdistance小于0，则nMC|＝128。求得的nMC如果在0-255之间，说明网格是有效的，即对应MC查找表中的各行。
针对第三包络网格组相对第四包络网格组的新增部分的每个网格，处理方式与上述公共部分的处理方式类似，区别在于还需存储其8个角点的标识ncorner[8]。
需要说明的是，该步骤的每个网格都可用CUDA并行计算，其坐标根据数量位置直接能得到，即是边长的整数倍，其颜色值为其对应点的颜色值。
S5，根据融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构确定融合后曲面。
具体的，三角网格顶点，以及三角网格的拓扑结构可用以下存储的信息表示，可以存储：角点标识、TSDF值及坐标；立方体对应的8个角点标识、对应nMC值、对应边标识、立方体的属性；边对应两个角点标识、对应三角网格顶点坐标、以及边的属性。其中，立方体的属性和边的属性均表示着是否对应存在着三角网格。
进一步的，根据融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构确定融合后曲面，包括：根据边的属性，确定对应有三角网格的边上三角网格的顶点坐标，根据立方体的属性，找到对应三角网格的立方体，并通过立方体对应的nMC值找到该网格的索引行数，通过该行的边标识找到每个三角网格的三个顶点在该立方体中对应的边，与第一步的顶点对应上，即确定了每个三角网格的三个顶点所在立方体中的边，及其坐标，从而确定了每个三角网格。由此，可得到对应的由每个三角网格组成的三维曲面。
在该实施例中，上述用到的MC曲面顶点标识的索引表(该表属于现有技术，故省略该索引表的具体内容)共256行，即有256种网格(由于第一行和最后一行是一样的，实际上是255种网格，分成256种是为了2的N次幂的快速计算)，每一行有16个标识(由于MC单个网格中最多有5个三角网格，故有5*3＝15个三角网格顶点的边标识，故有效的最多是前15个边标识，最后一个标识都是-1，分成16是为了2的N次幂的快速计算)。由于一个三角网格最多有12条边，故标识是0,1,2……，11，标识为-1说明无效。
通过上述方法，基于融合后曲面的三角网格顶点和该三角网格的拓扑结果可确定融合后的立体三维曲面。
本公开实施例的曲面融合方法，通过TSDF算法计算待融合曲面的第一TSDF值，获取已融合的曲面的第二TSDF值，根据第一TSDF值和第二TSDF值，计算融合后曲面对应的每个角点的第三TSDF值。用角点位置代替体素位置，减少了曲面融合的计算量。且根据第三TSDF值和角点坐标，以及目标曲面的三角网格顶点坐标和该三角网格的拓扑结构，计算得到融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，从而确定了融合后曲面。可递进式存储三角网格顶点及其拓扑结构，即可以多次配准融合但不需要多次曲面重建，曲面的之前存储数据包括拓扑结构不会变，相当于将曲面融合和曲面重建合并，大大 减少计算量。
基于上述的曲面融合方法，本公开提出了一种曲面融合装置。
在本公开的一个实施例中，如图9所示，曲面融合装置100包括存储器10、处理器20以及存储在存储器10上的计算机程序，计算机程序被处理器20执行时，实现如上述的曲面融合方法。
基于上述的曲面融合方法，本公开还提出了一种医疗影像设备。
在本公开的一个实施例中，如图10所示，医疗影像设备1000包括如上述的曲面融合装置100。
本公开实施例的曲面融合方法、装置和医疗影像设备，通过TSDF算法计算待融合曲面的第一TSDF值，获取已融合的曲面的第二TSDF值，根据第一TSDF值和第二TSDF值，计算融合后曲面对应的每个角点的第三TSDF值。用角点位置代替体素位置，较少了曲面融合的计算量。且根据第三TSDF值和角点坐标，确定融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，从而确定了融合后曲面。曲面融合计算的过程就是确定融合后曲面的三角网格顶点及三角网格拓扑结构的过程，相当于将曲面融合和曲面重建合并，无需单独重新构建曲面，大大减少计算量。
需要说明的是，在流程图中表示或在此以其他方式描述的逻辑和/或步骤，例如，可以被认为是用于实现逻辑功能的可执行指令的定序列表，可以具体实现在任何计算机可读介质中，以供指令执行***、装置或设备(如基于计算机的***、包括处理器的***或其他可以从指令执行***、装置或设备取指令并执行指令的***)使用，或结合这些指令执行***、装置或设备而使用。就本说明书而言，“计算机可读介质”可以是任何可以包含、存储、通信、传播或传输程序以供指令执行***、装置或设备或结合这些指令执行***、装置或设备而使用的装置。计算机可读介质的更具体的示例(非穷尽性列表)包括以下：具有一个或多个布线的电连接部(电子装置)，便携式计算机盘盒(磁装置)，随机存取存储器(RAM)，只读存储器(ROM)，可擦除可编辑只读存储器(EPROM或闪速存储器)，光纤装置，以及便携式光盘只读存储器(CDROM)。另外，计算机可读介质甚至可以是可在其上打印所述程序的纸或其他合适的介质，因为可以例如通过对纸或其他介质进行光学扫描，接着进行编辑、解译或必要时以其他合适方式进行处理来以电子方式获得所述程序，然后将其存储在计算机存储器中。
应当理解，本公开的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中，多个步骤或方法可以用存储在存储器中且由合适的指令执行***执行的软件或固件来实现。例如，如果用硬件来实现，和在另一实施方式中一样，可用本领域公知的下列技术中的任一项或他们的组合来实现：具有用于对数据信号实现逻辑功能的逻辑门电路 的离散逻辑电路，具有合适的组合逻辑门电路的专用集成电路，可编程门阵列(PGA)，现场可编程门阵列(FPGA)等。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本公开的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任何的一个或多个实施例或示例中以合适的方式结合。
在本公开的描述中，需要理解的是，术语“中心”、“纵向”、“横向”、“长度”、“宽度”、“厚度”、“上”、“下”、“前”、“后”、“左”、“右”、“竖直”、“水平”、“顶”、“底”“内”、“外”、“顺时针”、“逆时针”、“轴向”、“径向”、“周向”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本公开和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本公开的限制。
此外，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。在本公开的描述中，“多个”的含义是至少两个，例如两个，三个等，除非另有明确具体的限定。
在本公开中，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”、“固定”等术语应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或成一体；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系，除非另有明确的限定。对于本领域的普通技术人员而言，可以根据具体情况理解上述术语在本公开中的具体含义。
在本公开中，除非另有明确的规定和限定，第一特征在第二特征“上”或“下”可以是第一和第二特征直接接触，或第一和第二特征通过中间媒介间接接触。而且，第一特征在第二特征“之上”、“上方”和“上面”可是第一特征在第二特征正上方或斜上方，或仅仅表示第一特征水平高度高于第二特征。第一特征在第二特征“之下”、“下方”和“下面”可以是第一特征在第二特征正下方或斜下方，或仅仅表示第一特征水平高度小于第二特征。
尽管上面已经示出和描述了本公开的实施例，可以理解的是，上述实施例是示例性的，不能理解为对本公开的限制，本领域的普通技术人员在本公开的范围内可以对上述实施例进行变化、修改、替换和变型。

Claims (11)

  1. 一种曲面融合方法，其特征在于，所述方法包括：
    计算待融合曲面对应的每个角点的第一TSDF值；
    获取目标曲面对应的每个角点的第二TSDF值；
    根据所述第一TSDF值和所述第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值；
    根据角点的第三TSDF值和角点坐标，确定所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构；
    根据所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构确定所述融合后曲面。
  2. 如权利要求1所述的曲面融合方法，其特征在于，计算待融合曲面对应的每个角点的第一TSDF值，包括：
    确定所述待融合曲面的包络网格组，并针对所述包络网格组中的每个角点，计算该角点到所述待融合曲面上所有三角网格的最小距离；
    将所述最小距离作为所述待融合曲面对应的每个角点的第一TSDF值中的TSDF距离。
  3. 如权利要求2所述的曲面融合方法，其特征在于，确定所述待融合曲面的包络网格组，并针对所述包络网格组中的每个角点，计算该角点到所述待融合曲面上所有三角网格的最小距离，包括：
    针对所述待融合曲面的所有三角网格，确定包络所有三角网格的包络网格组，记为第一包络网格组，所述第一包络网格组即为所述待融合曲面的包络网格组；
    针对所述第一包络网格组中每个立方体的每个角点，确定该角点到所述待融合曲面上所有三角网格的投影点，并计算该角点到其各投影点的第一距离；
    将每个角点的第一距离中的最小值作为该角点到所述待融合曲面上所有三角网格的最小距离；
    或，
    针对所述待融合曲面上的每个三角网格，确定包络该三角网格的包络网格组，记为第二包络网格组，所有第二包络网格组即组成所述待融合曲面的包络网格组；
    针对每个第二包络网格组中每个立方体的每个角点，确定该角点到对应三角网格的投影点，并计算该角点到其各投影点的第一距离；
    将同一角点对应的第一距离中的最小值作为对应角点到所述待融合曲面上所有三角网格的最小距离；
    其中，用于计算所述第一距离的投影点为落入相应三角网格内的投影点。
  4. 如权利要求3所述的曲面融合方法，其特征在于，确定投影点的方式包括下述方式中的至少一者：
    方式一：针对每个角点，采用重心法和垂直向量法得到该角点到所述待融合曲面上所有三角网格的投影点；
    方式二：针对所述待融合曲面对应的每个网格，确定该网格相对面的中心连线与所述待融合曲面上相应三角网格的交点，并将该交点作为该网格8个角点到所述相应三角网格的投影点；
    方式三：针对所述待融合曲面上每个三角网格，确定该三角网格的包络网格组的网格面与该三角网格的交点，并将交点作为该三角网格的包络网格组中各网格的8个角点到该三角网格的投影点。
  5. 如权利要求3或4所述的曲面融合方法，其特征在于，当存在到所述待融合曲面上所有三角网格均没有投影点的角点时，将该角点相邻角点的投影点作为该角点的投影点。
  6. 如权利要求1-5中任一项所述的曲面融合方法，其特征在于，根据所述第一TSDF值和所述第二TSDF值，确定融合后曲面对应的每个角点的第三TSDF值，包括：
    确定所述目标曲面的包络网格组，记为第三包络网格组，并记所述待融合曲面的包络网格组为第四包络网格组；
    针对所述第三包络网格组和所述第四包络网格组的公共部分的每个角点，根据该角点的第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF距离，并根据该角点的第一TSDF值中的TSDF距离和TSDF权重得到该角点的第三TSDF值中的TSDF权重；
    针对所述第四包络网格组相对所述第三包络网格组的新增部分的每个角点，将该角点的第一TSDF值作为该角点的第三TSDF值中的TSDF距离，并得到该角点的第三TSDF值中的TSDF权重为预设值。
  7. 如权利要求6所述的曲面融合方法，其特征在于，根据第一TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF权重，包括：
    在第一TSDF值中的TSDF距离小于预设截断距离时，根据第二TSDF值中的TSDF权重得到第三TSDF值中的TSDF权重；
    在第一TSDF值中的TSDF距离大于或等于所述预设截断距离时，将第二TSDF值中的TSDF权重作为第三TSDF值中的TSDF权重。
  8. 如权利要求7所述的曲面融合方法，其特征在于，根据第一TSDF值中的TSDF距离、第二TSDF值中的TSDF距离和TSDF权重得到第三TSDF值中的TSDF距离，包括：
    计算所述第二TSDF值中的TSDF距离和TSDF权重的乘积，得到第一乘积；
    计算所述第一TSDF值中的TSDF距离与TSDF的单位阈值的乘积，得到第二乘积；
    计算所述第一乘积与所述第二乘积的加和，得到第一和值；
    计算所述第二TSDF值中的TSDF权重与所述单位阈值的加和，得到第二和值；
    计算所述第一和值与所述第二和值的比值，得到所述第三TSDF值中的TSDF距离；
    其中，根据第二TSDF值中的TSDF权重得到第三TSDF值中的TSDF权重，包括：将所述第二和值作为所述第三TSDF值中的TSDF权重。
  9. 如权利要求6-8中任一项所述的曲面融合方法，其特征在于，根据角点的第三TSDF值和角点坐标，确定所述融合后曲面的三角网格顶点坐标和该三角网格的拓扑结构，包括：
    根据所述角点坐标和所述第三TSDF值中的TSDF距离，确定所述融合后曲面的三角网格顶点坐标；
    根据所述第三TSDF值中的TSDF距离确定立方体网格对应的nMC值，根据nMC值和三角网格的顶点坐标，确定所述融合后曲面的三角网格的拓扑结构。
  10. 一种曲面融合装置，包括存储器、处理器以及存储在存储器上的计算机程序，其特征在于，计算机程序被处理器执行时，实现如权利要求1-9中任一项所述的曲面融合方法。
  11. 一种医疗影像设备，其特征在于，包括如权利要求10所述的曲面融合装置。
PCT/CN2023/115899 2022-09-30 2023-08-30 曲面融合方法、装置和医疗影像设备 WO2024066891A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211212199.5A CN117853679A (zh) 2022-09-30 2022-09-30 曲面融合方法、装置和医疗影像设备
CN202211212199.5 2022-09-30

Publications (1)

Publication Number Publication Date
WO2024066891A1 true WO2024066891A1 (zh) 2024-04-04

Family

ID=90476047

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/115899 WO2024066891A1 (zh) 2022-09-30 2023-08-30 曲面融合方法、装置和医疗影像设备

Country Status (2)

Country Link
CN (1) CN117853679A (zh)
WO (1) WO2024066891A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130083019A1 (en) * 2011-09-30 2013-04-04 Hon Hai Precision Industry Co., Ltd. Computing device and method for determining distance between two curved surfaces
US20150178908A1 (en) * 2013-12-23 2015-06-25 A.Tron3D Gmbh Method for capturing the three-dimensional surface geometry of objects
CN109903372A (zh) * 2019-01-28 2019-06-18 中国科学院自动化研究所 深度图超分辨率补全方法及高质量三维重建方法与***
CN114782628A (zh) * 2022-04-25 2022-07-22 西安理工大学 基于深度相机的室内实时三维重建方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130083019A1 (en) * 2011-09-30 2013-04-04 Hon Hai Precision Industry Co., Ltd. Computing device and method for determining distance between two curved surfaces
US20150178908A1 (en) * 2013-12-23 2015-06-25 A.Tron3D Gmbh Method for capturing the three-dimensional surface geometry of objects
CN109903372A (zh) * 2019-01-28 2019-06-18 中国科学院自动化研究所 深度图超分辨率补全方法及高质量三维重建方法与***
CN114782628A (zh) * 2022-04-25 2022-07-22 西安理工大学 基于深度相机的室内实时三维重建方法

Also Published As

Publication number Publication date
CN117853679A (zh) 2024-04-09

Similar Documents

Publication Publication Date Title
US6952204B2 (en) 3D computer modelling apparatus
US6778173B2 (en) Hierarchical image-based representation of still and animated three-dimensional object, method and apparatus for using this representation for the object rendering
JP2009032122A (ja) 画像処理装置、画像処理方法およびプログラム
BR112019011254A2 (pt) método para gerar uma imagem de intensidade de luz, aparelho para gerar uma imagem de intensidade de luz, e produto de programa de computador
US6978230B1 (en) Apparatus, system, and method for draping annotations on to a geometric surface
JP2022509329A (ja) 点群融合方法及び装置、電子機器、コンピュータ記憶媒体並びにプログラム
KR20050063994A (ko) 경계셀 축소 방식을 이용하여 3차원 측정점들로부터 3차원메쉬를 생성하는 방법
US6977652B2 (en) Polygon rendering device
JPH11504452A (ja) ２次元投影図に基づいて３次元の対象物を再現し、取り扱うための装置と方法
US6476819B1 (en) Apparatus and method for assigning shrinkage factor during texture mapping operations
WO2024066891A1 (zh) 曲面融合方法、装置和医疗影像设备
JP4229398B2 (ja) ３次元モデリング・プログラム、３次元モデリング制御プログラム、３次元モデリング・データ伝送プログラム、記録媒体および３次元モデリング方法
JP2000348206A (ja) 画像生成装置および画像優先度決定方法
JP2002520969A (ja) 動き画像からの自動化された３次元シーン走査方法
KR102455843B1 (ko) 영상 재구성 장치 및 방법
TWI687086B (zh) 時間一致可靠度傳遞系統
CN113112606B (zh) 一种基于三维实景建模的人脸校正方法、***及存储介质
Carnielli et al. Fast digital perspective shell rendering
JP4292645B2 (ja) ３次元データの合成方法及び装置
JPH1139506A (ja) 任意視点画像生成装置
CN116109781B (zh) 一种三维重建方法及***
JP2002259945A (ja) ３次元モデル化方法
JPH07210688A (ja) 多角形面積算出装置
Zhang et al. Range image registration via probability field
JPH11259682A (ja) 画像処理装置および方法、並びに提供媒体

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23870131

Country of ref document: EP

Kind code of ref document: A1